SWP-281 the Singapore Area License Scheme Public Disclosure Authorized

Relieving Traffic Congestion: SWP-281 The Singapore Area License Scheme Public Disclosure Authorized

World Bank Staff Working Paper No. 281

June 1978 Public Disclosure Authorized Public Disclosure Authorized

The views and interpretations in this document are those of the authors and should not be attributed to the World Bank, to its affiliated organizations or to any individual acting in their behalf.

Prepared by: Peter L. Watson and Edward P. Holland Urban Projects Department

SLCO1 8287

The views and interpretations in this document are those of the authors and should not be attributed to the World Bank, to its affiliated organizations, or to any individual acting in their behalf.

L1B

WORLD BANK f>'- R 8

Staff Working Paper No. 281 5

June 1978

RELIEVING TRAFFIC CONGESTION:

THE SINGAPORE AREA LICENSE SCHEME

This paper reports the results of the research program carried out by the World Bank to identify and measure the effects of Singapore's Area License Scheme on the transport system in particular and the urban system in general. The specific areas dealt with are: traffic performance, travel behavior, the business community, conditions for pedestrians, air pollution, and public opinion and attitudes. Under each category, the research methods are described and the results presented. In addition, the paper investigates the potential for developing similar license schemes for other cities and sets out the available options and variations, the decisions to be taken, and the prerequisites for successful implementation.

Prepared by: Peter L. Watson and Edward P. Holland Urban Projects Department

PREFACE

To carry out a study of the impacts of the world's first use of road pricing to reduce congestion was, for the authors, the chance of a lifetime. If our enthusiasm led us to attempt to answer all possible questions, we have clearly failed. We believe, however, that we have succeeded in producing a reasonably comprehensive picture of what happened in Singapore as a result of the introduction of the Area License Scheme. We have tried to provide enough policy discussion to satisfy the policy-maker and enough technical detail to satisfy the technician. Inevitably, for any given audience, there will be too much of one sort of information and not enough of another. Overall, we trust that each reader will be able to find the level of detail that he or she needs.

One word of caution is in order. Both authors were favorably disposed toward road pricing when the study began. As the results emerged, they became more and more convinced that the type of scheme introduced in Singapore should be carefully examined in all transport studies. Indeed, not to include it would be equivalent to ignoring the potential of traffic lights. Thus, we do not apologize for any hint of a crusading tone that may have inadvertently crept into our report. The Government of Singapore has provided a challenge to all cities. We trust that the research that is reported in this book will result in other cities being better equipped to meet that challenge. It would be a source of great disappointment to the authors if this initiative were allowed to simply die away.

ACKNOWLEDGEMENTS

It is obvious that the authors could not have carried out a project of this size without the assistance of a great number of people. It is equally obvious that a complete list of the people who helped cannot be accommodated on this page. The authors, nonetheless, would like to express their gratitude to all who participated in this study.

Special thanks are due to:

-- the Government of Singapore, which took the bold step of introducing area licensing, without which the study would not have been possible;

-- the officials of-the Gov-iernmentNXof Singapore, especially Lim Leong Ge Gopinath Mennon,)Joseph Yee, and Tan Guong Ching, who provided-informat-ion, data, and advice far beyond the call of duty;

-- the World Bank Research Committee, the United Nations Environment Programme, the U.S. Department of Transportation, the U.S. Environmental Protection Agency, and the U.S. Federal Energy Administration, for the funds required to carry out the study;

-- our superiors at the World Bank--Kim Jaycox, Harold Dunkerley, Tony Churchill, Alastair Stone, and especially Clell Harral-- who maintained their support for the project, even in difficult times;

-- the many people who reviewed the report, especially W. Vickrey, M. Webber, H. Levinson, D. Bayliss, A. May, and our colleague Alan Walters;

-- our many consultants, particularly Bruno Wildermuth, Gordon Geraghty, John Roberts, Roger McGlynn, and Dick Worral;

-- our research assistants: Gokmen Erg"un, Khanh Nguyen, Carolyn Gochenour, Max Pulgar-Vidal, Brian Smith, Jane Kozlowski, and to Betty Easter for organizing and checking the final version of the manuscript;

-- and finally our wives and families, for putting up with our long absences and our obsession with Singapore.

The responsibility for the results presented and the views expressed herein, remains, of course, with the authors. The views expressed do not necessarily represent the official positions of the World Bank or the funding agencies.

TABLE OF CONTENTS

Page No.

1. INTRODUCTION AND PRINCIPAL FINDINGS ...... 1

Singapore's Area License Scheme ...... 3 The Research Project and Findings ...... 3 Achievement of the Objectives of the Area License Scheme ...... 11 Problems ...... *...... 14 Policy Implications ...... 15

2. THE AREA LICENSE SCHEME ...... 17

Background Information on Singapore ...... 17 Development of the Area License Scheme ...... 21 Details of the License Scheme ...... 23 Complementary Measures ...... 24 Expectations ...... -.-.... . o ..... o...o 28

3. IMPLEMENTATION, UTILIZATION, AND FINANCES ... ooooo. 29

Implementation ...... 29 Utilization o-o-...... - ...... -...... -...... o.....o.....o 33 Finances ... ooo.. ... o...oo ...... 37

4. TRAFFIC PERFORMANCE, PARKING, AND ACCIDENTS ... oo-_ .... 40

Traffic Flows ...... -.... -...... 40 Vehicle Speeds -...... -. oo ...... oo-....61 Parking ..... _oo_ ... o ...... 68 Accidents .-...... oo ...... 73

5. PRINCIPAL CHANGES IN TRAVEL BEHAVIOR ...... 77

The Samples .... oo...... - - 77 Household Trip Rates .. - ...... 78...... - 78 Trips from Vehicle-Owning Households to Work in the Restricted Zone oo..oo ...... -oo- 83 Trips from Vehicle-Owning Households to Work Beyond the Restricted Zone .o o#o...... o...... 97 Trips to Work from Non-Vehicle-Owning Households 102 Summary -- Trips to Work ...... o,- ... o ...... 102 Trips Home to Vehicle-Owning Households from Work in the Restricted Zone ...... oo-o...... 105 Trips Home to Vehicle-Owning Households from Work Beyond the Restricted Zone . oo- ...... o-...... - 107 Trips Home from Work to Non-Vehicle-Owning Households ...... o .... ooo ..... oo.112 - vi -

Page No.

112 Summary--Trips Home From Work ...... Trips to the Restricted Zone for Non-Work Purposes ...... 112 117 Trips Outside the Restricted Zone ......

122 ANNEX to Chapter 5 ...... 122 Micro-level Analysis of Changes in Mode ...... Supplementary Evidence on Trip Scheduling ...... 125 131 Changes in Travel Time ...... Variationp by Income Group ...... 136 Non-Vehicle-Owning Households ...... 142

148 6. FURTHER ANALYSIS OF TRAVEL CHANGES ...... o ......

148 Car Trips to the Restricted Zone ...... The Evening Peak ...... 150 Travel Time and Speed Measurements ...... 158 160 Motorists' Responses to Price Increases ... o...... The Level of the License Fee ...... 165

168 7. EFFECTS ON PEDESTRIANS ...... ees......

Design of the Study ...... 168 171 Traffic and Pedestrian Volumes ...... Pedestrian Delay ...... 184 Conflicts between Pedestrians and Vehicles ...... 191 191 Conclusions ......

194 8. AIR POLLUTION ...... a......

The Air Quality Monitoring Program ...... 194 198 Results ...... * ...... 206 Conclusions ...... *......

210 9. IMPACTS ON THE BUSINESS COMMUNITY ......

Reactions to the Traffic Restraint Scheme ...... 212 Assessment of Impacts on the Business Community ...... 219 227 Conclusions ......

10. PUBLIC OPINION AND ATTITDUES ...... 228

The Survey ...... 228 231 Basic Sample Characteristics ...... Importances ...... 233 237 Perceptions of Change ...... Relations Between Importances and Perceptions of Change ...... 244 Perceived Effectiveness of Transport Policies ...... 249 Perceptions of Impacts on Different Groups ...... 253 - vii -

Page No.

11. DESIGNING AN AREA PRICING SCHEME ...... 258

Specification of Objectives ...... 258 Choices in the Design of the Scheme ...... 259 Complementary Measures ...... 271 Essential Factors in Implementing the Scheme ...... 272 Other Contributing Factors ...... 273

APPENDIX 1: A Plan for the Relief of Traffic Congestion in the City ...... 275

TABLES

3.1 Infringements of Area License Scheme ...... 32 3.2 Issue of Area Licenses ...... 34 3.3 Utilization of Fringe Car Parks ...... 35 3.4 Average Daily Shuttle Bus Revenue ...... 36 3.5 Blue Arrow (Express) Bus Ridership ...... 37 3.6 Capital Costs ...... 0...... 38

4.1 All Motor Vehicles Entering Restricted Zone ...... 41 4.2 Cars, Including Car Pools, Entering Restricted Zone ...... 43 4.3 Taxis Entering Restricted Zone ...... 44 4.4 Car Pools Entering Restricted Zone ...... 46 4.5 Goods Vehicles Entering or Observed in Restricted Zone ..... 48 4.6 Buses Entering Restricted Zone ...... 50 4.7 Motorcycles and Scooters Entering Restricted Zone ..# ...... 51 4.8 Motor Car Fees and Taxes ...... 53 4.9 All Motor Vehicles Entering Restricted Zone ...... 54 4.10 Cars Entering Restricted Zone ...... 56 4.11 Non-Car-Pool Cars Entering Restricted Zone ...... 57 4.12 Car Pools Entering Restricted Zone ...... 57 4.13 Taxis Entering Restricted Zone ...... 58 4.14 Evening Outbound Traffic on Eight Roads ...... 59 4.15 Evening Outbound Traffic on All Roads ...... 60 4.16 Classification of Moving-Car Loops ...... 65 4.17 Summary Mean Speeds from Moving Car Data ...... 65 4.18 Traffic Speeds with (Actual) and Without (Estimated) Area License Scheme in Morning Commuting Hours ...... 66 4.19 Summary of Mean Bus Speeds from License Plate Observations . 68 4.20 Parking Charges ...... 69 4.21 Monthly Parking Ticket Sales ...... 71 4.22 Hourly Parking ...... 72 4.23 Duration of Parking at Hourly Rates ...... 73 4.24 Accidents by Type and Severity ...... 74 4.25 Time of Occurrence of Accidents ...... 76 4.26 Severity of Accidents ...... 76 - viii -

Page No.

5.1 Trip Rates by Purpose ...... 79 5.2 Trip Rates by Origin/Destination Pattern ...... 80 5.3 Reasons for Not Replacing Vehicle ...... 82 5.4 Trip Rates by Purpose ...... 82 5.5 Mode and Time Changes for Pre-ALS Car Drivers Whose Trips started Between 7:31 a.m. and 10:15 a.m...... 97 5.6 Starting Times of Trans-Zone Work Trips ...... 100 5.7 Principal Impacts of Traffic Policies on Work-Trip Behavior ...... 103 5.8 Starting Times of Shopping Trips to the Restricted Zone .... 113 5.9 Starting Times of Personal Business Trips to the Restricted Zone ...... 115 5.10 Modal Split: School Trips to Restricted Zone ...... 117 5.11 Micro-Analysis of Mode Changes for Trips to Work in Restricted Zone ...... 123 5.12 Mean Travel Times Before and After ALS ...... 133 5.13 Factors Potentially Influencing Changes in Travel Time ...... 134 5.14 Mode Changes by Income Group For Trips to Work in the Restricted Zone ....00...... 138

6.1 Timing of Car Trips To and from Restricted Zone ...... 150 6.2 Percentage Modal Usage -- A.M. and P.M. Post-ALS ...... 154 6.3 Proportions of Car Drivers Starting and Finishing Work, and Making Work Trips at Various Times ...... 156 6.4 Price Elasticities of Demand for Area Licenses ...... 161 6.5 Traffic Flows, Monthly Travel Costs, and Price Elasticities of Demand ...... 163 6.6 Elasticities from Estimated Demand Curve ...... 165

7.1 Analysis of "Other" Traffic Modes for Site 7 ...... 179 7.2 Pedestrian Crossing Flows ...... 182

8.1 Pollutants Monitored and Locations ...... 195 8.2 Carbon Monoxide Levels ...... 202

9.1 Respondents' Business Affiliations ...... 211

10.1 Reported Monthly Household Income ...... 232 10.2 Mean Importance Scores for Travel and Environmental Factors ...... 236 10.3 Mean Scores for Perceptions of Change ...... 239 10.4 Mean Weighted Perceptin-of-Change Scores ...... 246 10.5 Mean Scores for Attitudes to Effectiveness of Traffic Policies ...... 251 10.6 Mean Scores for Impact of Area License Scheme on Different Groups ...... 254 - ix -

Page No.

11.1 Implementation of Different Bases of Charging ...... 261 FIGURES

2.1 Map of Singapore (IBRD 13180) ...... 19 2.2 Central Singapore: Restricted Zone, Ring Road, and Fringe Car Parks (IBRD 13183) ...... *...... 25

4.1 Speed Measurement Loops (IBRD 13182) ...... 63

5.1 Mode Use for Work Trips to the Restricted Zone ...... 85 5.2 Typical Reported Starting Times ...... 89 5.3 Starting Times for Work Trips to the Restricted Zone, all Modes ...... 0...... *...... 90 5.4 Starting Time of Car Drivers' Work Trips to the the Restricted Zone ...... 92 5.5 Starting Time of Car Passengers' Work Trips to the Restricted Zone ...... 93 5.6 Starting Times of Carpools' Work Trips to the Restricted Zone .... * ...... 0...... 94 5.7 Starting Times of Bus Riders' Work Trips to the Restricted Zone ...... 95 5.8 Mode Use for Trips to Work Beyond the Restricted Zone from Vehicle-Owning Households ...... 99 5.9 Combined Mode, Time, and Route Changes ...... 101 5.10 Mode use for Trips Home from Work in the Restricted Zone ... 106 5.11 Starting Times for Trips Home from Work in the Restricted Zone ...... 0...... 108 5.12 Mode use for Trips Home from Work Beyond the Restricted Zone to Vehicle-Owning Households ...... 110 5.13 Starting Times for Trips Home from Work Beyond the Restricted Zone ...... 111 5.14 Mode Use for Shopping Trips to the Restricted Zone from Vehicle-Owning Households ...... 114 5.15 Mode Use for Personal Business Trips to the Restricted Zone from Vehicle-Owning Households ...... 116 5.16 Mode Use for Shopping Trips outside the Restricted Zone from Vehicle-Owning Households ...... 119 5.17 Mode Use for Shopping Trips Outside the Restricted Zone from Non-Vehicle-Owning Households *...... 120 5.18 Times of Starting Trip, Arriving, and Starting Work ...... # 126 5.19 Times of Ending Work, Starting Home, and Arriving .* ...... 128 5.20 Mode Use for Trips to Work in the Restricted Zone from Non-Vehicle-Owning Households ...... 143 5.21 Mode Use for Trips Home from Work in the Restricted Zone to Non-Vehicle-Owning Households ...... 144 Page No.

5.22 Starting Times for Work Trips to the Restricted Zone from Non-Vehicle-Owning Households ...... 145 5.23 Mode Use for Work Trips Beyond the Restricted Zone from Non-Vehicle-Owning Households ...... 146 5.24 Mode Use for Trips Home to Non-Vehicle-Owning Households from Work Across the Restricted Zone ...... 147

6.1 Changes in Morning Peak Car Trips ...... 152 6.2 Vehicle Flows Into and Out of the Restricted Zone ...... 157 6.3 Demand Curve for Car Trips to the Restricted Zone During Restricted Hours ...... 166

7.1 Map of Central Singapore, Survey Sites for Pedestrian Study (IBRD 13181) .*...... 171 7.2 Views from the Camera ...... 174 7.3 Hourly Variations in Traffic Flows at Each Site ...... *****...... 176 7.4 Total Pedestrians Crossing at Each Site ...... 180 7.5 Peak Pedestrian Volumes near Each Site ...... 183 7.6 Mean Delays to Those Delayed, at Signals ...... 186 7.7 Proportion of Pedestrians Delayed, at Signals ...... 187 7.8 Mean Delay to Those Delayed, not at Signals ...... 189 7.9 Proportion of Pedestrians Delayed, not at Signals ...... 190 7.10 Proportion of Pedestrians Challenged While Crossing, not at Traffic Signals ...... 193

8.1 Map of Central Singapore, Air quality Monitoring Sites (IBRD 13185) ...... 195 8.2 Carbon Monoxide Variations through the Day ...... 200 8.3 Oxides of Nitrogen in the Restricted Zone ...... 204 8.4 Oxides of Nitrogen Near a Ring Road Junction ...... 205 8.5 Smoke in the Restricted Zone and Near a Ring Road Junction ..*...... , ...... 207 8.6 Haze in the Restricted zone and Near Ring Road Junction ...... 208

10.1 Map of Central Singapore, Survey Sites for Public Opinion Study (IBRD 13184) ...... 229 10.2 Mean Importance Scores for the Whole Sample ...... 235 10.3 Mean Perception-of-Change Scores for Whole Sample ...... 240 10.4 Largest Perceptions of Change by Travel Group ...... 242 10.5 Perceptions of Change by Intra- and Extra-Zone Respondents . 243 10.6 Weighted Perception-of-Change Scores ...... 247 10.7 Overall Index of Perceived Change Weighted by Importance ... 248 10.8 Impacts of Policies on Different Groups ...... 255 CHAPTER 1. INTRODUCTION AND PRINCIPAL FINDINGS

The motor vehicle and its use as a private means of transport have profoundly affected the style and character of life in many countries by expanding the range of opportunities for work, recreation, and social activity. However, there are adverse effects that cannot be ignored. The increasing use of motor vehicles has contributed to environmental pollution, to balance-of-payments deficits for oil-importing countries, to a deteri- orating quality of life in city centers, and to congestion on limited urban road space. In recent years, it has become increasingly obvious that the economic, social, and environmental costs of urban road construction programs are unacceptably high, especially for cities in the developing world where resources are scarce. Moreover, even the largest road construction programs have failed to solve the problem of congestion, as demands for road space continue to grow faster than the space can be expanded.

Since increasing the supply has failed to balance demand and supply, greater consideration has recently been given to the possibilities of containing the demand by methods generally referred to in the transport planning profession as traffic limitation or restraint. These methods include preferential treatment for high occupancy vehicles, streets reserved for pedestrians and/or buses, imposing delays on drivers of private cars, and road pricing. Attempts to encourage people to ride public transport or to form car pools have met with limited success, as have efforts to make car travel less convenient by imposing detours and delays. Thus, transport planners have turned to the more direct approach of making people pay more for journeys in vehicles that use scarce street space inefficiently, specifically cars with only one or two people in them.

The use of road pricing to restrain inefficient use of motor vehicles, to conserve scarce capital resources and fuel, and to avoid adverse environmental impacts may also be viewed in purely economic terms. In other sectors of the economy, demand and supply are balanced by market determined prices. This mechanism does not function adequately in the transport sector, because motorists generally are not required to pay the full social cost of any particular trip, i.e., the costs imposed on others in terms of delay, air pollution, noise, etc. The result is that, for many car trips, the cost to the traveler is less than the benefit he gets from the trip, while the total cost to society exceeds the benefit. When this is so, trips take place that -- from the social viewpoint -- should not be made, contributing to congestion. Road pricing is a way of making the motorist recognize the entire social cost of any trip by requiring him to pay the difference between that cost and the cost he would otherwise perceive. This should inhibit him from making trips for which the social costs exceed the benefits.

Cars do not only interfere with other cars. In addition, free access to road capacity leads to a level of car usage that has adverse effects on public transport. A vicious circle develops in which congestion resulting from the increased use of cars leads to poorer public transport service and, - 2 - hence, reduced ridership. People give up using the unreliable bus service and drive cars instead. This increases the number of cars on the road, further degrading the public transport service, and contributing to deficits for public transport operators. In general, they respond by increasing fares or cutting service or both, which exacerbates the problem. Increased public transport efficiency could result from reducing the amount of private car usage and promoting the use of public transport.

Finally, it should be noted that the present system of underpricing road use generally favors those who can afford to purchase cars and has adverse impacts on poorer people who are obliged to use a public transport system with declining service levels.

The World Bank, given its activities in lending for both urban and inter-city roads, has long been interested in the subject of road user charges. This interest resulted in a significant contribution to the theoretical and conceptual literature in this area when the Bank commissioned Professor Alan Walters to carry out a study of the applicability of the basic concepts to developing countries. 1/ This work was followed by an internal Bank study of the application of road user charges in the countries of Central America. 2/ The Bank also joined with the Organization for Economic Co- operation and Development (OECD) to commission a survey of the problems of cities with rapidly growing automobile use, carried out by Wilfred Owen of the Brookings Institution. 3/

In practice, through its role-as executing agent or supervisor of land use and transport studies, the World Bank has encouraged planning teams in developing countries and their consultant advisors to examine the possibilities of using road pricing in conjunction with other measures in attempts to solve urban transport problems. When the Singapore government decided to introduce an area pricing scheme, it was seen in the Bank as an opportunity to augment the meager stock of factual knowledge in this field by making empirical observations before and after the scheme was initiated.

1/ Published as "The Economics of Road User Charges", World Bank Staff Occasional Papers, Number 5, The Johns Hopkins University Press, 1968.

2/ "Road User Charges in Central America", A. Churchill, World Bank Staff Occasional Papers, Number 15, The Johns Hopkins University Press, 1972.

3/ "Automobiles and Cites: Strategies for Developing Countries", Wilfred Owen, World Bank Staff Working Paper Number 162, or O.E.C.D. (Paris) Environment Directorate Working Paper Number 5. - 3 -

SINGAPORE'S AREA LICENSE SCHEME

Despite widely expressed interest in urban road pricing by transportation and urban management experts, a variety of potential problems -- political, administrative, and enforcement -- have generally deterred public officials from adopting this approach to urban transport problems. At the time of this writing, Singapore is the only city in the world that has implemented a road pricing scheme. The scheme was introduced in June, 1975. Its essence is that a special supplementary license must be purchased and displayed on any car that is driven into a designated Restricted Zone during the morning commuting hours. The license scheme was complemented by increased downtown parking rates, and measures were taken to provide improved bus service, including park-and-ride arrangements. The scheme, including the prices of daily and monthly licenses, exemptions for certain classes of vehicles, the hours of restriction, etc., is fully described in Chapter 2, which also presents background information on Singapore, explains the basic decisions that were made in designing the scheme, and states the effects it was expected to produce.

Implementation of the scheme went very smoothly. Contributing to this success were a well-conducted, year-long public information campaign, excellent planning and management of details, and quick recognition and correction of the few problems that did arise. 1/ Also helpful was the dis- position of Singaporeans to believe that their government was acting in the general social interest and to abide by the rules imposed.

Unlike most government actions to improve transportation, Singapore's Area License Scheme is profitable. Initially, revenues (equivalent to nearly US$200,000) were nine times current costs, and the rate of return on the investment (equivalent to US$2.73 million) was about 77 percent per year. In 1976, after the fee was increased, the rate of return approached 95 percent.

THE RESEARCH PROJECT AND FINDINGS

Before-and-after studies of the effects of transport policies are rare, because future policy changes are not usually known to researchers far enough in advance to permit the gathering of an adequate data base before the change takes place. Research staff from the World Bank faced a rare opportunity when they were given more than a year's notice of the introduction of Singapore's Area License Scheme. Given this warning, they were able -- in cooperation with the Government of Singapore and with support from the United Nations Environment Programme, the U.S. Environmental Protection Agency, and the U.S. Department of Transportation -- to design and carry out both an extensive program of data collection before the scheme went into effect and

1/ For more details, see Chapter 3. - 4 - a follow-up program once the scheme was in operation. This made before-and- after comparisons possible on many different impacts of the measures, including the effects on travel behavior, traffic performance, business activity, the environment, and public opinion.

The objective of the study was to gather and analyze information on the directions and magnitudes of as many as possible of the changes resulting from the Area License Scheme. This report is intended to provide decision makers and planners in other cities, especially in the developing world, with information on the effects of this innovative addition to existing urban transport planning and policy measures, so that they can judge its potential usefulness as a solution to their own problems, and understand how it might be adapted to their local conditions. No attempt was made to carry out a comprehensive economic evaluation of the scheme.

In planning the research, discussions were held with a wide variety of urban transport experts to identify the areas in which changes were likely to result from the introduction of the Area License Scheme. In addition to data on the operational and financial aspects of the scheme (summarized above), the following areas of potential impact were identified, and a program was designed to obtain data in each of these areas both before and after implementation of the scheme:

Traffic Performance Pedestrian Activity Travel Behavior Air Pollution Business Activity Public Opinions and Attitudes

Some of the data were made available by the Government of Singapore from administrative records or from its own monitoring programs. The remainder came from measurements and surveys conducted by the Bank.

For each of these areas, the basic features of the research methods and the major findings are set out below. Presenting the results in this abbreviated format permits an efficient exposition, but the reader is warned that these statements of findings are presented out of context and without the qualifications that are contained in the specific chapters dealing with the separate studies. Thus, it is recommended that any findings that the reader considers important or interesting be examined further by reading the appropriate chapter.

TRAFFIC PERFORMANCE 1/

Flows and Composition

The Singapore Government set as a primary objective a 25 to 30 percent reduction in the volume of cars entering the central area during the morning peak. In order to monitor the achievement of this objective, the Road

1/ For more details, see Chapter 4. - 5 -

Transport Working Group in Singapore's Public Works Department (PWD) set up a program using machine counters to record traffic flows at entry points to the Restricted Zone and at selected points on the ring road. This program was in operation intensively for one month before implementation and less intensively for six months after implementation. The Public Works Department also carried out manual counts to record traffic composition as well as flows on roads within, entering, and outside the Restricted Zone. The data were used not only for research, but also to monitor the implementation of the scheme and to identify desirable modifications.

The number of cars entering the Restricted Zone between 7:30 and 10:15 a.m. (the hours of restriction) fell by 73 percent from 42,790 in March 1975 to 11,363 in October, 1975. The proportion of these cars that qualified as car pools by carrying four or more occupants rose from less than 10 percent to 44 percent. The volume of cars entering during the half hour before 7:30 a.m. rose by 23 percent as some people started their trips earlier to avoid paying the Area License fee. Traffic volumes in the evening peak fell by only 2 or 3 percent.

Vehicle Speeds

One of the major benefits expected from a traffic restraint scheme is an increase in speeds resulting from reducted congestion. Attempts to develop measures of changes in speed by using cassette recorders to record license numbers, vehicle types, and the times of passing specific points were initially not successful. Thus, before-and-after speed measurements are not available. After implementation of the scheme, improved procedures were developed, and new observations were made both during the restricted hours and in the afternoon peak. The improved license-recording procedure was cross- checked against concurrent moving-car observations, and the mean speeds agreed well. From these data, inferences were made about the impact of the Area License Scheme.

The data on speeds are at best approximate, since the average speeds conceal wide variations caused by localized congestion. Overall, however, it appears that, during the hours of restriction, speeds in the Restricted Zone increased by about 20 percent to 33 kilometers per hour, speeds on inbound radial roads increased by about 10 percent, speeds on outbound radials did not change, and speeds on the ring road fell by about 20 percent to 20 km per hour.

Urban Goods Movement

A detailed assessment of the effects of the Area License Scheme on urban goods movements was judged to be unrealistic in the light of the resources available for the study. Rather, the research already planned was modified to highlight two indicators of whether or not the movement of goods in Singapore's central area had been facilitated.

Traffic flows into the Restricted Zone were disaggregated to show truck movements. Unfortunately, the doubling of truck volumes reported by - 6 - the Public Works Department is scarcely credible and at variance with other evidence. Thus, while truck volumes may have been increased slightly, there was probably not a large-scale increase in the number of trucks operating in the Restricted Zone during the restricted hours. On the other hand, the responses of respondents to the Business Survey, who were involved in ware- housing or distribution within the Restricted Zone, indicate that the movement of goods during the restricted hours has been greatly facilitated.

Overall, it seems reasonable to conclude that, while the volume of goods moved has not been affected by the Area License Scheme, goods can now be moved more easily than they could before the scheme's introduction.

TRAVEL BEHAVIOR 1/

The direct and immediate impact of the Area License Scheme was expected to fall particularly on car owners who formerly drove into the Restricted Zone. However, many other groups -- bus riders, pedestrians, and other motorists -- were also likely to be affected. Changes in the travel behavior of those directly affected by the scheme might also have widespread ramifications, affecting other groups of travelers. For example, buses might be faster but more crowded as a result of a shift away from the private car.

To provide an overall picture of the impacts of the traffic restraint scheme on travel behavior, it was necessary to obtain information on the characteristics of trips by users of all modes traveling into, through, or outside the Restricted Zone before and after implementation of the Area License Scheme. A household survey program was designed to obtain this information. A sample of households was surveyed before the Area License Scheme was introduced. Based on Registrar of Vehicle records and a local sampling frame, about 6,400 households were interviewed. A household informant gave information on the household; each person was then interviewed to obtain personal information and data on all trips made on the day preceding the interview. Interviewing extended from October 1974 to May 1975 with interruptions for school holidays and Chinese New Year -- periods when travel patterns in Singapore depart from normal.

A second phase of data collection was carried out from September to November, 1975, to obtain essentially the same data with the Area License Scheme in effect. In this phase, about 2,000 of the original households were re-interviewed. Analysis of the data revealed the following major changes.

Trip Rates

The rate of trip-making by members of vehicle-owning households to destinations in the Restricted Zone fell by 17 percent compared with 8 percent for trips made completely outside the Restricted Zone. In the specific case

1/ For more details see Chapters 5 and 6. of work trips, trip-making to destinations in the Restricted Zone fell 11 percent, while trip-making to work outside the Restricted Zone fell by 3 percent. Discretionary trips for shopping, personal business, social, and recreational purposes fell considerably more than work trips. It is believed that the general decline in trip rates resulted largely from factors other than the Area License Scheme, specifically recession and high taxes on vehicle ownership. The greater decline in trip rates to the Restricted Zone may be attributed partially to the Area License Scheme and partially to the existing decentralization trend.

There is no evidence that the Area License Scheme had any effect on trip-making by members of non-vehicle-owning households. Modal Changes

For trips to work in the Restricted Zone by members of vehicle- owning households, an important change in modal split was observed, from 56 percent car and 33 percent bus to 46 percent each by car and bus. Within the declining car mode, the share of trips made in car pools increased by a factor of three. Travelers from non-vehicle-owning households did not change their behavior appreciably; nearly 90 percent of them traveled by bus both before and after the Area License Scheme was put into effect. For work trips across the Restricted Zone, the car share fell from 52.5 percent to 50 percent and the proportion of car pools in the car share rose from 9 to 28 percent. Trip Schedule Changes

There was also an important change toward earlier starting times by non-car-pool car users to avoid having to pay the Area License fee. For work trips to the Restricted Zone, the proportions of trips started before 7:30 a.m. changed from 28 to 42 percent for car drivers and from 19 to 38 percent for car passengers. For work trips across the Restricted Zone, the proportions of trips started before 7:30 a.m. rose from 50 to 60 percent for car drivers and from 40 to 60 percent for car passengers.

Route Changes

For people making work trips across the Restricted Zone by car, detouring around the Restricted Zone to avoid paying the Area License fee was an important option. Before the introduction of the scheme, 88 percent of these trips passed through the zone. Afterwards, only 66 percent drove through the zone. Moreover, only 13 percent of these drove through during the restricted hours; the remainder changed time to avoid the fee.

Travel Time Changes

Neither the reported times from the household survey nor the results from the speed survey permit small changes in travel times to be measured. Thus, only the large changes can be identified. For car drivers (work trips to Restricted Zone) who changed to the bus after the introduction of the Area License Scheme, the mean travel time went up from 30 to 39 minutes. - 8 -

Conversely, for bus riders who changed to the car, the mean travel time fell from 38 to 29 minutes. For travelers who did not change mode, changes in travel times were negligible.

IMPACTS ON PARTICULAR GROUPS

Commuters by Car

For travelers driving to the Restricted Zone during the restricted hours and paying monthly rates for parking, the Area License Scheme and higher parking fees increased the monthly cost of commuting from S$153 (US$64) to S$228 (US$95).

Taxi Riders

Taxi riders suffered from the dramatically reduced number of taxis plying in the Restricted Zone during the restricted hours. Thus, it became more difficult to find a taxi and the cost of the trip usually increased, although the amount of the increase depended on each driver's ability to pass on the cost of the Area License fee to the passenger. This situation improved in April 1977, when the Government reduced the Area License fee for taxis by 50 percent. In general, riders in taxis that enter the Restricted Zone benefit from higher speeds, but may have to pay more.

Motorcyclists

As they were exempt from the Area License fee, motorcylists were under no obligation to modify their behavior. It is somewhat surprising that their share of work trips to the Restricted Zone remained stable. It had been expected that motorcycles might have been used more as a substitute for cars.

Shoppers

The number of shopping trips made to destinations in the Restricted Zone fell by 25 percent, but not all of this can be attributed to the Area License Scheme, since the number of shopping trips outside the zone also fell by 14 percent. The information gathered in the business survey indicates that the difference of 11 percent is due to the decentralization of jobs and residences and the increased parking charges associated with the Area License Scheme. It seems unlikely that the contribution of the increased parking charges could account for more than a small amount.

The people who still made shopping trips to the Restricted Zone made more of them by bus, and those travelling by car almost completely avoided the restricted hours. Users of all modes reported that shopping conditions in the Restricted Zone had improved markedly.

Central Area Residents

Central area residents reported increased ease and safety when crossing roads, improved general conditions in the Restricted Zone, and reduced fumes. - 9 -

People Making Trips Outside the Restricted Zone

Although the analysis has in general concentrated on the larger changes that have occurred in the behavior of people making trips to the Restricted Zone, it is useful at this point to note that over 70 percent of all trips and over 60 percent of work trips are made outside the Restricted Zone. The majority of trips to work outside the Restricted Zone are made by bus, and the majority of shopping trips are made on foot. The people making these trips have been unaffected by the Area License Scheme. Those motorists who use the ring road have been somewhat adversely affected by the increased congestion and reduced speeds.

BUSINESS ACTIVITY I/

When the Area License Scheme was announced, there was considerable speculation regarding the resulting impacts on the business community, especially on business located in the central area. The study team initially considered the idea of undertaking a quantitative study of changes in business activity, but, after examining data sources in Singapore, abandoned this idea. Instead, businessmen were interviewed in depth. Eight businessmen were in the "before" sample; the original eight plus sixteen more made up the "after" sample. The interviews were structured around a set of probe questions designed to elicit information on impacts on (a) level of business activity (sales, number of customers, etc.), (b) customer access to places of business, (c) employee access to places of business, (d) the general "climate for business". An attempt was made to have respondents set their comments in the context of the current Singapore economic situation, so that impacts due to the Area License Scheme could be distinguished from impacts due to other economic factors.

The members of the business community interviewed reported that the Area License Scheme had not had an important adverse impact on the business climate. However,-they believed that the increased parking charges had further depressed central area retail sales that were already suffering from recession and decentralization. They also believed that the restrictions on car travel to the center would accelerate the existing trend towards decentralization. In both cases, the Area License Scheme and the increase in parking charges were viewed as adding to existing problems and not creating new ones.

ENVIRONMENTAL IMPACTS

A consulting group prepared a report for the Bank on the various potential environmental impacts of the Area License Scheme and on the techniques that could be used for measuring and evaluating them. For each type of impact that was considered important, three questions about measurement were asked:

(a) What measurement techniques exist?

1/ For more details, see Chapter 9. - 10 -

(b) Are the changes expected in Singapore large enough to yield clear and significant differences in measurements?

The consultants were also asked to consider equipment availability, staff requirements, and data collection designs for various environmental measurements.

On the basis of these considerations, and after discussions between the consultants and the study team, it was decided to conduct a study of pedestrian movements, to call upon the Police Department for statistics on accidents, to encourage and assist the Singapore Anti-Pollution Unit in monitoring air quality, and to carry out a survey of public opinions and attitudes which would, among other matters, discover people's perceptions of any change in amenity and visual effects.

Measurements of noise were not undertaken because of the cost involved and because of doubts that the results would be meaningful. A proposal to study the energy implications was set aside on the grounds that it was peripheral to the main purposes of the study.

Pedestrian Conditions 1/

The consultants who had made the preliminary study of environmental impacts were commissioned to carry out the pedestrian study. It was decided that the critical feature of pedestrian movement was the crossing of streets, especially where no pedestrian facilities exist. Pedestrian delay in street crossing was identified as a suitable indicator of the impacts of the Area License Scheme, and reduced congestion, on pedestrians. Later, the probability of pedestrian-vehicle conflict was found to be a better indicator.

Time-lapse films were taken of 10 selected intersections or crossing sites at different times of day, before and after the Area License Scheme was put into effect. These films were analyzed to derive free-flow crossing time, average delay both in the street itself and at the curb (stratified by time of day and whether at the crossing site or within 50 meters of it). The films also yielded data on traffic flow and composition so that the delays to pedestrians could be analyzed in relation to traffic volume.

The level of pedestrian activity in the Restricted Zone has increased since the introduction of the Area License Scheme. While the pedestrian study provides some indications that pedestrian delays may have decreased during the restricted hours, the evidence is inconclusive. Reduced congestion, however, seems to have resulted in less pressure and, hence, more leisurely crossing behavior. In terms of conflict with vehicles, the pedestrian has clearly gained. There was a significant decrease in the proportion of pedestrians who had to take evasive action to avoid being hit by a vehicle.

1/ For more details, see Chapter 7. Air Pollution 1/

As an Anti-Pollution Unit exists in the Singapore government, the measurement of air pollution impacts was left to them. Their equipment was supplemented for this task by four carbon monoxide analyzers loaned by the U.S. Environmental Protection Agency. The Anti-Pollution Unit set up a program of before-and-after measurements on carbon monoxide, nitrogen oxides, and smoke at selected sites within the Restricted Zone and on the ring road.

Carbon monoxide levels in the Restricted Zone during the restricted hours were found to have been reduced to well below those found in the middle of the day or in the evening peak. Measurements of the other pollutants, being all-day total amounts, gave less clear-cut results.

PUBLIC OPINIONS AND ATTITUDES 2/

The data collection efforts described so far were concerned with the measurement of physical or behavioral changes. While these items are clearly important in measuring the impacts of the Area License Scheme, they tell the evaluator little about what the people of Singapore felt about the scheme. With this in mind, consultants were commissioned to survey and analyze the attitudes to the Area License Scheme of a sample of users of the central area, including pedestrians, motorists, and bus riders. The survey utilized psychometric techniques to probe (a) the extent to which people felt that conditions in the central area had changed, (b) which policies they believed had been most effective in bringing about these changes, (c) their attitudes toward the changes they perceived, and (d) whether they felt the scheme was fair in terms of who gained and who was faced with additional costs or inconvenience.

The public opinion survey revealed that Singaporeans believed that conditions in central Singapore had improved as a result of reduced congestion, and that the Area License Scheme had been the most effective of the policies responsible for the change. Pedestrians, residents, bus riders, taxi riders, and motorcyclists believed that they personally were better off as a result of the scheme. Motorists reported that they were worse off but not greatly so. All believed that the effect on Singapore as a city was favorable.

ACHIEVEMENT OF THE OBJECTIVES OF THE AREA LICENSE SCHEME

At the most general level, the objective of Singapore's transport planners was to contain the growth of traffic congestion in order to avoid its

1/ For more details, see Chapter 8.

2/ For more details, see Chapter 10. - 12 -

undesirable effects on road users and on the environment within the city. According to the publicity booklet distributed by the Road Transport Action Committee (reprinted as Appendix 1):

"Daily traffic congestion results in delay and frustration to motorists, bus commuters, goods and emergency vehicles, and poses danger to pedestrians and other road users. It also causes deterioration of the environment through noise, air pollution, and visual blight."

Before the introduction of the Area License Scheme, off-peak traffic volumes were about 25 percent lower than peak volumes and were considered to be acceptable. On this basis, the transport planners set a target of a 25 to 30 percent reduction in peak hour traffic volumes entering the Restricted Zone. This corresponded to a 50 percent reduction in the number of cars entering the Restricted Zone. The Area License Scheme achieved considerable success in reducing congestion in the central area, largely by inducing the expected shifts toward public transport and car pools, by spreading the peak, and by diverting through traffic.

In fact, it was more successful than planners and observers had expected, achieving a reduction in the number of cars of over 70 percent, and a reduction in total traffic volumes of over 40 percent. The achievement of a reduction in excess of the target has resulted in a high degree of underutilization of roads inside the zone, and observers have concluded that the price was set too high. The transport planners in Singapore, however, have not reduced the fee, believing that achieving their long-run objective of modifying attitudes is more important than achieving short- run economic efficiency. In short, having got people out of their cars, they have no intention of inducing them back again in the name of economic efficiency.

The target reduction in traffic volumes was intended to reduce "delay and frustration" and avoid continued "environmental deterioration." These benefits were not articulated in quantitative terms by the planners of the Area License Scheme, but, rather, in such terms as"... [being] freed of the burden and frustration of crawling through traffic jams in the city." At this subjective level, motorists in the public opinion survey reported large improvements in the level of congestion, traffic speeds, and travel times in the Restricted Zone during the restricted hours. Bus riders, taxi riders, and motorcyclists also reported improvements in these categories. Thus, transport users themselves perceive traffic conditions to have improved and the frustration associated with travel to the Restricted Zone to have been diminished.

At a more quantitative level, the speed study revealed a 20 percent increase in mean car speeds in the Restricted Zone during the restricted hours. This increase, however, was not reflected in the travel time changes derived from the household survey data. According to that source, car drivers, overall, lost small amounts of time and bus riders only gained small amounts. Thus, it may not be concluded that the Area License Scheme led to significant reductions in delay. However, the future time savings associated with - 13 -

keeping congestion levels constant rather than allowing them to get worse (for example, under the traditional "minimum intervention" assumption of economic analyses) are likely to be extremely large.

Air pollution levels in the Restricted Zone have been reduced. The likelihood that a pedestrian will have to take evasive action to avoid being hit by a car has been reduced. According to the public opinion survey, people perceive noise to have been reduced, safety when crossing roads to have been increased, general shopping conditions in the CBD to have improved, and the impact on Singapore as a city to be favorable. It may be concluded, therefore, that the Area License Scheme has had a positive effect on the environment.

It should be remembered that traffic congestion in central Singapore was by no means extreme before the introduction of the Area License Scheme, and the scheme was introduced to prevent the situation from becoming severe. To achieve the long-run goals, the planners perceived that, in addition to explaining the rationale for more widespread use of public transport and other high occupany vehicles, they would have to induce motorists to review and fundamentally change their attitudes towards the ownership and use of cars. This revision of motorists' attitudes and, hence, behavior was expected to reduce the problems caused by congestion and at the same time create an environment in which public transport services could be improved.

It is clear that the Area License Scheme forced motorists to modify their behavior, at least in the short run. While the high level of taxation on cars has stabilized car ownership levels, the Area License Scheme has reduced congestion in the central area, largely by inducing shifts toward public transport and car pools, together with shifts to earlier times and routes that avoid the Restricted Zone. Whether these are simply short-term behavior modifications or whether they really represent fundamental changes in the attitudes of motorists cannot be determined at this point. It seems likely, however, that the continued use of such measures will result in a more widespread acceptance (rather than mere tolerance) of public transport, car pooling, and other alternatives in the long run.

The benefits of meeting these general objectives cannot be valued in money terms, but the creation of a breathing space and a significant probability that Singapore can be prevented from becoming intolerably congested is clearly an important and valuable achievement.

In addition, two transportation studies between 1967 and 1975 showed that attempting to provide road capacity to serve unrestrained growth of traffic would have entailed the widening and upgrading of all main roads into the city by 1992, bringing some of them up to expressway standards. This, in turn, would have required demolition of many buildings in addition to the expenditures on road construction. The consultants and Singapore counterparts who carried out the transportation studies considered that the investments required for upgrading the road system were so obviously infeasible that they did not even estimate the cost of that alternative. - 14 -

PROBLEMS

While the Area License Scheme, by and large, achieved the primary objectives of its designers, some of its effects did not turn out as expected; some problems arose and had to be dealt with.

Level of the License Fee

With no precedent to guide them, the officials had to set the fee on the basis of an "educated guess". That they set it within the right order of magnitude showed very good judgment. Nevertheless, in terms of the target specified they set it somewhat too high. That was probably better than setting it too low, but it did pose the question of whether to reduce it or leave it alone. Choice of the latter alternative was justified in terms of conditioning people's attitudes so as to postpone future congestion problems.

The Evening Peak

Singapore officials, the Bank research team, and consultants all believed that if morning peak traffic were reduced by the targeted amount, evening peak traffic would also drop considerably. All were surprised when that did not occur. There were discussions of the possibility of imposing restraints of some kind in the evening, either on inbound or on outbound traffic. However, there seemed to be no way to do that without unfavorably affecting business -- for example, in the restaurant and entertainment sectors as well as retail trade. No good solution to this problem was found, and no restraints have yet been imposed in the evening.

Park and Ride

The designers of the scheme believed that former car drivers looking for a different mode of transport to avoid the license fee would be attracted by the possibility of driving to the outskirts of the central business district, parking, and taking a shuttle bus downtown, especially if they could be assured of a seat on the shuttle bus, a short wait, and a reasonably quick trip. On this basis, they spent S$6 million (US$2.5 million) preparing fringe car parks with capacity for 10,000 cars. This was the major expenditure connected with the scheme. They also awarded franchises for the operation of shuttle buses, which were restricted to operating on specified routes from the fringe car parks with frequent service, limited stops, and no standing passengers. With the projected demand, this would have been a profitable operation.

In the event, these park-and-ride arrangements were not attractive to Singapore drivers. About six percent of the parking space was used, and the shuttle buses were so underutilized that the operators were in immediate financial difficulties. The government acted quickly to revise the shuttle bus operations, making them essentially a supplement to the regular bus service on profitable routes where more capacity was needed. Over a longer period, they found new uses for much of the car-park space. - 15 -

It would be hard to predict how popular a park-and-ride scheme would prove in another city with different trip distances, car ownership patterns, and driver attitudes; hence its failure in Singapore should not be general- ized. Nonetheless, the Singapore experience indicates that such schemes are not necessarily popular, and, thus, should be carefully evaluated, and if possible pilot tested, before a full-scale operation is undertaken.

Other Public Transport Alternatives

Other attempts to induce drivers to abandon their cars by offering something better than the standard bus service included the introduction of express buses on a number of routes and the use of air-conditioned sightseeing coaches at commuting times. The express "Blue Arrow" service was priced at a flat S$0.50 compared with about S$0.40 for a similar trip on the standard bus. Surprisingly, the average number of passengers per trip on these buses, while slowly increasing, was still less than 30 a year after they were introduced. The air conditioned coaches, at S$1.00 per trip, carried only about 10 passengers per bus, with no upward trend.

Ring Road Traffic

Although improvements had been made to some segments of the ring road prior to implementation of the Area License Scheme, other parts proved inadequate for the unexpectedly heavy flows that developed as drivers to destinations beyond the Restricted Zone sought to bypass it. The initial problem was somewhat alleviated by retiming the traffic lights or directing traffic manually at critical intersections between radial roads and the ring road. Further relief will be provided by long-run projects, including construction along an outer ring route, which will divert some of the peripheral traffic from the existing ring road, and on one of the main one-way radial routes, which will improve conditions at a critical intersection.

Taxis

Whether taxis should be exempt from the license fee as public transport vehicles or subjected to it as equivalent to private cars in their use of road space has been a matter of debate among theorists. In Singapore, an answer was arrived at by trial and error. Initially, taxis were exempt, and they entered the Restricted Zone in such numbers that the government felt they were defeating attempts to reduce congestion. Accordingly, three weeks later, taxis were made subject to the license requirement. After that, it was almost impossible to find a taxi driver willing to travel into the central area during the morning peak unless the passenger first agreed to pay at least a substantial part of the cost of an area license. After nearly two years of this inconvenience, the situation was much improved by establishing a distinctive license for taxis at one-half the private car price.

POLICY IMPLICATIONS

The planners and decision-makers of Singapore have been shown to be justified in taking the bold step of implementing the world's first - 16 - area pricing scheme. Another question is whether the lessons learned in Singapore can be applied elsewhere. The Singapore experience demonstrates that -- under certain conditions -- pricing can be used effectively to reduce congestion in specific areas at specific times. Thus, it must henceforth be recognized as one of the tools of traffic management and must be considered for inclusion in any set of measures aimed at solving urban transport problems.

In any city, a variety of considerations must be taken into account in deciding whether area pricing would help attain the desired objectives. If pricing is to be used, the price, the method of collection, the area to be restricted, the types of vehicles to be included, the times of operation, and other particuiars have to be specifically chosen for the city in question. The last chapter in this report sets out the factors that should be considered in deciding whether to use area pricing and -- given a positive decision -- in working out the specific design for a particular city. The final sections summarize factors judged to have been either essential or favorable to the successful implementation of the Area License Scheme in Singapore. - 17 -

CHAPTER 2. THE AREA LICENSE SCHEME

This chapter sets out the background to the development of the Area Licensing Scheme, the process by which the scheme was designed, and the details of the scheme itself. 1/

BACKGROUND INFORMATION ON SINGAPORE

The Republic of Singapore comprises the main island of Singapore and some 54 small islands within its territorial waters (see Figure 2.1). It is situated a little over 150 kilometers north of the equator at the tip of the Malay Peninsula. The main island is diamond-shaped, measuring 42 kilometers from east to west at its broadest and 22 kilometers from north to south, with an area of about 584 square kilometers. The urban area of Singapore is located on the southern coast of this island, and approximately 70 percent of the population live within a radius of 8 kilometers from the center. A similar proportion of the island's Jobs are also located in this area. Urban development extends to the east and west along the southern coast. Scattered development follows roads leading to the north on either side of the water catchment area located in the central part of the island. The population of Singapore is about 2.25 million of whom about 74 percent are Chinese, 14 percent Malay, and 8 percent Indian. Singapore's strategic position in Southeast Asia has given it considerable economic importance. Its port is the fourth busiest in the world. An active industrialization policy has resulted in rapid growth, leading to a per capita annual income in 1975 of S$6000 (US$2500). 2/

TRANSPORTATION IN SINGAPORE

At the end of 1975, 280,378 motor vehicles were registered in Singapore. Of these, 143,155 were cars, representing a ratio of one car to 16 persons. Of the remainder, 4,585 were public service buses, 42,008 were goods vehicles, 5,388 were taxis, and 83,145 were motorcycles or scooters. From 1962 to 1973 the average annual growth rate of cars had been 8.8 percent per annum. In more recent years, government efforts to limit the growth of ownership of cars stabilized the fleet at its 1974 and 1975 level of about 143,000 cars, and by the end of 1976 the number had declined to 136,574.

1/ This chapter draws heavily on a paper presented to the OECD conference on "Better Towns with Less Traffic" in Paris in 1975 by Lim Leong Geok, Assistant Director of the Public Works Department and Secretary of the Road Transport Action Committee of the Singapore Government.

2/ During the period of the study the exchange rate varied around S$2.40 for one U.S. dollar. This value has been used, for the sake of consistency, in this report. - 18 -

During the period 1967 to 1974 two major transport studies were carried out in Singapore. The first was a comprehensive land use and transportation study with the objective of preparing long range plans to guide Singapore's development. The second was the first phase of a mass transit study which involved a more detailed examination of Singapore's public transport requirements in order to select a mass transit system. Both studies independently reached the conclusion that restraints on both car ownership and car usage would be necessary before 1992. This implied that radical changes would be required both in terms of the policies of the government and in terms of the attitudes of individuals towards car ownership and usage. Both would have to be re-oriented towards a more wide-spread use of public transportation.

Against this background, a transportation strategy for Singapore emerged. The measures adopted include the following:

(1) the adoption of land use development strategies to minimize the need to travel, for example, efforts to integrate labor-intensive industries in high density public housing estates and the provision of amenities, such as schools, recreation facilities, and shopping centers, within these estates;

(2) the implementation of a modest road construction program within the framework of a long range road plan. The latter has been kept to a scale considered compatible with the urban environment while still aiming to provide reasonable accessibility to public transport, commercial vehicles, and to a lesser extent cars;

(3) traffic management measures to promote better utilization of existing road capacity;

(4) rationalization and improvement of existing public bus services, including integration of private bus companies into a single company, major bus route revisions to provide more direct travel between major activity areas, expansion of the bus fleet, improved maintenance to reduce the proportion of the fleet off the road, government assistance to reorganize bus company management, and the introduction of 14 kilometers of reserved bus lanes in the central area to speed up bus travel;

(5) expansion of bus services for schools and the provision of supplementary bus services during peak hours by using the school bus fleet and other private buses to carry commuters; IBRD 13180 103o40' \ 103°45' 103050' 103055' DECEMBER 1977

J hore Bahru

1025- BuP.Mna UBIN 03 9 ee~~~~~~~ oon A! LeaASnapr YaS ItA ~~~~~~~~~~~~~~~~~~~~~~~Palau ./ Ch\_>ong P; zanrot

SNG I -APORE~~~~~MLASI

Ckn PanjangoBukit 7 2/

1~ ~20, B 3 < . iyae < urong er,oyorayedosmn rport~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~n 1030.10. 103045.~~~~~~ukiT0,0 ora c pan e o uc o n are. 10 O 0. 0 0 ig r .

'x~) -ieN emenscof uthereaderdanot Keng-Ch i hlS tg.s thed.n Railroads

World Bnk an, ito ats, Y an -* 10d'ment o03the leals tu ofn tes tio Internationalpory

- 21 -

(6) declaration of the government's intention to restrain the growth rate of car ownership and the implementation of this policy through increments in the various taxes on car ownership;

(7) revision of regulations and planning standards on car parking supply and use within the central area, designed to discourage commuter traffic while still catering to the needs of business and shopping for short time parking;

(8) prohibition of all large vehicles with three or more axles within the central area during peak hours; and

(9) a national campaign to promote and encourage staggered work hours and car pooling.

Transport planning in Singapore is facilitated by the fact that there is only one level of government for the island. This avoids the problems of overlapping jurisdictions between urban and suburban areas that are common in, for example, the United States. It also avoids conflicts over priorities among municipal, state (or county), and national decision makers. Within this framework, the Singapore Government set up a high level, inter- ministerial Road Transport Action Committee to coordinate transport planning measures and to formulate future policies. The committee is supported by technical level staff in the Public Works Department of the Ministry of National Development. Early in 1974 the committee considered the problems caused by growing traffic congestion in the central area, particularly during the morning and evening peak hours. The committee concluded that, following the recommendations of earlier studies, it would be appropriate to introduce restraints on the use of cars in the central area as soon as possible. The technical staff were, therefore, instructed to carry out a review of the methods available for achieving this goal.

DEVELOPMENT OF THE AREA LICENSE SCHEME

In the development of the Area License Scheme, the Singapore government decided that the following factors should be taken into account:

(a) Accessibility to and mobility within the central area is of paramount importance to the area's economic life and vitality. Thus, any control measures undertaken should minimize the number of people discouraged from entering the area either as commuters, shoppers, or businessmen. The widespread use of cars, particularly by people commuting to and from work, should, however, be discouraged. - 22 -

(b) It is only necessary to introduce measures aimed at discouraging the use of cars at specific times and in designated areas where congestion is heaviest. The mobility offered by the car need not be hampered unnecessarily in areas where and at times when traffic congestion does not exist.

(c) The method of control to be selected should be easy to implement from an administrative point of view and easy to enforce.

(d) Efficient, reliable, and attractive alternatives should be provided to commuters who are discouraged from driving into the city. As this will be an alternative mode of transport aimed at car owners, it should provide a better level of service in terms of frequency and comfort than that normally available to bus commuters. The diverted car owners can pay and should pay a higher fare for the better service provided.

Various traffic limitation measures were examined by the Road Transport Action Committee. Fiscal measures involving increases in import duties on cars, registration fees, and taxes on petrol had already been stringently applied to control car ownership, but it was recognized that they were not appropriate for regulating car travel at congested times and in congested areas. In the specific context of congestion, four measures were singled out for closer examination. They were vehicle metering, toll roads, parking fees, and area licensing. Vehicle metering was subsequently ruled out as it was judged that the technical difficulties of mass production and the problems of dealing with meters that were faulty or had been tampered with still remained to be resolved. Toll roads were considered difficult to implement in Singapore as high capacity toll stations could not be built within the limited space available along the major approach routes into the central area. Moreover, it was judged that the toll stations themselves would take up scarce urban land and indeed make significant contributions to the congestion they were intended to reduce. Increased parking charges were considered a practical method of reducing car travel in the central area. The fees would be easy to collect and parking could be regulated and priced to penalize long-term parkers and to favor short- term users, such as shoppers and people transacting business. This measure, however, was thought to suffer from the inability to discourage unnecessary through traffic and travel by chauffeur-driven cars, both of which are quite prevalent in Singapore. A fourth option, which would overcome these defi- ciencies, was area licensing. This would entail payment of an extra fee for the use of cars in the central area, whether parked there or not.

Of the measures of restraint examined, increased parking fees and area licensing were chosen as particularly appropriate for Singapore. It was felt that this combination would be adequate to deal with both traffic going to destinations in the central area and through traffic. - 23 -

DETAILS OF THE LICENSE SCHEME

The Area License Scheme requires that a special monthly or daily license be purchased and displayed on the windshield of any car entering a designated restricted zone during the morning commuting hours. Specific details are given below.

THE RESTRICTED ZONE

In delineating the boundary of the Restricted Zone the following

elements were taken into consideration:

(1) The zone should include all problem areas within the central area, that is, areas of traffic congestion where poor levels of service for both private and public transport prevail;

(2) The zone should have as few entry points as possible so that the policing can be kept to a minimum and the closing of minor roads will cause minimal disruption to local traffic;

(3) Where possible, bypass routes should be available so that traffic not wishing to enter the Restricted Zone may easily bypass it.

(4) To minimize administrative problems, the extent of the Restricted Zone should be relatively small.

In addition, in order to assist in the provision of a reliable alternative for drivers discouraged from using their cars, the location of existing car parks or vacant land which can be used for fringe parking should be taken into account.

The extent of the Restricted Zone is shown in Figure 2.2. It includes the whole of the central business district and a commercial area along the narrow corridor of Orchard Road northwest of the central business district. It covers an area of approximately 500 hectares (2 square miles).

THE HOURS OF RESTRICTION

The Road Transport Action Committee examined the alternatives of applying restrictions over the whole working day or only for part of it. Restrictions over the full working day were considered undesirable in that people transacting business or shopping in the area would be affected to the same extent as work commuters. This could produce adverse effects on the economic life of the area with an accompanying loss of vitality for the central area. Given that workers commuting by car were the primary targets - 24 - of the scheme, it was decided that the period from 7:30 a.m. to 9:30 a.m. should suffice. The advantage of keeping the period of restriction short is that staggered work hours might be induced, resulting in a spreading of the peak hour travel. In fact, this did take place, and congestion developed in the period from 9:30 a.m. to 10:00 a.m. After a few weeks, the time period was extended to 10:15 a.m. to eliminate the peak at the end of the restricted hours.

CATEGORIES OF RESTRICTED VEHICLES

The requirement to display an area license does not apply to buses or commercial vehicles, in order to favor public transport and maintain commercial activity. To encourage higher vehicle occupancy and more efficient use of road space, car pools (defined as cars or taxis carrying at least 4 persons) are exempt from the license requirement. Motorcycles are also exempt. These exemptions counter objections that driving into the center becomes a luxury that only the rich can afford; others can also do it if they form car pools or ride motorcycles. Initially, taxis were also exempt but after a few weeks of operation this exemption was removed. The only other exemptions are for police and military vehicles and emergency vehicles such as ambulances and fire engines.

THE AREA LICENSE FEE

As Singapore was the first city in the world to introduce a road pricing scheme of this type, it had no previous experience to guide it in setting the license fee. A panel of experts convened by the World Bank was unable to find any empirical data on the responses of motorists to large cost changes. Thus, it was necessary to set the fee by judgment. In such a situation, it is of course essential to be prepared to modify the fee by trial and error if it proves not to have been set correctly. The price of the license originally was set at S$60 a month or S$3.00 a day (US$25 and US$1.25 respectively). Monthly licenses are available from the Registrar of Motor Vehicles and selected post offices. Daily licenses are sold during the restricted hours at 16 curbside sale booths set up on the major approach routes into the Restricted Zone.

COMPLEMENTARY MEASURES

PARKING POLICY

At the time of planning the Area License Scheme and its complementary parking policy, there were approximately 20,000 parking spaces within the Restricted Zone, just over 50 percent of them privately owned. Parking fees were generally S$0.40 per hour; monthly rates varied between S$40 and S$60. As the heart of the central business district experienced higher traffic congestion and higher parking demand, a special "core" area was defined and special attention was focused on parking in this area. The IBRD 131.' Thin mtrpha boeenprepre-d by the WolBn'-s staff escl-iray Yfor the convenience of the readers of DU,,Ar the reporn to whoch it,0 attached. 0 Th, d ...... tonsdthede cedd o h ~~~~~~~~~50h~ ~ Rood~~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~b a. M-h_ on 0 s ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~WorldBank end Ito a5lie-, an ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ,s~~~~~~~~~~~~~~~~dgmnt onthe legal status f C S zfi 7 v F X& 3 s r i > ~ aeny ter-otory or any end-orement or accptanc of Ouch btoudarie:

0~~~~~~~~~~~~~~~~

KILOMETERS

CENTRAL SINGAPORE

/ ~. Ob~, Restricted Zone, ring road, and fringe car parks

tt \ \ p - ~~~~~~~~~~~~Roads

\o s e S _ ~~~~~~~~~~~~~Rivers

...... - Restricted Zone boundary Zot) V ~ ~~~~~~~~~~~~~~~~/ -- Ring road r =K.Kpel Road V°s * + ]/ , te° \,.v^t;'t O~~~~~~~~~Fringe car parks

Figure 2.2 JANUARY 1978

- 27 - monthly parking fees in publicly operated car parks within this core were increased by S$20 per space per month. Outside the core area but still in the Restricted Zone, fees were raised by S$10 per space per month. The result is that the monthly rate for all day parking in the central area has been increased from between S$40 and S$60 to between S$50 and S$80. Private car park operators are required to charge at least the same fees as public car parks in the same part of the zone. The government also levied a surcharge on private car park operators; the surcharge is equivalent to the price increase. Hourly parking fees in public car parks were also increased. In the core of the Restricted Zone, the new rates are S$0.50 for the first hour, S$1.00 for the second hour, and S$1.00 for each subsequent half hour. In the non-core area, the rates are S$0.50 for the first hour and S$0.50 for each subsequent half hour. These rates were designed to favor short-term parkers. 1/

THE PARK-AND-RIDE SCHEME

It was believed that a prerequisite to the imposition of restraints on car travel was the provision of an attractive alternative mode of transportation, i.e. an effective public transport system providing a high level of service. It was acknowledged that, other than in quasi-private forms such as taxis or special chartered buses, public transport in general cannot match the car in door-to-door service. However, it was believed that car commuters would expect a rapid, safe, comfortable, and reliable public transport service as an alternative to travel by their own car. When the scheme was planned, bus services in Singapore were judged to fall short of this expectation. Moreover, plans to improve the quality of bus services did not aim to achieve standards that would be attractive to motorists diverted from their cars. Thus, an alternative was planned including the provision of parking spaces close to the central area from which motorists could catch a shuttle bus to destinations within the Restricted Zone. As the bus service would only operate over short routes, it was expected that schedules could be well maintained. Moreover, the routes on which the shuttle buses would operate would have limited stops and only seated passengers would be carried in an attempt to provide a fast, comfortable alternative to the car.

For the fringe car park component of the scheme, it was believed that the car parks should be distributed around the periphery of the area in order to serve car traffic along all approach corridors. Ten sites were selected around the periphery of the zone and about seven and half thousand new parking spaces provided. In addition, a further two and half thousand spaces were provided in existing multi-story garages and in existing parking lots at the National Stadium. Eleven different shuttle bus services were provided to link the car parks with major destinations within the Restricted Zone. A fleet of 90 buses with a total capacity of 2,250 seats was provided. The combined monthly cost of parking and using the shuttle bus was set at S$30 (US$13) per month.

1/ For further discussion, see Chapter 4. - 28 -

PREMIUM BUS SERVICES

Two other public transport services designed to be of a high enough quality to appeal to ex-car-drivers were introduced after the Area License Scheme had been in effect for several weeks. One was the "Air-Conditioned Coach" service, using sightseeing coaches for an inbound trip in the morning and an outbound trip in the evening. The other, called the "Blue Arrows" was an express service during a two-hour morning period and a two and a quarter hour period in the evening with a 15-minute frequency. The Air-Conditioned Coach service was instituted on June 23, 1975, and the Blue Arrows on July 28, 1975. Fares were S$1.00 per trip for the former and S$0.50 per trip for the latter. Trips of comparable length on the standard buses, which have a zone fare system, cost from S$0.30 to S$0.40.

EXPECTATIONS

The Road Transport Action Committee expected to establish a workable system of central area pricing that would be acceptable to the public. It was thought that the scheme would also provoke motorists to review their attitudes to car ownership and usage and that once the rationale behind the need for more widespread use of public transport and other high occupancy vehicles had been accepted, the task of inducing more radical changes in travel habits could be undertaken. An examination of peak and off-peak traffic volumes and conditions at the time the scheme was planned indicated that the peak hour traffic volumes on central area streets were about 35 percent higher than off-peak volumes. As the off-peak operating conditions were in general considered to be satisfactory, the Road Transport Action Committee believed that the traffic restraints could be regarded as successful if peak hour traffic volumes were reduced by 25 to 30 percent. Although there existed no empirical basis for forecasting the reduction in peak hour traffic, it was believed that the range of available alternatives -- park- and-ride services, car pools, existing and new bus services, staggered work hours, and diversion routes -- represented a framework within which the Area License Scheme and the increased parking charges could produce a significant reduction in peak hour traffic volumes. The Committee had high expectations of achieving the target. - 29 -

CHAPTER 3. IMPLEMENTATION, UTILIZATION, AND FINANCES

IMPLEMENTATION

The Singapore Government started an extensive publicity campaign approximately one year before the Area License Scheme was actually introduced. A small booklet (reproduced as Appendix 1) was prepared and distributed to the public to explain the rationale for the scheme, the way in which it would operate, and the benefits that the government expected would accrue to the citizens. The publicity stimulated extensive coverage in the press, including frequent statements from government officials on the objectives and operational characteristics of the scheme. In addition, the press -- in its news, editorial, and letters-to-the-editor columns -- carried a continuing debate on the merits and problems of the scheme.

The actual implementation was carried out in three stages: parking charges in the Restricted Zone were increased on May 1, 1975; the "park- and-ride" scheme was inaugurated on May 16, 1975; and the Area License Scheme itself went into operation on June 2, 1975. The implementation was carried out very smoothly and no serious problems were observed.

INITIAL OBSERVATIONS

In this section are presented some of the basic changes that were observed during the first few weeks of operation of the Area License Scheme. They are intended to provide an overall introduction to the more detailed results that are presented in the rest of this report. The most immediately noticeable effect was the reduction in traffic flows within the Restricted Zone during the restricted hours. Congestion was virtually eliminated, buses moved rapidly, pedestrian movement was facilitated, and environmental conditions appeared to have improved. Preliminary counts carried out during the first week of operation indicated that the number of cars entering the Restricted Zone during the restricted hours had fallen by about 76 percent, and that the total traffic volume entering the zone had been reduced by about 35 percent. The number of vehicles entering the zone just before and just after the hours of operation was observed to increase. The preliminary data also indicated that both taxis and commercial vehicles were taking advantage of the reduced congestion by entering the zone in increased numbers during the restricted hours.

In contrast to conditions in the Restricted Zone, certain sections of the ring road, i.e., the streets that circumvent the Restricted Zone, were extremely congested. Specifically, Paterson Road, Scotts Road, and Newton Circus (adding up to a little less than a third of the ring road) were affected. This was believed to be mainly due to what had been through traffic trying to find its way around the zone. The situation was made worse by the fact that the traffic lights on the ring road were programmed to accommodate heavy inbound flows on the radial arterial roads. Traffic light timings were - 30 - adjusted during the first three days of operation and the problem was considerably alleviated. However, congestion remained at certain bottlenecks on the ring road.

Traffic flows were observed to build up after the restricted hours. In particular, traffic flows during the evening peak were very nearly the same as before the Area License Scheme was introduced.

The response to the shuttle bus park-and-ride services from the fringe car parks to the central area was extremely poor. During the two- week trial period in May, ridership was around 250 trips per day and for the first two weeks of the Area License Scheme it was only about 800 per day. The use made of the fringe car parks was correspondingly low.

The regular bus services provided by the Singapore Bus Service reported an increase in ridership of about 10 percent. They also believed that the reduced congestion enabled them to maintain schedules and provide a better service.

The exemptions granted to car pools (i.e., cars carrying 4 or more persons) resulted in a significantly higher number of car pools entering the Restricted Zone during the restricted hours. The increase appeared to be of the order of 80 percent.

PROBLEMS AND GOVERNMENT RESPONSES TO THEM

The government moved to identify problems and develop solutions with the speed, efficiency, and pragmatism that are characteristic of the Singapore Government.

Ring-Road Congestion

As was noted above, traffic lights were modified in three days to alleviate congestion. Longer-run actions include traffic engineering measures and some limited construction to improve traffic flows on the ring road and other affected routes.

Park-and-Ride

The unpopularity of the park-and-ride services resulted in severe financial problems for the shuttle bus operators, who had been given franchises for the new shuttle bus routes and services. The problem was particularly severe as some of the operators were small businessmen operating their own buses. On June 12, 1975, the routes of several shuttle services were modified and extended to serve housing estates beyond the fringe car parks. As the buses on the extended routes tended to be full when they reached the fringe car parks, some buses continued to start their routes at the car parks during morning peak hours. The shuttle buses were allowed to pick up and set down passengers at all bus stops, standing passengers were allowed, and the fare was reduced. The result is that the shuttle buses became almost indistingui- shable from the regular bus services. - 31 -

As a result of the changes, ridership rose from 800 trips per day to a profitable level of 2,500 trips per day. In compensation for losses initially incurred by operators, the shuttle bus license status was modified and S$74,717 in road taxes were refunded.

In addition, the government took steps to convert the unused fringe car parks to other uses, such as lorry parks, bus parks, and hawker stall sites.

Taxis

The restrictions on cars but not taxis resulted in a great increase in the number of taxis plying in the Restricted Zone. Afraid that they would undermine the effects of the Area License Scheme, the government on June 22, 1975 eliminated the Area License fee exemption that had been granted to taxi operators. The result was a dramatic decrease in the number of taxis entering the Restricted Zone during the restricted hours.

Hours of Operation

Although the Area License Scheme was successful in reducing the peak flow of vehicles into the Restricted Zone during the restricted hours, new peaks arose just before and just after that period. The traffic volume between 7:00 a.m. and 7:30 a.m. increased by 23 percent, but congestion was limited to ten to fifteen minutes as motorists tried to arrive just early enough to avoid the scheme, and it was considerably less than in the worst half hour previously. Thus, the government decided not to do anything about this early peak. Another peak developed just after the restrictions ended at 9:30 a.m. The government feared that this would affect retail stores and other businesses. Hence, on August 1, 1975, they extended the restricted hours to 10:15 a.m.

ENFORCEMENT

Enforcement proved not to be a problem. Advance warning signs were set up to advise motorists that they were approaching the Restricted Zone, and prominent signs mounted on overhead gantries were erected at all entry points. In addition, two amber lights and the words "In Operation" are illuminated during the hours of operation, to ensure that motorists are aware of precisely when the scheme goes into operation. Restricted categories of vehicles have to display the Area License on the top left-hand side of the front windscreen (in Singapore, traffic drives on the left).

Since the license sticker is large enough to be seen from the roadside, vehicles are not stopped for inspection. Moreover, as vehicle owners are required by law to furnish particulars on who was driving the vehicle at any given time, it is not necessary to stop offending vehicles to issue citations. Police officers simply record (either into a tape-recorder or manually) the vehicle license numbers of offenders for subsequent action. The offending motorist is issued a summons which, unless contested, requires - 32 - payment of a fine of S$50. Enforcement requires that police officers be stationed at each entry point, but the reductions in traffic noted above and the low level of illegal entries allow one or two officers to adequately control each entry point. In fact, the enforcing officers are not from Singapore's police force, but from CISCO -- a government owned firm special- izing in security systems (equivalent to Securicor or Brinks). It should be noted at this point that successful enforcement is not totally dependent on the specific local legal structure that permits citations to be issued without stopping the vehicle. If traffic volumes are reduced, it is possible to stop vehicles without adding to congestion. In the busiest corridor in Singapore, only about 15 cars per day would have to be stopped. This would not contribute to congestion, nor would it imply a significant increase in enforcement costs.

Just after the introduction of the scheme, about 200 infringements occurred each day. As drivers gained familiarity with the scheme, infringements quickly fell (see Table 3.1).

Table 3.1: INFRINGEMENTS OF AREA LICENSE SCHEME

(Daily Average) Total as % of Period Cars Taxis Total Licenses in Circulation

1975 June 132 12 144 2.8%

July 65 3 88 1.7% August 93 2 95 1.3% September 48 1 49 0.7% October 43 1 44 0.6% November 53 2 55 0.7% December 51 3 54 0.7%

1976 January 56 2 58 0.9% February 63 2 65 1.2% March 69 2 71 1.3% April 64 2 66 1.2% May 79 1 80 1.4% June 60 1 61 1.2%

July 61 1 62 1.2% August 74 2 76 1.5%

There was a small increase when the restricted hours were extended on August 1, 60) but after that the level stabilized at between fifty and eighty (average = per day. - 33 -

UTILIZATION

In this section, the sales of Area Licenses and the utilization rates of fringe car parks, shuttle buses, and premium bus services are presented.

SALES OF AREA LICENSES

Figures on the sales of Area Licenses are presented in Table 3.2. A certain variation in the level of sales can be detected in the first few months due to the introduction of restrictions on taxis and the extension of the restricted hours. After that, sales stabilized at a level of about 7,400 per month (calculated as monthly licenses plus the daily average number of daily licenses during the month).

At the end of December, 1975, the government announced increases in a number of vehicle taxes -- road taxes, registration fees, and the Area License fee. For private cars, the fee was increased by 33.3 percent from S$60 to S$80 per month (S$3 to S$4 per day). For company cars (Q-cars), the increase was from S$60 to S$160 -- double the new rate for private cars. The increases were intended to slow down the increase in ownership and use of private cars in general and to extend to the Area License Scheme the principle of taxing company cars at a higher rate. The result was a drop in sales of Area Licenses from about 7,400 per month to between 5,000 and 5,500 per month. After fluctuations in the early months of 1976, the level of sales stabilized at just over 5,000 per month. Before the price increase, the proportions of private and company car licenses sold were almost equal. After the price increase almost 2 private car licenses were sold for every company car license. Table 3.2: ISSUE OF AREA LICENSES

Monthly Daily (Average) Total (Average) Period Cars Q-Cars/a Taxis Total Cars Q-Cars/a Taxis Total Cars Q-Cars Taxis Grand Total

1975 June 4742 373 5115

July 4390 36 4426 462 223 685 4852 259 5111 August 5427 324 5751 897 449 1346 5874 773 7097 September 3771 1719 412 5902 10-30 430 1460 6520 842 7362 October 2740 2741 437 5916 841 253 456 1550 3581 2994 893 7468 November 2783 2684 387 5854 892 278 476 1646 3675 2962 863 7500 December 2733 2643 407 5783 935 274 456 1665 3668 2919 863 7448

1976 January 2574 2343 297 5214 650 107 343 1100 3224 2450 640 6314 February 2385 1661 198 4244 687 130 364 1181 3072 1791 562 5425 March 2427 1622 238 4287 652 114 316 1082 3079 1736 554 5369 April 2409 1584 246 4239 622 111 321 1054 3031 1695 567 5293 May 2659 1563 236 4458 666 112 318 1096 3325 1675 554 5554 June 2392 1520 217 4129 627 110 281 1018 3019 1630 498 5147

July 2350 1523 206 4079 631 95 286 1012 2981 1618 492 5091 August 2282 1470 213 3965 653 102 328 1083 2935 1572 541 5046 September 2341 1461 216 4018 656 98 297 1051 2997 1559 513 5069 October 2286 1444 205 3935 676 97 299 1072 2962 1541 504 5007 November 2446 1465 168 4079 679 99 301 1079 3125 1564 469 5158

/a Q-cars are cars registered as owned by businesses rather than individuals. Their license numbers begin with the letter "Q". - 35 -

UTILIZATION OF FRINGE CAR PARKS

The average numbers of cars parked in the fringe car parks are shown in Table 3.3.

Table 3.3: UTILIZATION OF FRINGE CAR PARKS

Period Average Daily Utilization

1975 July 636 August 525 September 545 October 558 November 388 December 431

1976 January 429 February 421 March 387 April 418 May 339 June 350

July 389 August 359 September 353 October 399 November 357

A discontinuity in the figures occurs in November, 1975, when the car park at New Bridge Road was converted to a local car park with hourly parking rates. Eliminating this site changed the statistical base of the series. After that the average utilization was 386 per day. A comparison of this figure with the 10,000 fringe parking spaces originally provided by the government reveals the extent of the unpopularity of the park-and-ride scheme.

UTILIZATION OF SHUTTLE BUSES

Ridership figures are not available on a continuing basis for the shuttle bus services. Average daily revenue figures are presented in Table 3.4. It was noted above that the initial ridership of the shuttle bus services was very poor, and that the Government authorized revised routes and fare structures similar to those of the regular city bus services. This step improved ridership somewhat and it was expected that ridership, and hence, the financial position of the operators, would improve further as the public became more familiar with the new services. The revenue figures show a steady increase that indicates that this expectation has been fulfilled. - 36 -

Table 3.4: AVERAGE DAILY SHUTTLE BUS REVENUE

Period Average Daily Revenue (S$)

1975 July 5,625 August 7,675 September 8,886 October 9,432 November 10,044 December 10,970

1976 January 11,745 Feburary 11,896 March 12,462 April 12,520 May 12,704 June 12,816

July 13,266 August 12,828 September 13,051 October 13,265 November 13,415 December 13,645

1977 January 13,997

UTILIZATION OF PREMIUM BUS SERVICES

Neither of the premium bus services attracted as many riders as were expected. The maximum monthly average ridership in the Air Conditioned Coaches was 150 in the morning and 100 in the evening (in August, 1975), averaging only ten persons per bus trip. In 1976, after the number of bus trips was reduced from 24 to 17, the service was used by about 130 people in the morning and only 50 in the evening, averaging 11 passengers per bus.

The Blue Arrow express service ridership figures are shown in Table 3.5. A gradually increasing trend is noticeable, with evening patronage generally 10 to 15 percent below that in the morning. The numbers of passengers per bus are well below capacity, on the average. - 37 -

Table 3.5: BLUE ARROW (EXPRESS) BUS RIDERSHIP

Number Passengers Number of Passengers of bus per Bus Trip Period a.m. P.M. Total Trips (Average)

1975 August 810 725 1,535 120 13 September 1,052 950 2,002 119 17 October 1,178 1,070 2,248 119 19 November 1,269 1,060 2,329 117 20 December 1,255 1,064 2,319 118 20

1976 January 1,378 1,180 2,558 118 22 February 1,458 1,241 2,699 118 23 March 1,508 1,308 2,816 126 22 April 1,442 1,296 2,738 115 24 May 1,502 1,367 2,867 115 30 June 1,551 1,367 2,918 116 25

July 1,594 1,418 3,012 112 27 August 1,611 1,409 3,020 108 28 September 1,614 1,388 3,002 108 28 October 1,585 1,389 2,974 108 28 November 1,702 1,434 3,136 123 25 December 1,756 1,359 3,115 127 25

1977 January 1,975 1,629 3,604 120 30

FINANCES

CAPITAL COSTS

The capital costs of establishing the traffic restraint scheme are presented in Table 3.6. The total was about S$6.6 million (about US$2.73 million). However, only 5 percent (S$317,290) of this sum was directly related to the construction and publicity associated with the introduction of the Area License Scheme. The majority of the costs were associated with the "park-and-ride scheme" -- 92 percent for the fringe car park construction and 2 percent for the shuttle bus operation. It should be remembered that, in the latter case, the government was only responsible for the planning of routes and for the construction of bus stops, shelters, etc. The costs of purchasing the buses themselves were borne by the private sector. It is also noteworthy that a small proportion of the costs attributed to the traffic restraint scheme cover the government sponsored campaign to promote staggered work hours and car pooling.

Overall, the total cost was less than it would have cost to build two kilometers of 4-lane urban expressway. - 38 -

Table 3.6: CAPITAL COSTS

Singapore Dollars /a

1. Fringe Car Parks

Construction of car parks, bus shelters, provision of utilities, and landscaping 5,751,743 Agency fees 273,092 Directional signs 28,082 Subtotal 6,052,917

2. City Shuttle Service

Consultant fee for plan of operation 21,000 Construction of bus termini, shelters, and bus stops 124,652 Subtotal 145,652

3. Area License Scheme

Overhead gantry structures at entry points 256,130 Supply of electricity 28,646 Booths for sale of licenses 30,864 Subtotal 315,640

4. Campaign on staggered work hours and car pooling 44,979

5. Printing of booklets for publicity 1,650

TOTAL S$6,560,838 (approximately US$2,734,000)

/a One U.S. dollar is approximately equal to S$2.40.

Source: Public Works Department of the Government of Singapore (March 1976).

NET FINANCIAL RETURN TO THE GOVERNMENT

Before the increase in the Area License fee, the government's monthly revenue from Area Licenses was S$471,827 (US$196,000) (average, September to December, 1975). After deducting monthly operating costs of about S$50,000 for fringe car park and license sales booth attendants, agency fees for license sales, and the CISCO officers used for enforcement, the net revenue was about S$420,000 (US$175,000) per month. This is a monthly net revenue of 6.4 percent of the capital costs presented above, or an annual rate of return of 76.8 percent. - 39 -

After the price increase, gross monthly revenue rose to an average level of S$568,082 (US$236,700)(average, February to December, 1976). With approximately the same operating expenses, the net revenue came to about S$518,000 (US$216,000) per month, equivalent to a net cash rate of return on capital of 94.8 percent per annum. - 40 -

CHAPTER 4. TRAFFIC PERFORMANCE, PARKING, AND ACCIDENTS

The principal impact of the Area License Scheme, of course, was on the flow of cars into the Restricted Zone during the hours of restriction. A reduction in car traffic during these hours might be expected to increase operating speed for vehicles of all types. Other effects might occur in flows of vehicles other than cars and in flows of cars just before and just after the restricted period. It was expected, also, that there would be a drop in homebound commuter traffic in the evening. This chapter deals primarily with observed flows and speeds of different types of vehicles. Also presented are data on parking and accidents.

TRAFFIC FLOWS

The Road Transport Working Group, a unit in Singapore's Public Works Department, monitored traffic flows before and after introduction of the Area License Scheme. Some of their traffic counts were done by observers, who manually recorded the types of vehicles and, on some occasions, the occupancy of cars. Other counts were made by machines, which automatically recorded the total numbers of vehicles without distinguishing types. The primary focus of the monitoring was on morning traffic entering the Restricted Zone, but some data were also collected in the afternoon, mainly on outbound traffic. The following sections present results on the impact of the traffic restraint package on morning traffic, the impact of the increase in taxes and the Area License fee, and an examination of the impacts of the scheme on afternoon traffic.

INBOUND MORNING TRAFFIC: INITIAL IMPACTS

All Motor Vehicles

Table 4.1 summarizes the flows of traffic into the Restricted Zone at three different stages of the Area License Scheme: in March, before the Park-and-Ride Scheme and the Area License Scheme were introduced, in July, the second month of Area License Scheme operation, and in September-October, an average for the second and third month after the Area License Scheme period was extended from two hours to two and three-quarter hours. The vehicles counted include cars, taxis, goods vehicles, buses, motorcycles, and scooters. The upper part of the table gives the number of vehicles entering in each of the time periods indicated; in the lower part, the volumes have been converted to hourly rates of flow so that they can be compared between periods of different lengths. The flow rate in the peak half hour is also given. The percentage differences from the March value are shown in parentheses.

As expected, there was a large drop in the number of vehicles entering the Restricted Zone during restricted hours. The 47 percent reduction during the 7:30-9:30 a.m. period in July relative to March was considerably - 41 -

more than the 25 to 30 percent target set by the government. The reduction of 44 percent in vehicles entering between 7:30 a.m. and 10:15 a.m. after the restricted period was extended implies less congestion in the downtown area and greater ease in finding parking space later in the day for those with business or shopping to do.

lot all of the vehicles that stopped entering the Restricted Zone between 7:30 a.m. and 9:30 a.m. were left outside. Some were driven into the zone before the restricted period, as indicated by the 17 percent increase in traffic between 7:00 a.m. and 7:30 a.m. There was also a small increase in the period immediately following the end of the restrictions at 9:30 a.m.

Table 4.1: ALL MOTOR VEHICLES ENTERING RESTRICTED ZONE

NUMIBER OF VEHICLES (in each time period indicated)

March 1975 July 1975 Sept.-Oct. 1975 Time of Day (before ALS) (ALS in effect (ALS in effect (a.m.) 7:30-9:30 a.m.) 7:30-10:15 a.m.)

Half-Hour Before Restricted Hours 7:00- 7:30 9,800 11,510 (+17%) 11,073 (+13%)

Restricted 7:30- 9:30 55,313 29,532 (-47%) -- Hours 7:30-10:15 74,014 -- 41,198 (-44%)

Half-Hour After 9:30-10:00 12,775 14,041 (+10%) -- Restricted Hours 10:15-10:45 -- -- 13,925

RATE OF FLOW (vehicles per hour, average during period)

7:00- 7:30 19,600 23,020 (+17%) 22,146 (+13%)

7:30- 9:30 27,657 14,766 (-47%) 15,747 (-43%)

9:30-10:15 24,935 25,443 (+2%) 12,938 (-48%)

10:15-10:45 -- -- 27,850

10:45-11:15 -- -- 25,378

Peak 1/2-hour 29,948 28,082 (-6%) 27,850 (-7%) (8:00-8:30) (9:30-10:00) (10:15-10:45)

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from March, 1975. -- Indicates data not available. - 42 -

With the suppression of the peak rate of flow into the restricted area, which had formerly occurred between 8:00 a.m. and 8:30 a.m., the half hour starting at 9:30 a.m. became the new peak, with a flow rate only 6 percent below the old one. Although this post-restricted-hours peak was of shorter duration than before, the government, nevertheless, tried to reduce it further by extending the restricted period to 10:15 a.m., as of August 1, 1975.

The result of extending the restricted period may be seen by comparing the last column of Table 4.1 with the middle one. The impact in the 9:30-10:15 a.m. period was comparable to that of the original scheme in the 7:30-9:30 a.m. period -- a 48 percent reduction in number of vehicles (or average flow) during the extension period. Some increase occurred in the 7:30-9:30 a.m. period, partly due, perhaps, to people who had been traveling after 9:30 a.m. to avoid the restrictions finding that course no longer effective and shifting to an earlier time. The result for the 7:30-10:15 a.m. restricted period was a reduction of 44 percent from the traffic volume in the same hours before restraints were initiated.

Again, the peak flow was shifted to the half-hour immediately after the end of the restrictions and only slightly reduced. No further efforts were made to eliminate the morning peak.

Car and Taxi Traffic and Occupancy

The data discussed above are for all motor vehicles added together. The Area License, however, is required only for cars and (since June 22, 1975) taxis. Thus, it is reasonable to suppose that the changes in total traffic figures are primarily reflections of proportionately greater changes in car and taxi traffic. Tables 4.2 and 4.3 confirm that supposition. - 43 -

Table 4.2: CARS, INCLUDING CAR POOLS, ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

March 1975 July 1975 Sept.-Oct. 1975 Time of Day (before ALS) (ALS in effect (ALS in effect (a.m.) 7:30-9:30 a.m.) 7:30-10:15 a.m.)

Half-Hour Before Restricted Hours 7:00- 7:30 5,384 6,685 (+24%) 6,640 (+23%)

Restricted 7:30- 9:30 32,421 8,130 (-75%) -- Hours 7:30-10:15 42,790 -- 11,363 (-73%)

Half Hour After 9:30-10:00 7,059 7,082 (+0.3%) -- Restricted Hours 10:15-10:45 -- -- 6,642

RATE OF FLOW (vehicles per hour, average during period)

7:00- 7:30 10,768 13,370 (+24%) 13,279 (+23%)

7:30- 9:30 16,211 4,065 (-75%) 4,673 (-71%)

9:30-10:15 13,825 12,508 (-10%) 2,691 (-81%)

10:15-10:45 -- -- 13,285

10:45-11:15 -- -- 11,322

Peak 1/2-hour 18,428 14,164 (-23%) 13,285 (-28%) (8:00-8:30) (9:30-10:00) (10:15-10:45)

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from March, 1975. -- Indicates data not available. - 44 -

Table 4.3: TAXIS ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

March 1975 July 1975 Sept.-Oct. 1975 Time of Day (before ALS) (ALS in effect (ALS in effect (a.m.) 7:30-9:30 a.m.) 7:30-10:15 a.m.)

Half-Hour Before Restricted Hours 7:00- 7:30 1080 851 (-21%) 854 (-21%)

Restricted 7:30- 9:30 7397 1248 (-83%) -- Hours 7:30-10:15 10923 - 3787 (-65%)

Half Hour After 9:30-10:00 2452 2372 (-3%) -- Restricted Hours 10:15-10:45 -- -- 2643

RATE OF FLOW (vehicles per hour, average during period)

7:00- 7:30 2160 1702 (-21%) 1708 (-21%)

7:30- 9:30 3699 624 (-83%) 1252 (-66%)

9:30-10:15 4701 4476 (-5%) 1712 (-65%)

10:15-10:45 -- -- 5286

10:45-11:15 -- -- 5092

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from March, 1975. -- Indicates data not available. - 45 -

Almost every sizable change registered for "all motor vehicles" entering the Restricted Zone corresponds to a percentage change in both car and taxi volumes that is from 1.4 to 2.0 times as large. For example, the 47 percent drop in all-vehicle volume in the 7:30-9:30 a.m. period from March to July (Table 4.1) corresponds to a 75 percent drop in cars (Table 4.2) and an 83 percent drop in taxis entering the Restricted Zone (Table 4.3).

The exception is taxis entering in the 7:00-7:30 a.m. period. The fact that the volumes of taxis and automobiles did not move in parallel in the pre-7:30 a.m. period (when taxi entries dropped 21 percent while car entries rose by 24 percent) is probably explained by the fact that many taxi drivers, discouraged by the license scheme, simply did not take to the road to do business until after the restricted hours.

Since the license requirement is waived for car pools -- defined as cars with four or more occupants -- one of the options available to a car driver is to carry at least three passengers in order to be able to drive into the city during restricted hours without buying a license. That this option was widely used is evident from the absolute numbers of car pools and the percentage of car entries that they constituted, as shown in Table 4.4.

After the initial introduction of the Area License Scheme, the most striking fact is that the absolute number of cars with car pool occupancy entering the Restricted Zone between 7:30 a.m. and 9:30 a.m. went up by 53 percent. At the same time, the total number of cars fell to one-fourth of its former level. The proportion of cars carrying four or more occupants thus jumped from about 7 percent 1/ to 44 percent for this period. In the 7:00- 7:30 a.m. period, the number of car pools did not change much, while the total number of cars rose 24 percent. Thus, the percentage of car pools dropped slightly.

The large increase in the number of car pools in the 9:30-10:00 a.m. period (with almost no change in total number of cars) is a phenomenon that cannot be ascribed to the Area License Scheme. It must be presumed that the campaign to encourage car pooling, combined with increased parking charges, road taxes, etc., was inducing more car pooling independently of the license scheme. Nevertheless, it is clear that the increase was proportionately greater during the restricted hours than after, and the pattern of car pools as a percentage of all cars is quite clear, going from 10 percent to 44 percent and down again to five, for the pre-restricted, restricted, and post- restricted hours, respectively.

After the extension of the restricted hours to 10:15 a.m., still more car pools appeared, but not in the same proportion. The proportion in the period from 7:30-10:15 a.m. became 37 percent, still high compared with 11 percent just before and 5 percent just after the restricted hours.

1/ Although the percentage figures before the Area License Scheme are not known precisely (data on pools and on total cars are not available for the same month), it would not change the picture much if the percentages shown for May were in error by ± 50 percent. - 46 -

/a Table 4.4: CAR POOLS- ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

May 1975 July 1975 Sept.-Oct. 1975 Time of Day (before ALS) (ALS in effect (ALS in effect (a.m.) 7:30-9:30 a.m.) 7:30-10:15 a.m.)

Half-Hour Before Restricted Hours 7:00- 7:30 687 672 (-2%) 702 (+2%)

Restricted 7:30- 9:30 2334 3581 (+53%) -- Hours 7:30-10:15 -- 4083 4217

Half Hour After 9:30-10:00 258 332 (+29%) -- Restricted Hours 10:15-10:45 -- -- 352

PERCENT OF CARS ENTERING

7:00- 7:30 13# 10 11

7:30- 9:30 7# 44 ) ) 37 9:30-10:15 4# 5 )

10:15-10:45 -- -- 5

/a Car pools are defined as cars with 4 or more occupants; they are exempt from the license requirement.

Source: Public Works Department, Singapore.

Notes: -- Indicates data not available. # Estimated from March data for all cars and May data for car pools. The figures in parentheses are percentage changes from May 1975. - 47 -

Goods Vehicles

In addition to cars and taxis, the total vehicle volumes include goods vehicles, buses, motorcycles, and scooters. None of these categories is subject to the area license requirement. During peak hours, lorries (trucks) with three or more axles are prohibited from entering or moving within the Inner Ring Road System, which encloses the Restricted Zone plus other less congested areas. The number of vehicles in this category, however, is relatively small.

After the Area License Scheme was instituted, it was reported by the Public Works Department that inbound movements of goods vehicles during the restricted hours had greatly increased, presumably because operators were rescheduling trips to take advantage of the reduced congestion. Data from the Public Works Department counts (summarized in Table 4.5) indicated that the flow of goods vehicles into the Restricted Zone between 7:30 a.m. and 9:30 a.m. had risen by 126 percent between March and July 1975, and had remained close to the new level in September and October.

A more complete examination of the data, however, showed an increase from March to July of 90 percent during the half-hour just before the restricted hours, and of 112 percent during the 45-minute period just after the restricted hours. These large increases in goods-vehicle flows during periods of increased car traffic just outside the restricted hours are contrary to expectations and inconsistent with the hypothesis that goods vehicle trips were being rescheduled into the less congested restricted period. It is difficult to find any plausible explanation for an approximate doubling of this traffic during most of the morning, and it seems reasonable to suspect an error in the data of the Public Works Department.

Goods vehicle flows reported for August, September, and October remained close to the July levels except for a decline for the 7:00-7:30 a.m. period to 57% above the March figure. Thus, it is unlikely that the figures are in error for the months with the Area License Scheme in effect. It is conceivable that some error -- possibly a misplaced factor of two -- may have occurred in the basic calculation of the March figures. However, extensive checks have not revealed any error either in the original data or in values calculated therefrom. - 48 -

Table 4.5: GOODS VEHICLES ENTERING OR OBSERVED IN RESTRICTED ZONE (all figures shown as hourly flow rates)

I. P.W.D. Data: All Goods Vehicles Entering:

Time (a.m.) March 1975 July 1975 Sept.-Oct. '75

7:00-7:30 1762 3346 (+ 90%) 2810 (+ 59%) 7:30-9:30 1572 3557 (+126%) 3523 (+124%) 9:30-10:15 1937 4104 (+112%) 4075 (+110%)

II. TEST Data: Mean Flows at Each of Six Locations:

Medium and Heavy Goods Vehicles -- Light Vans Excluded

Time (a.m) March-April 1975 Feb.-March 1976

7:30-7:45 123 130 (+ 6%) 8:30-8:45 117 141 (+21%) 9:30-9:45 167 169 (+ 1%)

Note: The figures in parentheses are percentage changes from March or March-April 1975.

Some independent evidence is found in vehicle counts made during the pedestrian study at the sites where pedestrian movements were being observed. 1/ These are not data on entries into the Restricted Zone, but flows of different types of vehicle past points within the zone. A doubling of the rate of entry would certainly be reflected in a large increase in the flows observed. The counts were made in March and April 1975, and in February and March 1976 -- not the months for which the Public Works Department data are presented, but nevertheless a before-and-after comparison. The results, summarized in Table 4.5, show no large increase in goods vehicle flows at the locations monitored for the pedestrian study. Thus, the available data on goods vehicle movements are conflicting, and no firm conclusions are justified. It must be noted that the Public Works Depart- ment data on goods vehicles are included as part of the totals for all motor vehicles in Table 4.1, raising the possibility that the March totals may be understated. Doubling the goods vehicle component in that month, however, would not greatly alter any findings reported on the basis of that table.

1/ See Chapter 7. Figure 7.1 shows the locations where observations were made. - 49 -

Buses

It was not expected that the Area License Scheme would have any impact on the flows of buses into the city, inasmuch as the flows are essentially determined by the number of buses available and their routes and schedules. Table 4.6 presents the data on bus flows from the Public Works Department counts, which show only small changes except for the pre- 7:30 a.m. half-hour in July. The 20 percent drop during that period is unexplained, but the fact that the September-October level is close to that of March arouses some suspicion about the July figure.

Motorcycles and Scooters

Table 4.7 presents the data on motorcycles and scooters entering the Restricted Zone during the morning. With cars subject to the Area License fee and motorcycles and scooters exempt, an increase in use of the latter might have been expected. It would appear that a small increase took place in July during the restricted hours, but by September and October that increase had disappeared, and the number entering between 7:00 a.m. and 7:30 a.m. which had remained the same in July as in March, had dropped by twelve percent. It is surprising that these vehicles were not more used as a substitute for cars. - 50 -

Table 4.6: BUSES ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

March 1975 July 1975 Sept.-Oct. 1975 Time of Day (before ALS) (ALS in effect (ALS in effect (a.m.) 7:30-9:30 a.m.) 7:30-10:15 a.m.)

Half-Hour Before Restricted Hours 7:00- 7:30 750 599 (-20%) 717 (-4%)

Restricted 7:30- 9:30 3181 2982 (-6%) -- Hours 7:30-10:15 3980 -- 4122 (+4%)

Half Hour After 9:30-10:00 545 561 (+3%) -- Restricted Hours 10:15-10:45 -- -- 479

RATE OF FLOW (vehicles per hour, average during period)

7:00- 7:30 1500 1198 (-20%) 1434 (-4%)

7:30- 9:30 1591 1491 (-6%) 1662 (+4%)

9:30-10:15 1066 1035 (-3%) 1066 (0%)

10:15-10:45 -- -- 958

10:45-11:15 -- -- 942

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from March 1975. -- Indicates data not available. - 51 -

Table 4.7: MOTORCYCLES AND SCOOTERS ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

March 1975 July 1975 Sept.-Oct. 1975 Time of Day (before ALS) (ALS in effect (ALS in effect (a.m.) 7:30-9:30 a.m.) 7:30-10:15 a.m.)

Half-Hour Before Restricted Hours 7:00- 7:30 1705 1702 (0%) 1459 (-14%)

Restricted 7:30- 9:30 9170 10057 (+10%) -- Hours 7:30-10:15 11724 - 11824 (+1%)

Half Hour After 9:30-10:00 1725 1785 (+3%) -- Restricted Hours 10:15-10:45 -- -- 1807

RATE OF FLOW (vehicles per hour, average during period)

7:00- 7:30 3410 3404 (0%) 2918 (-14%)

7:30- 9:30 4585 5029 (+10%) 4639 (+1%)

9:30-10:15 3405 3320 (-3%) 3396 (0%)

10:15-10:45 -- -- 3614

10:45-11:15 -- -- 3438

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from March 1975. -- Indicates data not available. - 52 -

INBOUND MORNING TRAFFIC: IMPACT OF INCREASED FEES

On December 31, 1975, the Government of Singapore increased certain motor car taxes and fees as part of the program to discourage ownership and use. Registration fees and annual road taxes were raised sharply. In addition, the Area License fee for individually-owned cars and taxis was increased by one-third and that for cars registered by companies was increased to twice the fee for private cars, a 167 percent increase. 1/ The old and new rates are shown in Table 4.8.

Tfie impacts of the new Area License fees (together with other tax and fee increases) are shown in the next set of tables, comparing morning traffic inflows before (September-October 1975, average) and after (February to May 1976, average) the change in fees. Data for the month of August 1976, are also included to indicate whether there is any trend.

Table 4.9 shows the numbers of motor vehicles of all types entering the Restricted Zone during the half-hour preceding the restricted period, during the restricted period itself, and during each of the next two half- hour periods. In the aggregate, there was a drop of six percent in the restricted period flow for the February-May period, but by August 1976 the flow had come back to the late-1975 average. These changes are evidently not simple responses to the increased license fees, since they were roughly paralleled by changes in the flows before 7:30 a.m. and in the two periods after 10:15 a.m., when no license is required.

1/ Of course, the real company car rate is lower to the extent that companies can take an income tax deduction for the car. - 53 -

Table 4.8; MOTOR CAR FEES AND TAXES

AREA LICENSE FEES Before Dec. 31, 1975 After Dec. 31, 1975

Private Cars

Daily S$3 (US$1.25) S$4 (US$1.67)

Monthly S$60 (US$25) S$80 (US$33)

Company Cars

Daily S$3 (US$1.25) S$8 (US$3.33)

Monthly S$60 (US$25) S$160 (US$67)

ANNUAL ROAD TAX /a (cars registered to private individuals)

Engine Size

1000cc S$ 200 (US$83) S$ 350 (US$146)

1500cc S$ 375 (US$156) S$ 600 (US$250)

2000cc S$ 600 (US$250) S$ 900 (US$375)

2500cc S$1,000 (US$417) S$1,250 (US$521)

3500cc S$2,275 (US$948) S$2,800 (US$1,167)

INITIAL REGISTRATION FEE 55% ad valorem 100% ad valorem

/a The official rates are expressed as a fee per cubic centimeter. The above figures are examples from each of five cubic capacity groups. The road tax for company cars is twice that shown. - 54 -

Table 4.9: ALL MOTOR VEHICLES ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

/a Time (a.m.) Sept.-Oct. 1975 Feb.-May 1976 Aug. 1976

7:00-7:30 11,073 10,527 (-5%) 11,433 (+3%)

7:30-10:15 (ALS Period) 41,198 38,788 (-6%) 41,250 (+0%)

10:15-10:45 13,925 13,158 (-6%) 13,759 (-1%)

10:45-11:15 12,689 11,986 (-6%) 12,982 (+2%)

RATE OF FLOW (vehicles per hour, average during period)

7:00-7:30 22,146 21,054 (-5%) 22,866 (+3%)

7:30-10:15 (ALS Period) 14,981 14,105 (-6%) 15,000 (+0%)

10:15-10:45 27,850 26,316 (-6%) 27,518 (-1%)

10:45-11:15 25,378 23,972 (-6%) 25,964 (+2%)

/a Excluding April 1976.

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from September- October 1975. License fees changed on December 31, 1975. See Table 4.8. - 55 -

If indeed, there were any changes due to the increased license fees they would, of course, show up more sharply in the flows of vehicles for which the license is required, i.e. cars and taxis (with less than four occupants). Table 4.10 shows the entries into the zone by cars. It is somewhat surprising that the relative changes in these flows were only slightly greater than those in the aggregate, the main difference being that the inflow of cars during restricted hours did not rise in August to such an extent as the aggregate inflow.

However, the above car data include car pools, and the picture is considerably clarified if the two categories, which respond differently to changes in the Area License fee, are examined separately. Table 4.11 shows the data for cars that did not carry enough people to qualify as car pools. From these data, it is clear that the increase in the Area License fee had an important effect on the number of people who bought a license to drive into the Restricted Zone during the restricted hours. The data on car pools in Table 4.12 shows that the number of car pools increased to offset partially the reduction in non-car-pool cars, thus increasing the car pool percentage of total car traffic.

Table 4.11 also shows that the number of non-car-pool cars entering the Restricted Zone also fell from four to seven percent just before and just after the restricted hours. This indicates that some other factors must have operated to reduce the flows of cars throughout the morning. Thus, the reduction in non-car-pool car traffic may not be entirely due to the increased Area License fee. Nevertheless, the latter effect is at least in the range from ten to thirteen percent. This is in line with the thirteen percent reduction in the number of area licenses sold over the same period. - 56 -

/a Table 4.10: CARS ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

/b Time (a.m.) Sept.-Oct. 1975 Feb.-May 1976 Aug. 1976

7:00-7:30 6640 6143 (-7.5%) 6618 (0%)

7:30-10:15 (ALS Period) 11363 10515 (-7.5%) 10676 (-6%)

10:15-10:45 6642 6304 (-5%) 6550 (-1%)

10:45-11:15 5661 5237 (-7.5%) 5565 (-2%)

RATE OF FLOW (vehicles per hour, average during period)

7:00-7:30 13279 12286 (-7.5%) 13236 (0%)

7:30-10:15 (ALS Period) 4132 3824 (-7.5%) 3882 (-6%)

10:15-10:45 13285 12608 (-5%) 13100 (-1%)

10:45-11:15 11322 10474 (-7.5%) 11130 (-2%)

/a Including company cars and car pools.

/b Excluding April 1976.

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from September- October 1975. License fees changed on December 31, 1975. See Table 4.8. - 57 -

Table 4.11: NON-CAR-POOL CARS ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

Time (a.m.) Sept.-Oct. 1975 Feb.-May 1976 /a Aug. 1976

7:00-7:30 5938 5536 (-6.8%) 5802 (-2.3%) 7:30-10:15 7159 5940 (-16.9%) 5466 (-23.6%) 10:15-10:45 6290 5993 (-4.7%) 6196 (-1.5%)

/a Excluding April 1976.

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from September- October 1975. License fees changed on December 31, 1975. See Table 4.8.

Table 4.12: CAR POOLS/a ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

Time (a.m.) Sept.-Oct. 1975 Feb.-May 1976 /b Aug. 1976

7:00-7:30 702 607 (-14%) 816 (+16%) 7:30-10:15 (ALS Period) 4217 4569 (+8%) 5210 (+24%) 10:15-10:45 352 311 (-12%) 354 (+1%)

PERCENT OF PASSENGER CARS

7:00-7:30 11 10 12 7:30-10:15 (ALS Period) 37 44 49 10:15-10:45 5 5 6

/a Cars with 4 or more occupants, exempt from area license requirement. /b Excluding April 1976.

Source: Public Works Department, Singapore.

Notes: License fees changed on December 31, 1975. See Table 4.8. The figures in parentheses are percentage changes from September-October, 1975. - 58 -

The biggest impact of the increased license fee was on the number of taxis entering the zone in restricted hours. This is seen in Table 4.13 as a drop of 24 percent in the first part of 1976, followed by a small additional decline by August 1976. It may be recalled that the flow of taxi trips during these hours had already fallen 65 percent in 1975; from March 1975 to August 1976 the overall decline during the 7:30-10:15 a.m. period was 75 percent.

Table 4.13: TAXIS ENTERING RESTRICTED ZONE

NUMBER OF VEHICLES (in each time period indicated)

Sept.-Oct. 1975 Feb.-May 1976 /a Aug. 1976

Time (a.m)

7:00-7:30 854 836 (- 2%) 913 (+ 7%) 7:30-10:15 (ALS Period) 3787 2867 (-24%) 2724 (-28%) 10:15-10:45 2643 2381 (-10%) 2448 (- 7%)

RATE OF FLOW (vehicles per hour, average during period)

7:00-7:30 1708 1672 (- 2%) 1826 (+ 7%) 7:30-10:15 (ALS Period) 1377 1043 (-24%) 991 (-28%) 10:15-10:45 5286 4762 (-10%) 4896 (- 7%)

/a Excluding April, 1976.

Source: Public Works Department, Singapore.

Notes: The figures in parentheses are percentage changes from September-October, 1975. License fees changed on December 31, 1975. See Table 4.8.

OUTBOUND EVENING TRAFFIC

The Area License Scheme was designed to be in operation only during the morning commuting period and to apply only to inbound traffic. It was believed that if commuters could be induced to give up driving to work, they would not have cars at work to drive home in, and that the evening traffic peak would thereby be reduced appreciably as well as the morning peak. It was surprising, therefore, that the evening traffic peak remained close to its previous level after the Area License Scheme was implemented. - 59 -

Data from automatic counters, given in Table 4.14, show how little change there was in traffic leaving the Restricted Zone during the evening.

Table 4.14: EVENING OUTBOUND TRAFFIC ON EIGHT ROADS

(Vehicles of all types, counted mechanically)

Time (p.m) March 1975 June 1975 July 1975 Sept.-Oct. 1975 (before ALS) (ALS in effect (ALS in effect (ALS in effect 7:30-9:30 a.m) 7:30-9:30 a.m) 7:30-10:15 a.m)

4:00-5:00 15900 15650 (-2%) 15070 (-5%) 15820 (-0.5%) 5:00-6:00 19275 18250 (-5%) 17840 (-7%) 19050 (-1%) 6:00-7:00 16395 17510 (+7%) 15760 (-4%) 15575 (-5%)

4:00-7:00 51570 51410 (-0.3%) 48670 (-6%) 50445 (-2%) (Total)

Source: Public Works Department, Singapore.

Note: The figures in parentheses are percentage changes from March 1975.

Unlike the inbound flow data, which cover all entry points to the Restricted Zone, these data cover only eight selected outbound arterial roads. Never- theless, these roads are the most important ones, normally carrying about two-thirds of all outbound traffic, and changes in these flows should be representative of changes in the total flow.

The figures indicate no appreciable change in the total three-hour flow in June as compared with March, a decline of six percent in July, and a recovery to within two percent of the before-ALS level in the September- October average. These are small changes compared to the 38 percent reduction in inbound flow between 7:00 a.m. and 10:15 a.m. (see Table 4.1). Moreover, the pattern of variation does not suggest any relation to the introduction of the Area License Scheme at the beginning of June. It is puzzling that the level of evening traffic, which must have included a large flow of home-bound commuters, was apparently not affected by the measures that induced many commuters not to drive into the Restricted Zone during the morning restricted hours.

As soon as it was recognized that the reduction in morning traffic was not reflected in reduced evening volume, the Road Transport Working Group began collecting more complete information on evening traffic, including the flows on all roads leading out of the Restricted Zone. These were counted manually, and cars were distinguished from other vehicles. Unfortunately, no comparable data were collected before the Area License Scheme. Therefore it is not possible to ascertain whether the scheme brought about a change in - 60 - composition, such as a reduction in car traffic offset by an increase in some other category. Results are shown in Table 4.15 for the month of October only; corresponding figures for June, July, August and September differ by only a few percent.

Table 4.15: EVENING OUTBOUND TRAFFIC ON ALL ROADS

October 1975

Time (p.m) All Vehicles Cars Cars as %

4:00-5:00 22834 9975 44% 5:00-6:00 27767 14211 51% 6:00-7:00 23390 11866 51% 4:00-7:00 73991 36052 49% (Total)

Source: Public Works Department, Singapore.

In order to measure the volume of crosstown traffic going through the Restricted Zone in the afternoon peak, the Public Works Department, with assistance from the World Bank's traffic survey team, observed and matched license numbers of vehicles entering the Restricted Zone at the eight most heavily traveled entry points on the south and west sides of the Restricted Zone and leaving it at various exit points on the northeast side. The survey was carried out from 4:30 p.m. and 6:30 p.m. on one day in September 1975. While this procedure missed any through traffic in the opposite direction and any that entered by secondary streets, it captured the most important flows. In order to keep the data processing operation small enough to be done manually, only license numbers ending with pre-specified digits were recorded. The numbers of vehicles observed both entering and leaving the Restricted Zone were then factored to estimate the total crosstown flows.

This survey indicated that 52 percent of cars entering the Re- stricted Zone at the eight selected points between 4:30 p.m. and 6:30 p.m. went across and out of the zone again. The number of cars crossing the zone in this direction was estimated as 6,357 in the two-hour period. If these flows followed the same pattern as the total outbound traffic, then in the three-hour period from 4:00 p.m. and 7:00 p.m. the number of crosstown car trips would be in the neighborhood of 9,000.

Even if none of these trips had counterparts passing through the Restricted Zone in the morning -- an exaggerated assumption -- the number of crosstown afternoon trips is only a fraction of the number of trips that were suppressed in the morning. Therefore additional explanations must be sought to account for the level of evening traffic flow having remained so nearly the same as it was before the Area License Scheme. This topic will be taken up again in Chapter 6. - 61 -

VEHICLE SPEEDS

MEASUREMENT METHODS AND PROBLEMS

Because some of the benefits of restraining traffic were expected to come from faster operating speeds of cars, buses, and goods vehicles, the research plan included measurement of vehicle speeds before and after the institution of the Area License Scheme. The conventional moving-car method of collecting such data was considered but was set aside on grounds that reliable before-and-after comparisons required a larger sample of observations on any stretch of road within any time period than could practically be obtained by that method. Instead, a license-matching technique was decided upon.

In brief the technique consisted of recording license numbers and accurate time readings at several points along a road, ascertaining the time elapsed for those vehicles observed at successive points, and thence calculating the average speed of each vehicle between the points. In a ten minute observation period, a group of observers using hand-held tape recorders could record the license numbers of up to several hundred vehicles, classified by vehicle type. Not all of the license numbers recorded at one point would be observed again at the next, because some vehicles would stop or turn off on side streets, and so on, but enough would usually be matched to allow calculating the mean speeds (and standard deviations) of a set of 20 to 200 cars and of smaller numbers of other vehicles.

While the method is fundamentally valid and feasible -- as was demonstrated later -- defective procedures and data processing problems resulted in unusable data for both the "before" and "after" stages of data collection. Therefore, no direct before-and-after comparisons of speeds can be based on these data.

POST-ALS SURVEY

After the license-match data collection problems had been identified, further work was done to develop reliable procedures through a series of pilot tests. Then a new set of speed measurements was made (after the introduction of the Area License Scheme) on selected roads during the restricted hours and during the afternoon peak, to get comparative data on uncongested and congested conditions.

Throughout this new program, license-matching and moving-car methods were used simultaneously, so that the two methods could be compared. The moving car was driven at what the driver judged to be the average speed of the surrounding traffic while an observer in the car recorded the time of passing each of 20 or 30 checkpoints along the route. In one hour, the moving car would make several circuits of a relatively long loop, while - 62 - the license-match team would record several hundred license numbers on one segment of that loop. 1/

Survey Coverage

Figure 4.1 shows the loops around which the observer's car was driven for the moving-car speed measurements. They were chosen to include principal radial arterial roads, part of the ring road, and some of the main streets within the Restricted Zone. Also shown are the segments -- generally one on each side of each loop -- on which license numbers were recorded.

All of the principal roads within the Restricted Zone are either one-way or divided. Most of the moving-car loops were planned to include both directions on a main one-way pair or a divided main road. On Loop Seven, the ring road varies from an undivided two-way, two-lane road to two completely separate four-lane one-way roads. The license observations for the clockwise direction on the ring road were made on the four-lane half of a divided road, while the counterclockwise segment started on a two-way, two- lane stretch, and then merged with another branch into a divided road with three lanes each way.

On each license-matching segment, data were collected for ten out of each fifteen minutes between 8:00 a.m. and 9:00 a.m. and between 5:00 p.m. and 6:00 p.m. Data were also recorded between 10:00 a.m. and 11:00 a.m. on eight of the eighteen segments. Whenever these data were being collected, the moving car traveled around the loop that included the segment under survey, starting about a half-hour before and continuing for about a half-hour after the license observation period.

Results -- Car, Van, and Lorry Speeds

For each ten-minute period, the license-matching results yielded complete counts of the vehicles passing each end of the segment, classified as cars, vans, lorries, buses, and motorcycles, as well as the mean speed of each type except the last.

Car speeds from license matching and those from simultaneous moving- car runs over the same segments were scattered over about the same ranges, with no appreciable bias between the two sets of results. In view of these similarities, the comparisons presented below are based on the moving-car data, which cover longer stretches of road, amounting to a more representative sample of the different road types in the area than the shorter segments on which licenses were matched.

1/ The procedures for both methods, as well as the moving-car loops and license-matching locations, are specified in detail in a report by Survey Research Singapore and Richard Worral: Speed/Flow Measure- ments, Procedures, June 1976. IBRD 13182 This map has beenprepared by the f> Wand Bant s 0staff \ zit exclusively jlSt \ vVor/o' far Srlk I ~~~~~~~~~~~:5-,,~~~~~~~~~~~~~~~~~~~~~~ ~~~the can veience of the readeraof - - the roped ta whiChNTSItGAPOREisl Qlcl jr~~~~~~~~~~Sah D Thedenaminatians used and the ad.A...... X ..| 4ao=j * Roatd boundaries shown n this map Z 0 ~~~~~~~~"''%e5db07 do not imply, an the part sf the CENTRAL u sot. ..Oe World Bank and its affiliates, any SECTION a n- c judgment os the lagal status of -I- REPEATED V5ary acetanceorof suc bnsosarea.t any2,

atanry%.; X g*;; { 7%- od KILOMETERS

- /d'¼ ~~~~CENTRAL SINGAPORE

:- 7/74 aht-,R YcpelRad ! ."Z( MSpeedS Measurement Loops

>" n ) ->/ NRoads Rivers

- One way streets

R~~~~ -*'- ~~~~~~~~Restricted Zone boundary

Kcppel ____ -'-.-.- ~~~Loops surveyed by moving car V-A' ~.p-LM----eSegment \...e.K surveyed by license matching

d,j ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Figure4.1

JANUARY 1978

- 65 -

Analysis of the license-matching data showed that car and van speeds were approximately equal, with lorries about five percent slower. In general, these three went up or down together. Thus, changes in the moving-car speeds may be taken as a good approximation of changes in speeds of vans and lorries (or "goods vehicles"). Bus speeds did not follow the same pattern; they will be taken up in the next section of this chapter.

For analysis of the results, different parts of the loops surveyed were classified as inside the Restricted Zone, radial, and ring road, with directions distinguished for the radials. Table 4.16 specifies which parts of the various loops were included in each category.

Table 4.16: CLASSIFICATION OF MOVING-CAR LOOPS

Category Loops

Inside the Restricted Zone 1, 2, 3 and part of 6 Inbound Radial, outside the Zone Inbound portions of 4, 5, and 8 Outbound Radial, outside the Zone Outbound portions of 4, 5, and 8 Ring Road 7

Mean speeds were calculated for each of these road categories from the moving-car data taken in the morning restricted hours, the unrestrained afternoon peak, and, in some cases, the late morning period. The results are shown in Table 4.17.

Table 4.17: SUMMARY OF MEAN SPEEDS FROM MOVING CAR DATA (kilometers per hour)

Time of Day

Roads 7:30-10:15 10:15-12:30 4:30-6:30 (restricted hours)

Inside Restricted Zone 33 25 27 Inbound Radials 32 n.a./a 31 Outbound Radials 35 n.a./a 29 Ring Road 20 24 25

/a n.a. = not available

It is notable that the mean speed inside the Restricted Zone during restricted hours is considerably higher than at other times of day, and that the speed on inbound radials outside the restricted zone is as high during the restricted period as it is in the afternoon when the main flow is outward. The slowest traffic is that on the ring road in the morning, when that road is being used by many cars detouring around the Restricted Zone. - 66 -

From the figures in Table 4.17 and a few plausible assumptions, the impacts of the Area License Scheme on speeds in the morning restricted period on the different sets of roads may be estimated. Since traffic flows in the afternoon peak changed very little when the scheme took effect, it may be concluded that speeds in that period were unaffected. For the midday period there are no data on flow changes, and it can only be said that large impacts of the license scheme seem unlikely. For the restricted period itself, the following assumptions are put forward:

1. Without restrictions, traffic flows and therefore speeds within the Restricted Zone in the morning would be comparable to those in the same area in the afternoon peak;

2. Without restrictions, traffic flows and speeds on the inbound radials in the morning peak would be comparable to those on the outbound radials in the afternoon peak;

3. Traffic flows and speeds on outbound radials in the morning are not significantly affected by the Area License Scheme; and

4. In the absence of the diversions caused by the Area License Scheme, morning traffic on the ring road would be equivalent to the observed afternoon traffic, traveling at the speed observed in the afternoon.

On the basis of these assumptions, a set of estimated "without- ALS" morning peak speeds was arrived at. The actual "with-ALS" speeds are compared with them in Table 4.18.

Table 4.18: TRAFFIC SPEEDS WITH (ACTUAL) AND WITHOUT (ESTIMATED) AREA LICENSE SCHEME IN MORNING COMMUTING HOURS (kilometers per hour)

Without ALS With ALS Percent Change (estimated) (observed)

Inside Restricted Zone 27 /a 33 +22 Inbound Radials 29 lb 32 +10 Outbound Radials 35 /c 35 0 Ring Road 25 Id 20 -20

/a From assumption 1. in text. /b From assumption 2. in text. /c From assumption 3. in text. /d From assumption 4. in text. - 67 -

If the assumptions stated above are valid, the Area License Scheme has increased speeds during the restricted hours within the Restricted Zone, as would be expected, and to a lesser degree, on the inbound radial roads. At the same time, traffic diverted to the ring road has reduced speeds there.

The speeds inside the zone without restrictions -- 25 to 27 kilometers per hour -- are not low enough to indicate severe congestion, which in other cities is characterized by speeds of 15 to 20 kilometers per hour, or less. Nevertheless, the increased speed within the Restricted Zone indicates improved freedom of traffic movement there during the restricted hours. However, Table 4.17 shows that the speed later in the morning is even lower than in the afternoon peak, suggesting that the improvement is confined to the restricted hours.

For inbound travelers in the period of restrictions, the 10 percent increase in speed on the radials and the freer movement inside the zone indicate smoother-flowing, less frustrating traffic conditions, which drivers are likely to regard as benefits. Conversely, drivers using the ring road in the morning are encountering increased congestion and the annoyance that goes with it.

Results -- Bus Speeds

The average speed of buses on any road generally differs from that of other vehicles. Stopping for passengers not only increases travel time by the time spent at the bus stop, but also introduces acceleration and decel- eration phases during which the average speed is only about half the running speed between those phases. On the other hand, where buses have reserved lanes and other traffic is badly congested, it is possible for the average bus speed to exceed that of other vehicles in spite of the stops. Therefore, the moving-car speeds cannot be assumed to represent the speeds of buses on the same roads.

In the license-matching survey, buses were observed separately from other vehicles. Their average speeds are shown in Table 4.19 for road segments classified in the same way as the moving-car loops. As was observed for other traffic, speeds inside the Restricted Zone and on outbound radials were appreciably higher during the morning restricted period than in the afternoon. A lower mean speed was recorded in the morning than in the afternoon on the inbound radials outside the Restricted Zone, whereas the car speeds on these roads were nearly the same in both the morning and the afternoon. The biggest contrast between bus and car speeds occurred on the ring road, where the bus speed in the morning exceeded that in the afternoon by four kilometers per hour while the car speed was five kilometers per hour lower in the morning than in the afternoon. - 68 -

Table 4.19: SUMMARY OF MEAN BUS SPEEDS FROM LICENSE PLATE OBSERVATIONS (kilometers per hour)

Time of Day

Roads 8:00-9:00 10:00-11:00 5:00-6:00 (within restricted hours)

Inside Restricted Zone 25 21 20 Inbound Radials 19 n.a. 22 Outbound Radials 27 n.a. 21 Ring Road 23 n.a. 19

For the ring road, examination of the separate figures that constitute the average shows a wide range of variation. By direct observation, it has been noted that some of the lower-capacity sections of this circuit and especially some of its intersections are often on the verge of unstable congested flow when traffic is heavy, so that one unresponsive driver or a less- than-expert traffic policeman may introduce sizable delays into the traffic flow at some times. Possibly this phenomenon accounts for some of the apparent inconsistences between results for cars and for buses.

PARKING

Another instrument of traffic restraint, second only to the Area License Scheme as a measure specifically aimed at reducing car trips into the city center, was the policy on parking fees. On May 1, 1975, one month before implementation of the Area License Scheme, a new set of parking charges was put into effect in government-owned car parks and for on-street parking. In addition, private car park operators were required to increase their fees to at least the level charged in government-owned car parks in the same zone. The rates, before and after May 1, 1975, are shown in Table 4.20. - 69 -

Table 4.20: PARKING CHARGES (Singapore Dollars)

Before May 1, 1975

First Subsequent Monthly Hour Half-Hours Rate

Core of Restricted Zone .50 .50 50 generally

Subsequent Full Hours

Rest of Restricted Zone .40 .40 30 to 50

After May 1, 1975

First Second Subsequent Hour Hour Half-Hours

Core of Restricted Zone .50 1.00 1.00 70 minimum (some charge 80)

Subsequent Half-Hours

Rest of Restricted Zone .50 .50 60 minimum

The new fees were higher in the "core" or most congested part of the Restricted Zone than in the rest of the zone, and the hourly rates were made progressive with time, especially in the core of the zone. Thus, in the core, the second hour costs twice as much as the first, and each succeed- ing half-hour costs as much as the second full hour. This rate schedule is intended to discriminate against commuters and other long-term parkers and have relatively less impact on short-term parkers, who are assumed to be shoppers or clients of businesses, banks, lawyers, doctors, and so on. Dis- crimination against commuters, however, was not carried to its logical extreme. "Season" (i.e. monthly) rates, although increased, still were only equivalent to the cost of about three hours a day at hourly rates.

The increased parking rates were part of a package of policy changes which, although not simultaneous, were announced in advance and implemented over the course of a month. One may look for the separate - 70 - impact of the parking charges by comparing data for May with earlier data on trips downtown and on parking. However, it is not clear that the comparison shows the full adjustment that would have taken place over a longer period with no other changes in policy. After June 1, of course, the impact of the Area License Scheme was superimposed on that of the parking charges, and the two must be considered jointly.

ANALYSIS

While the data on flows into the Restricted Zone, already examined, show what happened to vehicle trips downtown, the parking data shed light on the degree to which the reductions in inflow represented reduced numbers either of people driving in to spend the day or of those making short visits. They also give some indication of the extent to which the limited hours of the license requirement have deferred some trip-making to a later time.

Table 4.21 shows what happened to long-term parking in terms of sales of monthly tickets at a sample of five parking garages in the core of the Restricted Zone, three in the "non-core" area of the Restricted Zone, and one just outside the zone. From the changes in May relative to the January-to-April average, it can certainly not be said that the increased rates caused a drop in monthly parking. One car park in the core suffered an eight percent decline, one gained thirty-four percent, and the total for the five together scarcely changed. In the outer part of the zone, the combined sample of the three car-parks gained thirteen percent. Outside the zone, the decrease of two cars is probably not significant.

The longer run combined impact of parking rates and Area License fees should be seen in the comparison of the June-to-September average sales with those for January-to-April. Surprisingly, only one car-park showed a loss of more than five percent in monthly ticket sales, and several showed substantial gains. With a drop of about 60 percent in cars entering the Restricted Zone between 7:00 a.m. and 10:15 a.m., one might expect a large drop in all-day parking, assuming most of the cars in question were those of commuters. Such an effect did not occur in the car parks included in this survey.

Why monthly parking ticket sales in these car parks did not reflect the decline in cars entering the zone is a matter for speculation. Data on downtown parking are by no means comprehensive enough to answer the question. Obviously there must have been a decline in the number of cars parked somewhere. One plausible hypothesis is that the total demand for all-day parking space on a regular basis (before the Area License Scheme took effect) greatly exceeded the supply of space available on a monthly basis, so that many commuters were paying hourly rates for all-day parking. If this were the case, any reduction in total parking would probably show up in the numbers parking on an hourly basis, with hourly parkers shifting to a monthly basis as fast as others vacated the monthly space. This assumes that the allocation of space to monthly parking is administratively limited, although the same physical space should be usable on either basis. - 71 -

Table 4.21: MONTHLY PARKING TICKET SALES

(Average Number of Monthly Tickets)

CORE OF RESTRICTED ZONE

Overseas Multi- U.I.C. Union, Storey, Building, Colombo High St. Collyer Quay Market St. Shenton Way Court Centre Total

1975 Jan.-Apr. 647 726 300 260 150 2083

May 593 720 310 251 201 2075

Change /a -8 -1 +3 -4 +34 -0.4

June-Sept. 483 691 332 252 187 1945

Change /a -25 -5 +11 -4 +25 -7

RESTRICTED ZONE, OUTSIDE CORE

Peace O.G. Bldg., Centre, Upper Selegie Selegie Rd. Cross St. Complex Total

Jan.-Apr. 132 25 76 233

May 162 27 74 263

Change /a +23 +8 -3 +13

June-Sept. 152 25 76 253

Change /a +15 0 0 +9

OUTSIDE RESTRICTED ZONE

International Building, Orchard Road

Jan.-Apr. 82

May 80

Change /a -2

June-Sept. 86

Change /a +5

/a Percent change from January-April average. - 72 -

If the hypothesis stated above were true, there should be a drastic decline in the number of vehicles parked on an hourly basis. Unfortunately there are only four car parks for which there are comparable before-and-after data on hourly parking. For these four, figures on vehicles entering for hourly parking are given in Table 4.22 for March and September 1975. In addition to the all-day totals, figures are given for the numbers of cars parking before 8:00 a.m., having presumably entered the Restricted Zone before 7:30 a.m., those that parked between 8:00 a.m. and 10:30 a.m., roughly approximat- ing those that would have entered the Restricted Zone during the restricted hours, and those parking between 10:30 a.m. and 11:30 a.m., just after the restricted hours.

Table 4.22: HOURLY PARKING

(Number of Entries per Day into Car Park)

LOCATION

Overseas Colombo High St. Selegie 1975 Union Court Centre Complex Total

Entries Before 8:00 a.m.

Feb. or March 2 0 0 1 3 September 11 15 9 14 49

Entries Between 8:00 and 10:30

Feb. or March 162 217 136 129 644 September 81 182 73 51 387 Change /a -50 -16 -46 -60 -40

Entries Between 10:30 and 11:30

Feb. or March 102 196 75 94 467 September 95 220 62 81 458 Change /a -7 +12 -17 -14 -2

Entries During Whole Day

Feb. or March 705 1609 513 584 3411 September 439 1465 424 506 2834 Change /a -38 -9 -17 -13 -17

/a Percent change from February or March to September.

As expected, Table 4.22 indicates a large decrease in the number of cars entering each of the four car parks between 8:00 and 10:30 a.m. for hourly parking. Even these reductions, however, amount to only about half the percentage reductions in the observed flows of cars into the Restricted Zone during the restricted hours. Thus hourly parking in these particular car parks is not where the impact of the reduced flow of cars is falling. - 73 -

The hypothesis that many long-term parkers were paying hourly rates is not supported by data from these four car parks. Table 4.23 shows the proportions of hourly-rate parkers parking for different periods of time, and it can be seen that the numbers parking for more than four hours made up only a small percentage of the total both before and after the traffic restraints were instituted.

Vacant parking spaces must have been left somewhere by those who stopped driving into the Restricted Zone when the Area License Scheme took effect, but relatively few of the vacancies, proportionately, were left in the parking garages covered by this survey.

Table 4.23: DURATION OF PARKING AT HOURLY RATES

(Percent Remaining for Time Indicated)

LOCATION

Overseas Colombo High St. Selegie 1975 Union Court Centre Complex

Up to One Hour

Feb. or March 62 67 50 78 September 62 71 62 73

One to Four Hours

Feb. or March 36 30 38 20 September 30 26 30 24

More than Four Hours

Feb. or March 2 3 11 2 September 8 3 9 3

ACCIDENTS

It was expected that, after institution of the Area License Scheme, there would be fewer cars in the central city but that they would be traveling at higher speeds. Whether these changes would lead to a reduction in accidents was a matter for speculation. One opinion was that the number of accidents would go down but that those that occurred would be more severe, especially where pedestrians were involved. - 74 -

To investigate whether the Area License Scheme had a significant effect on accidents, particularly regarding rate of occurrence, degree of seriousness, and involvement of pedestrians, data were compiled on the incidence of accidents throughout the city. Data were compiled only for accidents involving injuries, because reporting of less severe accidents was believed to be unreliable. A distinction was made between accidents involving only motor vehicles and those which involved motor vehicles and pedestrians. This information has been further categorized to see if there has been a shift in concentration of accidents during different time periods of the day.

NUMBER OF ACCIDENTS

Table 4.24 illustrates the pattern of occurrence of minor, serious, and fatal injury accidents (a) which involved only motor vehicles, and (b) which involved pedestrians, during the months January through September 1975.

Table 4.24: ACCIDENTS BY TYPE AND SEVERITY

Two or more Vehicle and Vehicles Pedestrian

1975 Minor Serious Fatal Minor Serious Fatal Total

January 120 34 3 106 59 4 326 February 75 38 10 61 53 9 246 March Before 95 37 6 77 46 6 267 ALS April 79 24 3 103 38 8 255 May 82 27 6 108 36 4 263

June 111 27 2 74 41 3 258 July With 103 29 2 95 56 8 293 ALS August 97 43 7 105 57 9 318 September 97 30 1 90 40 4 262

Mean, before ALS 90 32 6 91 46 6 271 Mean, with ALS 102 32 3 91 49 6 283

Comparison of the mean numbers of accidents per month before and after introduction of the Area License Scheme shows no difference in some categories and differences well within the range of scatter in others. The small increase (from 46 to 49) in serious accidents involving pedestrians cannot be attributed with any certainty to the Area License Scheme, because, - 75 -

as the month-to-month variations indicate, extraneous factors can cause much larger differences. The same may be said for the increase in minor vehicle- only accidents and the decrease in fatalities in that category. Thus, there is no firm evidence of any impact of the scheme on the frequency of accidents or on their types.

TIMING AND SEVERITY OF ACCIDENTS

Accidents were then analyzed according to the time at which they occurred, specifically with respect to the restricted hours, to see if there was a change in frequency during those hours after the introduction of the scheme. Table 4.25 shows the percentages of accidents occurring during four periods of the day -- early morning pre-restricted hours, 7:30-9:30 a.m. (restricted hours for June and July), 9:30-10:15 a.m. (extension of restricted hours for August and September), and the rest of the day.

For the 7:30-9:30 a.m. period, the mean number for January through May is 32, and for June through September it is 38. A "t" test showed that this difference is not significant. For other time periods the differences are even smaller.

Data on the relative proportions of minor, serious, and fatal accidents are presented in Table 4.26. There is no appreciable shift in the distribution of accidents among the three severity categories. Hence, it may be concluded that the Area License Scheme had no measurable effect on either the timing or the severity of accidents. - 76 -

Table 4.25: TIME OF OCCURRENCE OF ACCIDENTS

Number of accidents in each period (Percent of month's total shown in parentheses)

Midnight 7:30 to 9:30 to 10:15 a.m. to 1975 to 7:30 a.m. 9:30 a.m. 10:15 a.m. midnight

January 40 (12) 33 (10) 3 (1) 250 (77) February 23 ( 9) 25 (10) 10 (4) 188 (76) March 35 (13) 30 (11) 9 (3) 192 (72) April 34 (13) 39 (15) 12 (5) 170 (67) May 23 ( 9) 33 (13) 7 (3) 200 (76)

June 24 ( 9) 38 (15) 14 (5) 182 (71) July 37 (13) 43 (15) 9 (3) 204 (70) August 38 (12) 34 (11) 10 (3) 236 (74) September 36 (14) 37 (14) 5 (2) 184 (70)

Mean Jan.-May 32 (12) 32 (12) 8 (3) 200 (73) June-Sept. 34 (12) 38 (14) 8 (3) 201 (72)

Table 4.26: SEVERITY OF ACCIDENTS

(Percent of each month's accidents in each category)

1975 Minor Injury Serious Injury Fatal

January 69 29 2 February 55 37 8 March 64 31 4 April 71 24 4 May 72 24 4

June 72 26 2 July 68 29 3 August 64 31 5 September 71 27 2

Mean Jan.-May 67 29 4 June-Sept. 68 29 3 - 77 -

CHAPTER 5. PRINCIPAL CHANGES IN TRAVEL BEHAVIOR

The foregoing chapter reported on phenomena that are physically observable and can be measured, like the flows of different types of vehicles. Two of the most important variables -- the flow and occupancy of automobiles -- depend upon decisions made by many individual trip makers about where to go, when, and by what mode of transport. In order to gain a better understanding of these decisions, household surveys were carried out before and after implementation of the Area License Scheme, gathering detailed data on travel behavior and facts about the households and individuals interviewed. The main results of analyses of these surveys are presented in this chapter. Because of the complexity of the information, some of the secondary findings are presented in an annex to this chapter, so that they will not obscure the main findings but will still be available to those who want them.

In the body of the chapter, after a brief explanation of the samples, the analysis begins with an examination of changes in the rates of trip making for different purposes and origin-destination patterns. Then, trips to work in the Restricted Zone from vehicle-owning households are analyzed. Changes in mode, trip starting time, and travel time are presented and discussed. The next main section contains a similar analysis for work trips for which the origins and destinations lie on opposite sides of the Restricted Zone (trans-Restricted-Zone trips), so that by-passing the zone is an additional option. The examination of work trips is completed by a look at those from non-vehicle-owning households. Then trips home from work are examined, first those from inside the Restricted Zone, then trans-Restricted-Zone trips (both by members of vehicle-owning households) and finally trips home from work by people from non-vehicle-owning households. Two final main divisions of the chapter cover trips to the Restricted Zone for non-work purposes, and trips entirely outside the zone for various purposes.

THE SAMPLES

The results in this chapter are based on interviews carried out before and after the introduction of the Area License Scheme. There are two principal samples -- members of non-vehicle-owning and of vehicle-owning households. Subsamples were extracted from the latter for more detailed analysis. In the definition of these samples, "vehicle" means any motor vehicle, including a motorcycle or scooter, and "owning" means having the use of a vehicle based at the household, even though it may be registered as the property of a firm.

The preliminary contacts that were made to establish the non- vehicle-owning sample showed that 41 percent of households in Singapore were in the vehicle-owning category, and 59 percent were non-vehicle-owning. - 78 -

However, in order to permit more detailed analysis where the most important changes were expected, that is, by people using vehicles, interviews were conducted at about three times as many households with vehicles as without. Moreover, for most purposes the vehicle-owning sample was enriched by the addition of a subsample of households whose cars had been observed entering the to-be-restricted area before the scheme was inaugurated. In addition, the ratio of multi-vehicle to single-vehicle households in the sample is twice that observed in the population as a whole. The numbers in the samples used for the aggregate (or "macro") before-and-after analysis are:

Household Class Number of Households Number of People Interviewed (before and after) (after)

Vehicle-owning 1,554 9,728 Non-vehicle-owning 499 3,049

When the analysis would have been biased by using the vehicle-owning household sample just described, for example, in analyzing origin-destination patterns of trips made outside the Restricted Zone, the extra sample of households based on cars entering the area was deleted. The remaining sample is referred to as the unweighted vehicle-owning sample. It includes 1,127 households. The non-vehicle-owning sample required no such alteration, since it was ran- domly selected.

While the aggregate samples thus far described included the same households before the scheme and afterward, some households had gained or lost members. Moreover, the origins, destinations, and purposes of the trips reported by any given household member were not necessarily the same at both times. To get a more precise picture of the changes in the way specific trips were made by those affected by the Area License Scheme, a subsample was extracted from the vehicle-owning sample, consisting of 720 people who made trips from home outside the Restricted Zone directly to work inside the zone both before and after the scheme was instituted. Analysis of the trips by this group is referred to as the "micro" analysis.

HOUSEHOLD TRIP RATES

Two of the possible results of a policy that restrains cars from going downtown (even if only at certain hours) are (a) a change in the absolute rate of making trips, and (b) a change in the relative rates of trip- making into that area and into other areas. The extent to which these responses occurred was examined by comparing the trip rates (number of trips per day per household) for particular purposes and origin-destination patterns. The trip rates include trips on foot and by bicycle. For this analysis, the after-ALS unweighted sample of vehicle-owning households was divided into two groups to distinguish a small group of 76 households that sold or scrapped their vehicles and did not replace them between the two surveys, thus becoming "ex-vehicle-owning households", from the 1,051 households that kept or replaced their vehicles and thus continued as vehicle-owning households. - 79 -

In comparing trip rates before and after the institution of the Area License Scheme, it is important to remember that other policies were being implemented at about the same time which were intended to have a more general impact on ownership and use of cars without regard to any specific geographical area. These policies, as well as a slackening in the rate of economic growth, would be expected to reduce trip rates to some degree. In addition, the mean number of respondents per household fell by one percent. Thus, the effects of the Area License Scheme and the associated parking rate increases must be sought mainly in terms of differential effects on trips into the Restricted Zone and to other destinations.

CONTINUING VEHICLE-OWNING HOUSEHOLDS

The overall trip rate per household per day fell by 12 percent in the period during which the Area License Scheme was implemented. This reduction, however, was the result of a series of changes of different sizes and directions in different types of trips. The following analyses break the trip rate down by purpose and origin/destination pattern.

Analysis by Purpose

The trip rates for different purposes are shown in Table 5.1:

Table 5.1: TRIP RATES BY PURPOSE (All modes) (Continuing Vehicle-Owning Households)

Personal Social/ To Work To Shop Business Recreational To School Total

Before 2.72 0.44 0.37 0.37 1.82 5.73 After 2.48 0.36 0.24 0.16 1.77 5.01 Percent Change -9% -19% -35% -57% -2% -12%

Although work trip rates fell by nine percent, it is difficult to understand how this could be the result of the Area License Scheme. Singapore was in a recession at the time the second survey was undertaken, so that work trips may have been affected by a decline in employment. However, appropriate data on employment levels are not available.

Unemployment figures from the Singapore Monthly Digest of Statis- tics for the periods of the two surveys show a small increase in the number of males registered with the Employment Service as unemployed and a decrease in the number of females registered. The total number registered fell from 41,005 to 39,967. This change is only about one-eighth of a percent of the size of the work force (employed plus unemployed), and the work force itself may have changed by more than that, either upward or downward. Lacking data on changes in the work force, it is impossible to determine how much employment changed and therefore how much of the reduction in work trip rates should be attributed to the recession. - 80 -

The numbers of shopping, personal business, and social/recreational trips all fell by large amounts. The sample sizes for all of these trip purposes are rather small. Thus, the results are somewhat unreliable. Moreover, the relatively large reductions in what might be termed "discretionary" trips -- personal business, social/recreational -- may well be the result of other government policies aimed at restraining the use of cars.

Analysis by Origin/Destination Pattern

The key question in this analysis is whether trip rates for trips into the Restricted Zone fell more than trip rates for trips with other origin/destination patterns. To examine this, the rates for trips to the zone are compared with those for trips made completely outside the zone. The data are presented in Table 5.2:

Table 5.2: TRIP RATES BY ORIGIN/DESTINATION PATTERN (All modes) (Continuing Vehicle-owning Households)

Personal Social/ To Work To Shop Business Recreational To School Total

TO THE RESTRICTED ZONE

Before 0.75 0.04 0.07 0.07 0.20 1.14 After 0.67 0.03 0.04 0.03 0.18 0.95

Percent Change -11% -25% -43% -57% -10% -17%

OUTSIDE THE RESTRICTED ZONE

Before 1.56 0.35 0.23 0.22 1.49 3.86 After 1.51 0.30 0.15 0.11 1.47 3.54

Percent Change -3% -14% -35% -50% -1% -8%

In all cases, the fall in trip rates to the Restricted Zone was greater than the fall in trip rates outside the zone. If the decline outside the Restricted Zone may be assumed to represent the effect of taxes on car ownership and of the recession, then the difference in the declines in trip rates indicates the upper limit of the effect that might be due to the Area License Scheme:

DIFFERENCES IN TRIP RATE DECLINES TO RZ AND OUTSIDE RZ

Personal Social/ To Work To Shop Business Recreational To School Total

-8% -11% -8% -7% -9% -9% - 81 -

To the extent that the decline in trips to the Restricted Zone was the result of decentralization that would have occurred without the scheme, the above data overstate the extent to which trips to the Restricted Zone declined as a result of the Area License Scheme.

Summary: Continuing Vehicle-Owning Households

Overall, the rate of trip-making declined over the period covered by the household surveys, especially for purposes other than work and school. Questions regarding the relative effects of recession, decentralization, policies to discourage car ownership and use, and possible under-reporting cloud the issue of how much of the decline can be attributed to the Area License Scheme. However, for every purpose, the decline in the rate of making trips to the Restricted Zone was somewhat greater (by 7 to 11 percentage points) than the decline in trip-making outside the zone. It seems reasonable to conclude that part of this differential was due to the Area License Scheme and part to the decentralizing trend that was already under way.

NON-VEHICLE-OWNING HOUSEHOLDS

The conclusion that factors other than the Area License Scheme were operating to reduce trip rates is supported by the data on trip rates for non-vehicle-owning households. Overall, trip rates fell from 5.21 to 4.89, a six percent reduction. This reduction reflects a three percent increase in work trips offset by reductions in trips for other purposes, especially a large drop in social-recreational trips.

The increase in work trips is somewhat surprising, given the reports of a recession, but, as already noted, there are no data to show whether the number of jobs actually rose or fell.

Geographically, the work-trip increase took place entirely outside the Restricted Zone; the rate of making work trips into the zone did not change. For other purposes, the numbers of trips into the zone were too small for meaningful analysis of the changes by destination and purpose. When all purposes are combined, however, it is found that, unlike the work- trip rate, the overall trip rate declined by the same proportion -- six percent -- for both origin/destination patterns.

From this evidence, there is no reason to believe that the Area License Scheme affected the trip-making behavior of people from non-vehicle- owning households.

EX-VEHICLE-OWNING HOUSEHOLDS

Out of the 1,127 original vehicle-owning households in the unweighted sample, 76 gave up their vehicles and did not replace them between the two surveys. The reasons stated for giving up vehicles are shown in Table 5.3: - 82 -

Table 5.3: REASONS FOR NOT REPLACING VEHICLE

Expense too high 20 ALS 5 Road tax 10 Petrol price 6 Can't afford/no money 20 License suspended 2 Nobody to drive 17 Parking problem 7 No need for car 4 Other 9

Sixty-one percent cited the expense of owning and using a car as the reason for not replacing their cars. However, only five percent specifically attributed the problem to the Area License Scheme, less than either road tax or petrol prices. Although for some the Area License fee would represent an element in the general categories of "too expensive" and "cannot afford", the responses indicate that the Area License Scheme was not explicitly viewed as a major factor in the increased cost of owning and operating a car. It may be concluded, therefore, that the Area License Scheme was not a primary cause of a reduction in car ownership.

The trip rates for the 76 households that gave up their vehicles between the two surveys are shown in Table 5.4:

Table 5.4: TRIP RATES BY PURPOSE (All modes) (Ex-Vehicle-Owning Households)

Personal Social/ To Work To Shop Business Recreational To School Total

Before 2.72 0.44 0.37 0.37 1.81 5.71 After 1.93 0.39 0.32 0.14 1.66 4.45

Percent change -29% -11% -14% -62% - 8% -22%

This group suffered a dramatic reduction in the number of trips thiat they made. It is scarcely credible that this change was s;rply the result of the decision to give up owning a car. In that case, it would be reasonable to expect most work trips to continue to be made. In fact, the data indicate that work trip rates fell more than school, shopping, and personal business trip rates. One possible explanation is that both the decline in work trips and giving up the car were the result of unemployment, with many school and shopping trips still being made. - 83 -

Summary: Ex-Vehicle-Owning Households

It seems unreasonable to attribute these dramatic reductions in trip rates to the introduction of the Area License Scheme. Except to the limited extent that the Area License fee may have been one of the elements of the increase in car costs that led some people to give up their cars, it may be concluded that the Area License Scheme did not have an important influence on the decisions of people to give up their cars or on their subsequent trip-making behavior.

TRIPS FROM VEHICLE-OWNING HOUSEHOLDS TO WORK IN THE RESTRICTED ZONE

As expected, the main effects of the Area License Scheme and the increased parking charges on work trips were not on the number of trips made but on how they were made. The analysis of these trips focuses mainly on those people who made trips to work in the Restricted Zone and, within this group, on those coming from vehicle-owning households, who might, therefore, have had access to a car.

The options available for members of this group included a variety of modes and a range of times at which to schedule their trips. Not every one had the full range of choices available to him. For example, if a household owned one car and had several members going to work, not all of them (independently) would have the option of driving the car. In addition, fixed work schedules constrained the timing of work trips for many people, although the option was open of making the trip earlier than the work schedule required. Although most individuals may have had somewhat limited choice, nevertheless within the group as a whole there was considerable flexibility in trip scheduling and in choosing among modes.

CHOICE OF MODE

The categories of mode recorded in the interviews reflect the variety of alternatives open to the downtown commuter in Singapore. They were:

1. car driver (non-car-pool); 2. car passenger (non-car-pool); 3. car-pool driver (defined as driver of a car with at least 4 occupants); 4. car-pool passenger (in a car with at least 4 occupants); 5. bus rider; 6. motorcycle driver or passenger; 7. taxi rider; 8. trishaw rider; 9. bicylist; 10. pedestrian; - 84 - and after the introduction of the Area License Scheme:

11. shuttle-bus rider.

Preliminary tabulations showed that the predominant modes were the first six in this list. Thus, these modes were distinguished for the analysis. In addition, shuttle-bus riders were identified where appropriate. In most cases, the remaining modes were grouped together and labeled "other."

Members of vehicle-owning households making work-trips to the Restricted Zone, but not necessarily during the restricted hours, chose the different modes in the proportions shown in the "modal split" diagrams of Figure 5.1.

Before the introduction of the Area License Scheme, 56 percent of people making work trips to the Restricted Zone from vehicle-owning households utilized a car either as drivers or passengers, compared with 33 perceniL who used the bus. After the introduction of the Area License Scheme, the relative bus and car shares changed significantly, the bus share rising from 33 percent to 46 percent (including shuttle bus) and the private car share falling from 56 percent to 46 percent. Thus, post-ALS, the shares were equal. Use of the park-and-ride facilities was negligible, and the shuttle buses were used mainly as a supplement to other bus services.

The fall in the proportion of car-driver trips from 34 percent to 27 percent (including both car-pool and non-car-pool drivers) indicates that the number of cars being driven on work trips to destinations within the Restricted Zone declined by about 20 percent. (This refers to trips at all hours, not just during the restricted period. Changes in trip timing are reported below.)

The number of people from vehicle-owning households who made their work trips into the Restricted Zone by car, either as drivers or as passengers, fell by 18 percent, nearly in proportion to the decline in the number of cars. This appears to conflict with the increase in car-pooling observed in the vehicle occupancy counts. It may also appear strange that, with the Area License Scheme in effect, the number of car-pool passengers was less than twice the number of car-pool drivers, although a car pool, by definition, must have at least three passengers and only one driver. The explanation is that two important groups of car-pool passengers were excluded from the above analysis -- passengers from non-vehicle-owning households and passengers making trips for purposes other than going to work. With the exemption of car pools from the license requirement, people who wished to drive to work had an incentive to take passengers with them, and many of the passengers were members of these other groups. - 85 -

Figure S.1 Mode use for work trips to the Restricted Zone from vehicle-owning households.

BEFORE ALS

Driver (2%) Other (4%) Car Passenger (16%) Passenger (6%)

Motorcycle (7%)

AFTER ALS

4 ~~~Bus (43%) tX

|~~~~~~~~~aDrver (20%) /

Shuttle (3%) ri/|ver) > Other (2%) Passenger

Motorcycle (6%) C

Car Passenger (7%)

Note. Sample sizes before ALS. 1,199 trips. after ALS, 1,026 tips - 86 -

The aggregate numbers presented above should be interpreted with caution: the changes in modal proportions suggest that 10 percent of commuters switched from driving and riding in cars to riding buses. In fact, the changes are much more complex. For example, some car drivers became car passengers; some passengers formed larger groups to qualify as car pools; some shifted to the bus; and some bus passengers joined car pools. For a more detailed examination of modal changes a subsample was extracted, comprising 720 individuals from vehicle-owning households each of whom reported making a trip directly from home to a work-place in the Restricted Zone both before and after the introduction of the Area License Scheme. 1/

Out of this sample, 242 traveled to work by bus before the introduction of the Area License Scheme, and 326 did so afterward. As illustrated in the diagram below, 208 of the original group (86 percent) continued to use the bus, and 34 (14 percent) shifted to other modes. The 208 were joined by 118 new bus riders who formerly used other modes of transport to go to work. Thus the 35 percent increase in bus riders (in this vehicle-owning sample) was the net result of a 14 percent loss and a 49 percent increment.

Before 2

208 All Other Modes

118

After 326 Units = trips

1/ See the Annex to this Chapter for a more detailed analysis of the "micro" subsample. - 87 -

A similar analysis of car use for trips to work by car drivers and passengers (including car pools) is summarized by the following diagram.

Before C

303 Other

After ar 338 Units = trips

Of the 405 people who formerly traveled to work in cars, 303 (75 percent) continued to do so, and were joined by 35 more who switched from other modes, so that the net decline was only 67 (17 percent).

Going down to still finer detail, the micro analysis revealed that the greatly increased number of car poolers (both drivers and passengers) making work trips included only 61 percent of the original group of car poolers. The numbers leaving, staying in, and joining this category are shown below.

Car Before Po

16 9 39on

79 24

After ool 142 Units =trips - 88 -

Evidently, car-pool arrangements are not very stable, and new ones must be continually formed even to maintain a given level of car pooling, while expansion requires correspondingly more.

CHANGES IN THE TIMING OF TRIPS

Another alternative for those travelers who were unwilling or unable to change travel mode or pay for area licenses was to make their journeys earlier, to avoid entering the Restricted Zone during the hours of operation of the Area License Scheme (theoretically, they could also travel later, but very few could adjust their work schedules to enter the Restricted Zone after 10:15 a.m.). The extent to which travelers chose to travel earlier was investigated through an analysis of the times at which journeys to work were made before and after the introduction of the Area License Scheme. The reported starting times for the journeys were examined for car drivers, car passengers, and bus riders.

Throughout the analysis of starting times, it must be remembered that the times are those reported in interviews. It is well known that people are not aware of precise times and that they report times rounded to the nearest five or ten minutes at best, and in many cases to the nearest quarter- or half-hour. This is clearly evident in Figure 5.2, which presents the 7:00 -8:30 a.m. portion of the distribution of times of starting the trip to work (before the Area License Scheme) as reported by a sample of car drivers. While it may well be true that more drivers started at about 8:00 a.m. than at about 7:00, 7:30, or 8:30, it is very unlikely that none of them started at 7:55 or 8:05, or that the numbers were as low as indicated between the quarter-hour points.

Given this lack of precision, no attempt was made to ascertain the time of crossing the boundary of the Restricted Zone, either by questioning or by adjusting the reported starting times to allow for the spatial distribution of starting points. Instead, it was assumed that changes in the proportions of people who reported starting times up to 7:30 a.m. were adequate indications of changes in the proportions who actually entered the Restricted Zone before 7:30 a.m. - 89 -

Figure 5.2. Typical reported starting times.

30 -

20-

Reported starting time (a.m.)

The distributions of starting times for journeys to work in the Restricted Zone by members of vehicle-owning households before and after the introduction of the Area License Scheme are shown in Figure 5.3. Overall, no great changes can be seen, the distributions being similar in shape, with their peaks between 7:30 a.m. and 8:00 a.m. The percentage in the peak half- hour declined from 28.6 percent before the scheme was introduced to 25.5 percent afterward, and there was a small movement toward earlier starting times. More pronounced changes can be detected in the starting time distributions of the users of specific modes before and after the introduction of the Area License Scheme, as analyzed in detail below. - 90 -

Figure 5.3 Starting times for work trips to the Restricted Zone, all modes.

350-

Key 300 - Before ALS 0J After ALS

250

200-

150f

50 W

6 01 6:31 7 01 7 31 8 01 8131 9 01 9.31 10 01 10 31 to to to to to to to to to to 6.30 7.00 7 30 8.00 8 30 9 00 9 30 10 00 10 30 11 00 Time period (a in.)

300

Z 15

I 0

6 01 631 701 7.31 801l 831 901 9 31 10 01 10 31 to to to to to to to to to to 6 30 7,00 7 30 8.00 8 30 9.00 9 30 10 00 10 30 11 00 Time period (a.m)

d Sample 119 trips before, 985 trips after, from vehicle-owning households - 91 -

Car Drivers (non-car-pool). The detailed starting time distributions, expressed in terms of both the absolute numbers and the percentage of car driver trips starting in each period, are presented in Figure 5.4. From this figure, it is clear not only that the number of car driver trips fell (as detailed in the above section on modal changes) but also that the distribution of starting times for car trips changed significantly, with an increase in the proportion of trips starting before 7:30 a.m., when the Area License Scheme goes into operation. The changes are highlighted in the histogram showing the percentages of car trips starting in each time period. The proportion of car trips that started before 7:30 a.m. jumped from 28 percent before the introduction of the Area License Scheme to 42 percent afterwards.

Car Passengers (non-car-pool). The distributions for car passengers are presented in Figure 5.5. There is a noticeable reduction in the peak. This is caused partly by a reduction in the number of trips made as car passengers, but it is also the result of a shift towards earlier starting times. Before the introduction of the Area License Scheme, 19 percent of non-pool car passenger trips started before 7:30 a.m., compared with 38 percent afterward.

Car Pools. While the number of car-pool drivers and passengers increased markedly, the proportion of them who started their journeys to work in the Restricted Zone before 7:30 a.m. remained virtually constant (29 percent before; 30 percent after). (See Figure 5.6.)

Bus Riders. The distribution of starting times for trips by bus riders after the introduction of the Area License Scheme shows more of a peak than the beforehand distribution, as shown in Figure 5.7. The number of trips started before 7:30 a.m. did not change, but because the total number of trips by bus increased, the percentage starting before 7:30 a.m. declined from 38 percent to 32 percent. The percentage of bus trips made in the hour after 7:30 a.m. increased from 50 percent to 55 percent, numer- ically, an increase from 183 to 246 trips. - 92 -

Figure 5.4 Starting times of car drivers' work trips to the Restricted Zone.

Key _00- Before ALS J After ALS

100-

6 01 6:31 7 01 7:31 8 01 8'31 9.01 9:31 10:01 10 31 to to to to to to to to to to 6 30 7 00 7 30 8.00 8.30 9-00 9 30 10:00 10 30 11:00 Time period (a.m.)

25k

6.01 6 31 7 01 7.31 8 01 8 31 9 01 9 31 10 01 10 31 to to to to to to to to to to 6:30 7.00 7:30 8-00 8.30 9.00 9 30 10 00 10 30 11 00 Time period (a.m )

a Sample 360 trips before, 193 trips after, from vehicle-owning households - 93 -

Figure 5.5 Starting times of car passengers' work trips to the Restricted Zone. 150

Key 100 Before ALS After ALS

6.01 6.31 7:01 7:31 8 01 8 31 9.01 9131 10.01 10 31 to to to to to to to to to to 6:30 7:00 7.30 8 00 8:30 9 00 9:30 10 00 10 30 11 00 Time period (a.m.)

35r

25K

2 20_

5- [e X

6 01 6 31 7 01 7 31 8 01 8 31 9 01 9 31 10 01 10 31 to to to to to to to to to to 6 30 7 00 7.30 8.00 8 30 9 00 9.30 10.00 10 30 11 00 Time period (a.m.) a Sample t89 trips before, 69 trips after, from vehicle-owning households - 94 -

Figure 5.6 Starting times of carpools' work trips to the Restricted Zone.

150 -

~~~~~~~~~~~Key. 100 - Before ALS SD After ALS

50 _

O A1_ f 91 1§l731 8:0 6 01 6.31 7 01 7 31 8:01 8.31 9:01 9.31 10 01 10 31 to to to to to to to to to to 6:30 7-00 7:30 8.00 8.30 9.00 9.30 10 00 10 30 11 00 Time period (a.m.)

35 _

30-

25-

' 5

0 6 01 6 31 7 01 7.31 8 01 8 31 9 01 9 31 10 01 10 31 to to to to to to to to to to 6 30 7 00 7 30 8 00 8 30 9 00 9 30 10 00 10 30 11 00 Time period (a m.)

a Sample 94 trips before. 195 trips after, from vehicle-owning households - 95 -

Figure 5.7 Starting times of bus riders' work trips to the Restricted Zone.

150

Key iOO Before ALS CM After ALS

6:01 6 31 7:01 7 31 8 01 8 31 9.01 9 31 10 01 10 31 to to to to to to to to to to 6.30 7 00 7.30 8 00 8.30 9 00 9 30 10:00 10 30 11:00 Time period (a. )

35r

30K

25-

20-

15 -

6 01 6 31 7 01 7 31 8 01 8 31 9.01 9 31 10 01 10 31 to to to to to to to to to to 6 30 7 00 7 30 8.00 8 30 9 00 9 30 10 00 10 30 11 00 Time period (a ni)

a Sdmple 364 trips before. 451 trnps dater, from vehitle-owning households - 96 -

COMBINED MODE AND TIME ANALYSIS FOR PRE-ALS CAR DRIVERS

In order to examine in more detail the changes in mode and timing on the part of those most directly affected, an in-depth analysis was made of the behavior of respondents who made home-to-work trips into the Restricted Zone during the restricted hours (7:31-10:15 a.m.) as car drivers (non-car--pool) before the introduction of the Area License Scheme. The mode that they used and the time at which they started the trip after the introduction of the scheme are presented in Table 5.5.

It is particularly interesting (a) that 78 percent of the sample continued to travel to work by car (69 percent as drivers and 9 percent as passengers), and (b) that 74 percent of the sample continued to start their journeys during the restricted hours (whatever mode they used).

Out of the sample of car drivers who used to drive into the Restric- ted Zone during the restricted hours, only 36 percent bought licenses and continued to do as before. Eighteen percent continued to drive, but outside the hours of restriction. Fifteen percent more continued driving and carried three or more passengers to qualify for the car-pool exemption. A further 9 percent became car passengers (4 percent in car pools, 5 percent as non-car- pool passengers). It is interesting that three times as many drivers became car-pool drivers as became car-pool passengers.

Besides the 18 percent of these drivers who changed time and continued to drive, another 8 percent changed both time and mode.

CHANGES IN TRAVEL TIME

The Singapore Area License Scheme achieved the objective of reducing congestion -- at least in the morning commuting period. On the face of it, the achievement of this objective might have been expected to lead to reduced travel times. Therefore, a comprehensive analysis of the starting and arrival times reported in the interviews was carried out in order to find any large differences that might have been perceived between travel times before and after the Area License Scheme was introduced. This analysis is presented in the Annex to this chapter. The question of the consistency of the speed and travel time measurements is discussed in Chapter 6.

The changes in travel time for travelers who did not change mode were very small. Car drivers, on average, reported taking about one minute longer, bus riders from vehicle-owning households reported taking, on average, half a minute longer. However, bus riders from non-vehicle-owning households, who constitute the vast majority of bus riders in Singapore, reported taking, on average, about one minute less. Car drivers who changed to the bus took an average of nine minutes longer, and bus riders who changed to the car took an average of nine minutes less. - 97 -

Table 5.5: MODE AND TIME CHANGES FOR PRE-ALS CAR DRIVERS WHOSE TRIPS STARTED BETWEEN 7:31 and 10:15 a.m.

(non-car-pool)

Time of Starting Trip Post-ALS

Before 7:31- After 7:30 10:15 10:15 Total Car Driver (Non-Pool) 14% 36% 4% 54% Mode Car Passenger (Non-Pool) 1% 3% 1% 5% Used Car Pool: Post- Driver 1% 14% 0% 15% Passenger 1% 3% 0% 4% ALS Bus 2% 16% 1% 19%

Other 1% 1% 1% 2%

Total 20% 74% 6% 100%

Purpose: Home to Work. O/D Pattern: To Restricted Zone. Sample Size: 172 trips

Explanations for the changes in travel time of people who changed mode are straightforward. However, it is difficult to relate changes in travel time by a given mode to the changes in traffic conditions that resulted from the Area License Scheme.

TRIPS FROM VEHICLE-OWNING HOUSEHOLDS TO WORK BEYOND THE RESTRICTED ZONE

After travelers to work places in the Restricted Zone, the group most affected by the Area License Scheme was made up of automobile travelers who lived on one side of the Restricted Zone and worked on the other. If their routes before the Area License Scheme went through the Restricted Zone, they were directly affected in the same way as those with destinations inside the Restricted Zone, and the same set of alternatives was available to them -- pay for the license, form car pools, change the time of making the trip, take the bus, and so on. In addition, they had the option of altering their routes to bypass the Restricted Zone. In order to study the responses - 98 - of this group, an examination was made of the behavior of a macro-level sample of respondents whose homes and work places lay on opposite sides of the zone that was to be restricted. The results of analyzing changes in mode, time of starting the trip, and route for members of this subsample are presented below.

CHOICE OF MODE

The proportions of travelers using the different modes are shown in Figure 5.8. Before the introduction of the Area License Scheme, 52.5 percent of trans-Restricted-Zone work trips from vehicle-owning households were made by car, compared with 31.5 percent by bus. After the introduction of the Area License Scheme, the proportion of trips by bus rose to 40 percent, but that of car trips fell only 2.5 points, to 50 percent. Much of the increase in bus ridership was at the expense of motorcycles and "other" modes.

Within the car category, a significant increase in car pools is revealed. Before the Area License Scheme, 5 percent of all car drivers in the sample qualified for car-pool status by carrying 3 or more passengers; afterward the proportion was 16 percent.

This increase, however, did not represent many drivers combining into pools. That would have converted many former drivers into passengers, but the car-driver proportion (pool and non-pool) declined only slightly, from 39.5 percent to 37 percent. Also, it is noted that the total car passenger share remained unchanged at 13 percent, and the ratio of car-pool passengers to car-pool drivers in the sample was less than 1.5. Thus, the new car pools must have incorporated many passengers who were excluded from the sample either because they were from non-vehicle-owning households or because their trips were not for the purpose of going to work. - 99 -

Figure 5.5 Mode use for trips to work beyond the Restricted Zone from vehicle-owning households.

BEFORE ALS

\ g M~~otorcycle | Car e°O\/c Other (3%) (13%) Passenger (I | (10%) /\ Driver (2%) Passenger (3%)

AFTER ALS

Other (I) otrc)l

Driver (6%) Car Passenger (5%) Passenger (8%)

Noie Sample sizes before ALS, 372 trips. after ALS. 275 trips - 100 -

CHANGES IN THE TIMING OF TRIPS

The proportions of travelers who started their trans-Restricted- Zone journeys to work before 7:30 a.m. (the start of the restricted hours) are presented in Table 5.6.

Table 5.6: STARTING TIMES OF TRANS-ZONE WORK TRIPS

Percentage starting before 7:30 a.m.

Car Car Bus Driver Passenger Rider Other Total

Before ALS 50 40 73 65 57 After ALS 60 64 65 50 61

Sample: Vehicle-Owning Households. Sample Sizes: Before = 340; After = 254.

The most interesting features are the results for car drivers and passengers:

Car Drivers. Before the introduction of the Area License Scheme, about 50 percent of the trips by car drivers (including car pool drivers) started before 7:30 a.m. After the introduction of the scheme, this proportion had risen to 60 percent.

Car Passengers. A greater change can be seen for car passengers. Before the Area License Scheme, 40 percent of trips were started before 7:30 a.m.; after the introduction of the Area License Scheme, this had risen to over 60 percent.

In both of these categories, the absolute numbers had fallen as a result of the modal shift away from private cars noted above. For those who remained in their cars, it is clear that many took advantage of an early start to avoid paying the Area License fee.

CHANGES IN ROUTE

Commuters in the trans-Restricted-Zone category also had the alternative of changing route and detouring around the Restricted Zone to avoid paying the license fee. Of course, people do not always travel by the shortest route. Before the introduction of the Area License Scheme, 12 percent of car drivers making trans-Restricted-Zone journeys to work used circumferential routes and did not pass through the area that was to become the Restricted Zone; 88 percent did drive through the zone. After the introduction of the scheme, the proportion driving around the Restricted Zone rose to 34 percent. The other 66 percent drove through the zone, but not necessarily during the restricted hours. - 101 -

COMBINED ANALYSIS

The analysis of mode, time, and route changes have indicated that, in each case, important changes have taken place. Thus, a combined analysis was made to reveal the rather complex pattern that results. The analysis recreates the choices of an imaginary group of 100 car drivers. The before- and-after mode, time, and route proportions are those reported above. Two key assumptions were made: (a) that the proportion driving through the Restricted Zone before the Area License Scheme was the same for both the pre-7:30 a.m. and the post-7:30 a.m. periods. (Note that should this assumption be false, with more people driving around after 7:30 a.m. due to downtown congestion, the conclusions drawn below are strengthened); (b) that, with the Area License Scheme, there would be no incentive for pre-7:30 a.m. commuters to change their route, so that the proportion of them driving around the Restricted Zone would be the same as before the Area License Scheme. Based on these assumptions, the exercise deduced the proportions using circumferential versus through routes, in such a way as to preserve the relevant control totals. The results are shown in Figure 5.9.

Figure 5.9: COMBINED MODE, TIME, AND ROUTE CHANGES

MODE TIME ROUTE

BEFORE: CAR DRIVERS Pre-7:30 a.m. Around 6 (Overall) 50 Around 12% Through 44 100 50 Around 6 Through 88% Post-7:30 a.m.

Through 44

AFTER: CAR DRIVERS Pre-7:30 a.m. Around 7 (Overall) Around 57 34%

94 Through 50 (6 changed mode) 37 Around 25 Through 66% Post-7:30 a.m.NX

Through 12

While the overall route split with the license scheme in effect was 66:34 in favor of through routes, the split for those traveling after - 102 -

7:30 a.m. was just the reverse -- 32:68 in favor of circumferential routes. with Comparing the latter ratio -- 32 percent through: 68 percent around -- that before the license scheme -- 88 percent through: 12 percent around -- could reveals a much more dramatic shift towards circumferential travel than times. be detected from the analysis of route choice without specifying the by It is clear that much of the congestion on the ring road can be explained this change.

CHANGES IN TRAVEL TIME

The mean travel times, based on reported times of starting and arrival, for trips made before and after the introduction of the Area License Scheme are:

Bus Riders Car Drivers Car Passengers

Before 48 mins. 33 mins. 33 After 49 mins. 32 mins. 30

Thus, there is no evidence that trans-Restricted-Zone travel times changed appreciably.

TRIPS TO WORK FROM NON-VEHICLE-OWNING HOUSEHOLDS

As expected, only minor changes are apparent in the travel choices or of people from non-vehicle-owning households who go to work either in were beyond the Restricted Zone. Between 85 and 90 percent of these travelers (The bus riders both before and after inauguration of the license scheme. For com- "vmodal split" diagrams are presented in the annex to this chapter.) either muters into the zone, there was a slight shift toward riding in cars, share) with or without car-pool occupancy, while taxis (formerly a two percent for were no longer used. No statistically significant changes were recorded trans-zone workers. The timing of work trips showed no appreciable change for either group of commuters from non-vehicle-owning households.

SUMMARY -- TRIPS TO WORK

The impacts on people making work trips to the Restricted Zone in The a variety of travel behavior categories are summarized in Table 5.7. principal changes resulting from the Area License Scheme were the changes in the behavior of drivers traveling into the Restricted Zone, and bypassing of the zone by drivers with destinations beyond it. - 103 -

Table 5.7: PRINCIPAL IMPACTS OF TRAFFIC POLICIES ON WORK-TRIP BEHAVIOR

WORK TRIPS TO THE RESTRICTED ZONE TRANS-RESTRICTED-ZONE WORK TRIPS ------A- - - - - C------BUS riders before ALS CAR usersabefore ALS CAR users/ before ALS BUS riders before ALS Cost of driving to RZ during Cost of driving thru RZ ALS hours rose from S$153 to during ALS hours rose by COSTS AND No change in coat S$228 per month. ALS fee only. REVENUES No change in cost These two groups pay S$245,000/month in Area License fees. Bus share of trips rose from 33% to 46%. Car share fell from Bus share of trips rose from 31.5% to 4-0T%. 56% to 46%. 62% of tripmakers did not change mode. Car share of trips fell from 52.5% to 50% CHA-NGES Of bus riders, 86% did not Carpool share of car trips Carpool share of car trips IN change, 8% joined carpools. rose from 16% to 42%. rose from 9% to 28% MODE For non-vehicle-owning house- Of car drivers: For non-vehicle-owning holds, bus share remained 55% did not change households, about 88%. bus share 19% formed/joined carpools remained about 90%. 19% changed to the bus

Percentage leaving home before Percentage leaving home be- Percentage leaving home Percentage of bus riders 7:30 am fell from 38% to 32% fore 7:30 am: before 7:30 am: leaving home before CHANGES largely due to increased 7:30 bus car drivers - up from car drivers - up from am fell from 73% to 65% IN TRIP ridership after 7:30 am. 28% to 42% 50% to 60% SCHEDULING passengers - up from passengers - up from 19% to 38% 40% to 64%

Bus riders from vehicle-owning Car users who remained in The average car driver households lost an average of the car mode (driver,passen- saved 1.2 minutes. CHANCES 0.6 minutes if they stayed on ger,or carpool) lost about The average car passenger IN the bus and saved an average The average bus rider one minute on average, saved 2.8 minutes. lost 0.8 minutes. TRAVEL of 9 minutes if they change TIME to a car mode. Car users who switched to the bus lost an average of Bus riders from non-vehicle- 9 minutes. owning households saved an average of 1.2 minutes.

CHANGES Work trip rate to the Restricted Zone fell by 11% from IN TRTP ~~~~~~~~~~~~~~~~~~~~Notavailable RATES 0.75 to 0.67 per household per day.

% of drivers who drove ROUTE Not applicable Not applicable restrictedaround the RZ during from Not applicable CHANGES hours rose fo 12% to 68%. |

/a Car users" includes drivers and passengers, pool and non-pool, Note: based on interviews; vehicle-owning households unless otherwise noted. - 104 -

Of trips made by members of vehicle-owning households to work inside the Restricted Zone, the proportion made by car, either as driver or passenger, declined from 56 to 46 percent, while the proportion made by bus rose from 33 percent to 46 percent. Many people who continued to drive cars either carried three or more passengers in order to qualify for the car-pool exemption or changed their schedules to avoid the restricted hours.

About half of those who used to drive to work-places in the Re- stricted Zone during the restricted hours with less than car-pool occupancy continued to drive at that time, 36 percent of the original non-pool drivers buying licenses and 14 percent taking three or more passengers. Another 19 percent continued driving but rescheduled their trips, and the remaining 30 percent gave up driving for other modes of travel.

People from vehicle-owning households who formerly rode to work in the zone as car passengers changed their travel practices in the same ways as, but to lesser degrees than, the drivers. Car pooling increased notably.

Few car drivers going to work places beyond the Restricted Zone changed mode. Some rescheduled their trips to go through the zone before 7:30 a.m., some formed car pools, and about two-thirds of those who made the trip during restricted hours detoured around the Restricted Zone.

Commuters from non-vehicle-owning households to work in or beyond the Restricted Zone reported few changes. Most of them traveled by bus both before and after institution of the Area License Scheme.

People who used to make trips by car to the Restricted Zone during the restricted hours before the introduction of the Area License Scheme suffered a large part of the adverse impact of the Area License Scheme. Overall, about 26,600 cars are involved. This implies about 37,000 people. A large proportion of these people changed their behavior. Not only did this involve inconvenience to the traveler, but also, in many cases it resulted in more time being spent on travel. Those who chose to pay the Area License fee instead of changing their behavior had less traffic to contend with, but they did not gain in terms of travel time.

Like car users making trips to the Restricted Zone, those who used to drive across it during the restricted hours also had to make changes in their travel behavior or pay the extra cost of area licenses. About 16,000 cars (22,400 people) are involved. The existence of the ring road made the change somewhat easier for them, and even though ring road congestion was quite heavy, the trans-Restricted Zone travelers reported a small improvement in travel time.

Bus riders were also affected relatively little by the scheme, particularly in the sense that they did not have to change their travel habits or incur higher cost. Those who managed to join car pools enjoyed savings in travel time. Those from vehicle-owning households reported - 105 -

a very small increase in travel time, but those from non-vehicle-owning households reported a small decrease.

Overall, it appears that the majority of travelers (mainly bus riders from non-vehicle-owning households) have been made slightly better off. Most car users were made worse off, but none greatly so. There is no evidence that the people who were made worse off were from lower income groups (see annex to this chapter). Thus, it may be concluded that no in- equity was caused by the distribution of benefits and disbenefits.

TRIPS HOME TO VEHICLE-OWNING HOUSEHOLDS FROM WORK IN THE RESTRICTED ZONE

CHOICE OF MODE

For their trips home from work in the Restricted Zone, members of vehicle-owning households chose the different modes in the proportions shown in the mode split diagrams of Figure 5.10.

Similarities to Work-Trips to Restricted Zone

The changes that emerge from this before and after comparison parallel closely the changes observed in the modal splits for work-trips to the Restricted Zone, with the same two prominent shifts: first, from private to public transport and, second, to car pools.

The proportion of trips made by car as drivers or passengers (including trips in car pools) fell from 53 percent to 43 percent, while the proportion traveling by bus rose from 36 percent to 48 percent. This means that, before the Area License Scheme, for every trip made by bus, 1.5 trips were made by car. With the scheme in effect, this ratio changed to 0.90:1, i.e., for every trip made by bus, only 0.90 trips were made by car. As in the case of trips to work in the Restricted Zone (where the ratio fell from 1.7:1 to 1:1), the modal shares have moved from car predominance to virtual equality.

Not only did the number of people making homeward trips by car fall; the number of cars used also changed appreciably. Before implementation of the Area License Scheme, the car-pool share was 5 percent of all trips (9.5 percent of car trips); afterward, it had risen to 12 percent of all trips (28 percent of car trips). This represents a major movement toward shared riding. The proportion of trips made as car driver (car-pool and non-car-pool) went down from 37 percent to 27 percent, indicating that the number of cars being driven home from work in the Restricted Zone fell by about 27 percent. - 106 -

Figure 5.10 Mode use for trips home from work in the Restricted Zone to vehicle-owning households.

BEFORE ALS

Bus (36%) Car Driver (35%)

Other (4%) ar Passenger (13%) ~~~~Driver (2%)

Motorcycle (7%) Passenger (3%)

AFTER ALS

/ ~~~Bus (46%) / <

/~~~~~~~~~~~~a Driver (23%) /

Shuttle (2%) -//\ ,\ Other (3%) Car CO

Motorcycle (6%) Driver (4%)

Pdssenger (8%)

Note Simple sizes before ALS. 1,149. altcr ALS. 1.019 trips - 107 -

Differences from Work-Trips to Restricted Zone

One interesting feature distinguishes the modal splits for work- trips to the Restricted Zone from the return trips home. Both before and after the license scheme, the proportion of trips made as members of car pools was smaller in the evening than in the morning, i.e. more people used car pools to get to work than used them to get home:

Trips in Car pools (Percent)

.Home to Work in Restricted Zone Home from Work in Restricted Zone

Before 8 5

After 19 12

TIMING OF HOMEBOUND TRIPS FROM THE RESTRICTED ZONE

Figure 5.11 shows the distributions of the times when respondents from vehicle-owning households reportedly started home from work in the Restricted Zone. Compared with the distributions shown in Figure 5.3 for starting the inbound trip, those for the trip home are considerably more sharply peaked. An analysis in the annex to this chapter compares distributions of the times for leaving home in the morning and arriving home in the evening, for arriving at and departing from the work place, and for starting and stopping work. As pointed out there, it is logical to expect the distributions to be counterparts in the pairs just named, and to expect the first pair to be the least sharply peaked and the last pair most sharply peaked. These relationships are in fact found in the data, as the annex shows. Thus, it should not be surprising that the times of starting home are more concentrated than those of starting the trip to work.

Comparing the distributions of homebound trip starting times before and after initiation of the Area License Scheme shows no appreciable change. The analysis of trips to work, earlier in this chapter, showed definite shifts toward earlier starting times, and some evidence of corresponding shifts in timing of the trips home was expected, but was not found. A detailed examination (in the annex) of the times reported by car drivers -- the group which changed time the most for the trip to work -- also failed to show any significant change in timing of the trip home.

TRIPS HOME TO VEHICLE-OWNING HOUSEHOLDS FROM WORK BEYOND THE RESTRICTED ZONE

This section presents the basic data for trans-Restricted-Zone trips home from work. The implications of the changes reported here for evening peak traffic patterns are discussed in Chapter 6. - 108 -

Figure 5.11 Starting times for trips home from work in the Restricted Zone.

450 _

400 Key: 350 - Before ALS M |After ALS

300-

250-

-00 200

150

100

4 01 4 31 5:01 5:31 6 01 6 31 to to to to to to 4:30 5 00 5 30 6.00 6 30 7 00 Time period (a.m.)

50_

so lot o ol 4: 20

o ~ ~ t~ 4 01 4 31 5-01 531 6 01 6 31 to to ~~~to to to t 4 30 5 00 5-30 6 00 6 30 700 Time period (a.m a Sample 943 trips before, 854 trips afer, from vehicle-owning households - 109 -

CHANGES IN MODE

The proportions of trans-Restricted Zone commuters from vehicle- owning households who chose the different modes for their journeys home from work are shown in Figure 5.12.

With minor exceptions, the proportions are close to those for the trip to work by the trans-Restricted Zone group. The overall car share remained the same with the license scheme as before -- 51 percent. A small shift toward the bus can be seen, but it took place at the expense of motorcycle and "other mode" use rather than at the expense of car use. Thus, the car-bus ratio did not change as noticeably as for the trip to work.

Among car users, there was an increase in the use of car pools, although not as great as the increase for the trip to work. The proportion of car trips made in car pools rose from 9.8 percent to 19.6 percent. In the work-bound direction the increase was from 9.3 to 28.0 percent. It appears that the increase was largely due to regrouping by passengers who formerly rode in smaller groups. The proportion of car drivers (non-car-pool and car- pool) and, hence, the number of cars on the road, did not change.

CHANGES IN TIMING OF TRIPS

The distributions of times at which trans-Restricted-Zone commuters from vehicle-owning households left work for the journey home are shown in Figure 5.13.

There is no important difference between the before and after distributions, indicating that the staggering of the times of traveling to work was not repeated for the journey home.

CHANGES IN ROUTE

Before the Area License Scheme, 90 percent of trans-Restricted-Zone car-using commuters traveled home by a route that passed through the Restric- ted Zone. After the introduction of the scheme, 80 percent traveled through the Restricted Zone and 20 percent used a circumferential route. Compared with the 34 percent bypassing the zone on the trip to work, the drop to 20 percent bypassing it on the return trip clearly reveals the preference of many drivers for the through route when the restraints are not in force. There is no obvious reason why the proportion using the ring road for the homeward trip should have increased as it did from 10 percent to 20 percent. Possibly some of those who took to the ring road in the morning to avoid the license requirement discovered unexpectedly that they preferred it to the through route. - 110 -

Figure 5.12. Mode use for trips home from work beyond the Restricted Zone to vehicle-owning households.

BEFORE ALS

Bus (34%) E\ Car Driver (36%)|

\/Motorcycle | \S S 5 Other (2%) (13%) Car \ o

Driver (3%) Passenger (2%)

AFTER ALS

4 ~~~Bus(38%)|

t t X ~~~~~~~~~rDriver (35%) )

.- Motorcycle/\/ Other (1%) -_ 'c\°

Driver (4%) Passenger (6%) Car Passenger (6%)

Note Sample sizes before ALS. 353 trips. after ALS. 275 trips Figure 5.13 Starting times for trips home from work beyond the Restricted Zone.

50-

100 - ~~~~~~~~Key- Before ALS After ALS

* 4 0? 4.31 5 01 5:31 6 01 6 31 to to to to to to 4'30 5-00 5 30 6.00 6.30 7 00 Time period (a m.)

60 F

0- IV

4 01 4 31 5 01 5:31 6 01 6 31 to to to to to to 4:30 5:00 5 30 6 00 6.30 7 00 Time period (a m) - 112 -

TRIPS HOME FROM WORK TO NON-VEHICLE-OWNING HOUSEHOLDS

Again (as in the case of trips to work), members of non-vehicle- owning households reported very little change in their travel choices for the trip home from work. Between 85 and 90 percent traveled by bus for trips home from within the zone or beyond the zone both before and after the license scheme. As shown in the annex to this chapter, the modes used changed only by slight amounts. There was no discernible change in timing.

SUMMARY -- TRIPS HOME FROM WORK

In general, the changes in mode that were reported for the trip to work in the morning were maintained for the trip home in the evening, except that there was less car pooling for the homebound trip, both before and after the Area License Scheme took effect.

The shift to earlier starting times for the morning trip to work, however, was not reflected in earlier times for the trip home. Evidently, most of those making earlier trips in the morning had not had their working hours adjusted to earlier times. They were entering the Restricted Zone early to avoid the license requirement but were still working the same hours as before and going home at the same time as before.

Of car drivers returning home from work on the far side of the Restricted Zone, the proportion using the ring road rather than driving through the zone increased from 10 percent to 20 percent for no obvious reason.

As expected, very little change was registered by commuters from non-vehicle-owning households.

TRIPS TO THE RESTRICTED ZONE FOR NON-WORK PURPOSES

Near the beginning of this chapter it was seen that about two-thirds of all trips into the Restricted Zone were trips to work and about one-third were made for all other purposes. Most trips for shopping, personal business, social and recreational purposes, and to attend school had destinations outside the Restricted Zone. Nevertheless, the number of such trips that went into the Restricted Zone was enough to be important, especially since there were some notable changes in how they were made.

SHOPPING TRIPS TO THE RESTRICTED ZONE

Only nine shopping trips to the Restricted Zone were reported by the 499 non-vehicle-owning households. Therefore, this analysis is restricted to trips made by members of vehicle-owning households. - 113 -

Changes in Mode

The modal splits are presented in Figure 5.14.

The "other" category in each case includes several different modes -- walk, motorcycle, taxi, shared taxi, trishaw, etc. -- none of which represents more than a few percent. An insignificant number of trips were made in car pools, which is not at all surprising for shopping trips.

After the introduction of the Area License Scheme the bus share rose 13 percentage points, from 21 percent to 34 percent, while the car (driver and passenger) share fell from 56 percent to 45 percent. This represents a significant modal shift, from a situation in which 2.7 car trips were made for every bus trip to one in which only 1.3 car trips were made for every bus trip. Almost all of the gain in the bus share was at the expense of the car driver share, the same number (proportionately) of passengers riding in fewer cars.

Changes in the Timing of Trips

Given the size of the sample, hourly distributions of starting times are not meaningful, so the information for vehicle-owning households has been grouped into four time periods as shown in Table 5.8:

Table 5.8: STARTING TIMES OF SHOPPING TRIPS TO THE RESTRICTED ZONE

Auto Drivers Auto Passengers Bus Riders Before After Before After Before After %__ %7. %7 %i % %

Before 10 a.m. 43 - 38 27 45 46 10 a.m. - 12 noon - 16 7 18 9 - 12 noon - 4 p.m. 36 34 1 - 46 54 After 4 p.m. 21 50 54 55 - -

Sample: Vehicle-Owning Households. Sample Sizes: Before = 56; After = 38.

The most remarkable change is the dramatic drop in the proportion of drivers making shopping trips before 10:00 a.m. from 43 percent to zero. The fact that 50 percent of their trips are now made after 4:00 p.m. is interesting and may support the hypothesis that these shoppers are timing their trips to pick up other family members and take them home from work, helping to account for the continued high level of evening peak traffic flows. - 114 -

Figure 5.14. Mode usefor shopping trips to the Restricted Zonefrom vehicle-owning households.

BEFORE ALS

/9g C~~~ar Driver (26%)

Bus (21%) / / \ ~~Car Passenger (30%)

< O~~ther(23%) \

AFTER ALS

Bus (34%) S

Note Sdtmple sizes. before ALS, 54 lnp>,. after ALS, 38 trip-, - 115 -

PERSONAL BUSINESS TRIPS TO THE RESTRICTED ZONE

"Personal Business Trips" are defined as trips made to visit a doctor, lawyer, bank, etc. Over 80 percent of personal business trips made by people from non-vehicle-owning households were made outside the Restricted Zone, mostly by bus. There is no evidence that origin-destination patterns for these trips have changed since the introduction of the Area License Scheme, or that any important changes occurred in mode or timing of the trips that did go into the Restricted Zone.

Changes in Mode (Vehicle-owning households)

The proportions of personal business trips to the Restricted Zone made by people from vehicle-owning households by the different modes are shown in Figure 5.15.

The proportion of trips by private car fell by almost ten percent, but almost all of the reduction was in trips by car passengers, not drivers. It seems likely that the reduced number of cars entering the Restricted Zone has resulted in reduced opportunities for getting a ride to the central area to carry out personal business. The taxi share also fell dramatically from nearly ten percent to zero. These two changes resulted in an increase in the bus share of 15 percentage points.

Changes in Timing (Vehicle-owning households)

The proportions of travelers starting their personal business journeys at different times of the day are shown in Table 5.9.

Table 5.9: STARTING TIMES OF PERSONAL BUSINESS TRIPS TO THE RESTRICTED ZONE

Before ALS After ALS

Before 7:30 a.m. 5% 7% 7:30 - 10:00 a.m. 36% 34% 10:00 a.m. - 12:00 noon 26% 19% 12:00 noon - 2:00 p.m. 6% 9% 2:00 - 4:00 p.m. 9% 8% 4:00 - 6:30 p.m. 5% 11% After 6:30 p.m. 13% 11%

Sample: Vehicle-Owning Households. Sample Sizes: Before = 86; After = 53.

Overall, few large changes have taken place. The figures for the 7:30-10:00 a.m. period show little change. The only large changes are the reduction in trips between 10:00 a.m. and noon and the increase between 4:00 and 6:30 p.m. A detailed investigation of the data shows that this change is almost entirely due to changes by bus riders. Unless they travel later in order to be brought home by someone else in a car, it is difficult to explain this change. - 116 -

Figure 5.15 Mode use for personal business trips to the Restricted Zone from vehicle-owning households.

BEFORE ALS

Bus (42%) /\\ / ar Drive (17 C°5%\

Other (2%) Walk (1%) Taxi (9.5%)

Motorcycle (6%)

AFTER ALS

< \~~~~~~~~~ar Driver ( 17%)

Other (2%) \ / \/ / ~~~Passenger\//Car\// Walk (4%) (13%/

Motorcycle (7%) / ,_

Note Sample sizes before ALS, 86 trips. aiter ALS, 53 trips - 117 -

SOCIAL/RECREATIONAL TRIPS TO THE RESTRICTED ZONE -- VEHICLE-OWNING HOUSEHOLDS

"Social/recreational" trips are those made for the purposes of visiting friends, attending or participating in cultural, recreational, or sporting events, and eating out. In our sample of vehicle-owning households, 21 percent of social/recreational trips were trips into the Restricted Zone, 68 percent were made completely outside the zone, and all other origin- destination patterns together made up the remaining 11 percent.

Changes in Mode and Time

The car share fell from 53.5 percent to 44 percent while the bus share rose from 38.5 percent to 46 percent. The number of these trips made before noon rose very slightly, but the major change was a fall in both the number (36 to 10) and the proportion of trips made after 6:30 p.m. In the latter case, the changes were very largely due to time changes by car drivers and passengers. It is difficult to understand how these changes could be related to the introduction of the Area License Scheme.

SCHOOL TRIPS TO THE RESTRICTED ZONE

Clearly, children making trips to school do not have the same range of modes available to them as adults do. Nevertheless, the modal shifts noted above for work trips are reflected in the school trip data of Table 5.10, which show a move from car to bus.

Table 5.10: MODAL SPLIT: SCHOOL TRIPS TO RESTRICTED ZONE

Before ALS After ALS

Bus 66.2% 72.0% Car passenger 14.5% 8.4% Car-pool passenger 11.6% 10.5% Other 7.7% 9.1%

An examination of the times at which children started their trips to school reveals no evidence that the Area License Scheme has resulted in these trips being moved to before 7:30 a.m. This is probably because school schedules are less flexible than work schedules.

TRIPS OUTSIDE THE RESTRICTED ZONE

Obviously, the Area License Scheme had its greatest effects on travel into the Restricted Zone and from one side of the zone to the other. These primary effects, however, might have repercussions that would change the nature of trips that went neither into the Restricted Zone nor from one side of the zone to the other. One hypothesis, for example, was that cars - 118 - left home by people who formerly drove them to work would be used by other members of the family for trips outside the zone. To investigate possible secondary effects of that sort, trips entirely outside the zone and outside of the trans-zone category were analyzed for changes in mode and timing. The results are reported below, categorized by trip purpose.

WORK TRIPS OUTSIDE THE RESTRICTED ZONE

The modal shares for people from vehicle-owning households making work trips outside the Restricted Zone were:

Car Car Car Bus Walk Driver Pass. Pool Motorcycle Other

Before ALS 32% 8% 29% 7% 11% 10% 3% After ALS 34% 9% 31% 5% 9% 9% 3%

In addition, 56 percent of work trips were started before 7:30 a.m. before the introduction of the Area License Scheme compared with 55 percent afterward.

For people from non-vehicle-owning households the modal shares were:

Bus Walk Bicycle Car Pass. Other

Before ALS 67% 21% 6% 2% 4% After ALS 68% 19% 7% 6% 0%

Before the introduction of the Area License Scheme, 66 percent of trips were started before 7:30 a.m. compared with 65 percent afterward.

These data show clearly that the travel behavior of people making work trips outside the Restricted Zone has been virtually unaffected by the Area License Scheme.

SHOPPING TRIPS OUTSIDE THE RESTRICTED ZONE

Changes in Mode

For people making shopping trips outside the Restricted Zone from both vehicle-owning and non-vehicle-owning households, the dominant mode is walking, as can be seen in the modal splits presented in Figures 5.16 and 5.17. - 119 -

Figure5.16 Mode usefor shopping trips outside the Restricted Zone from vehicle-owning households.

BEFORE ALS Car Driver (7%)

Other(5%) 5% Car Passenger (8%)

AFTER ALS Car Driver (4%) Other (2%) I 9 y 8 Car Passenger (5%)

Note Sdmple sizes before ALS, 474 trips, after ALS, 423 trips - 120 -

Figure5.17 Mode usefor shopping trips outside the Restricted Zone from non-vehicle-owning households.

BEFORE ALS - Other (2%)

Walk (90%)

AFTER ALS rI Other (1%)

Walk (94%)

Note Sdmple sizes before ALS, 214 trips. dfter ALS. 168 trips - 121 -

The interesting change to note is that the proportion of trips made by car by members of vehicle-owning households fell from 15 percent to 9 percent. This decline was balanced mainly by an increase in the proportion of trips made on foot. It seems reasonable to assume that this is the result of the increasing costs of automobile operation, rather than the result of the Area License Scheme.

The evidence refutes the prediction that cars left at home by commuters would be used to make additional shopping trips.

Changes in the Timing of Trips

Virtually no difference can be detected in the distributions of starting times for shopping trips outside the Restricted Zone, although it is interesting to note that many trips were being made in the morning peak period.

PERSONAL BUSINESS TRIPS OUTSIDE THE RESTRICTED ZONE -- VEHICLE-OWNING HOUSEHOLDS

For trips outside the Restricted Zone, the car share rose from 46 percent to 52 percent, while the bus share fell from 26 percent to 22 percent. At the same time the proportion of trips made before 10:00 a.m. rose from 36 percent to 51 percent, while the proportions in the 10:00 a.m. to noon and the after 6:30 p.m. categories fell (13 percent to 7 percent and 23 percent to 15 percent, respectively).

SOCIAL/RECREATIONAL TRIPS OUTSIDE THE RESTRICTED ZONE -- VEHICLE-OWNING HOUSEHOLDS

The car share rose from 46 percent to 59.5 percent, while the bus share fell from 28 percent to 21 percent. No large changes took place in the times when trips were made.

SCHOOL TRIPS OUTSIDE RESTRICTED ZONE -- VEHICLE-OWNING HOUSEHOLDS

Children from vehicle-owning households who made school trips outside the Restricted Zone were virtually unaffected by the Area License Scheme. The proportion who traveled by bus rose from 49 percent to 51 percent. No change was observed in the times at which the trips were made; the proportion starting before 7:30 a.m. remained constant at 50 percent.

SUMMARY -- TRIPS OUTSIDE THE RESTRICTED ZONE

None of the changes noted in trips outside the Restricted Zone was very large, nor can any of them be attributed to the Area License Scheme. There was no evidence that cars left home by commuters were used for shopping outside the Restricted Zone by other family members. In fact the use of cars for shopping trips both into and outside the zone declined, perhaps because of increased car operating costs other than the Area License fee. - 122 -

ANNEX TO CHAPTER 5

The analysis of the data obtained from the household survey was extremely widespread. Thus, it was neither possible nor appropriate to present all the results in the text. Some topics, however, are of particular interest to the reader who is interested in pursuing certain points in more detail. With this in mind, more detailed analyses of certain issues are presented in this annex. The topics treated are:

- Micro-level analysis of modal changes;

- Supplementary distributions related to the times at which trips were made;

- Detailed analysis of travel time changes;

- Analysis of changes by different income groups (trips to work from vehicle-owning households).

MICRO-LEVEL ANALYSIS OF CHANGES IN MODE

The aggregate, macro-level results presented in the body of the text should be interpreted with caution: the changes in modal proportions on the surface appear to show that about ten percent of commuters switched from driving cars to riding buses. Tn fact, the changes were much more complex.

A more detailed analysis of these changes was made on the basis of a special sample of 719 work-trips to the Restricted Zone made by people from vehicle-owning households both before and after the introduction of the Area License Scheme. The sample contains only direct home to work trips; all triangular or multi-purpose trips were eliminated.

For these trips, the data on the modes used before and after the Area License Scheme are presented in Table 5.11. Each cell in Table 5.11 includes the number of people corresponding to that particular before-and- after mode combination, together with the row percentages (showing the number in the cell as a proportion of the total number using that mode before) and the column percentages (showing the number in that cell as a proportion of the total using that mode afterwards). The cells on the diagonal of the matrix represent trips by people who did not change mode. Each row shows how those who used a given mode before the Area License Scheme were distributed over the different modes after the Area License Scheme, and each column shows the distribution among "before" modes for those who used a particular mode afterward. - 123 -

Before proceeding to the detailed analysis, a word of caution is in order. While this table shows the changes that took place over the period during which the Area License Scheme was introduced, it would not be correct to conclude that all the changes necessarily resulted directly and exclusively from the Area License Scheme. Modal shares are not static and it would be reasonable to expect, over any given time period, to find some bus riders who changed to the car and vice versa. In other words, the table shows the combined effects of the Area License Scheme and whatever "normal" changes took place during the same period. No evidence exists to indicate what the magnitude of the "normal" changes might be. However, experience with transport systems indicates that they are unlikely to be large in the short run. Thus, when the changes noted in the table are large, they may be attributed mainly to the Area License Scheme. When they are small, firm conclusions should not be drawn, as the changes may simply constitute "normal" adjustments in the system.

Table 5.11: MICRO-ANALYSIS OF MODE CHANGES FOR TRIPS TO WORK IN RESTRICTED ZONE

Vehicle-owning households

Mode after ALS

| Mode II I I I I | Before | | Car/a I Car/a I Car I I TOTAL I I Als Bus I driver lpassenger*l pool I Other I BEFORE I

I Bus | 208 (86%)l 5 (2%)| 3 (1%)| 19 (8%)| 7 (3%) | 242 (100%)| I (64%) I (4%) I (6%) I (13%) I (13%) I (34%) I I I I I I I 1 Car 1 43 (19%)l 123 (55%)! 11 (5%)! 43 (19%)l 4 (2%)l 224 (100%)! I driver/a | (13%) | (87%) | (20%) I (30%) | (7%) I (31%)

I Car/a I 38 (32%)! 11 (9%)| 31 (27%)l 36 (31%)! 1 (1%)| 117 (100%)I Ipassenger 1 (12%) 1 (8%) 1 (57%) | (25%) 1 (2%) 1 (16%)

1 Car ! 16 (25%)! 3 (5%)I 6 (9%)l 39 (61%)l 0 (0%)! 64 (100%)! I pool ! (5%) I (2%) I (11%) e (28%) ! (0)% ! (9%)

1 Other 1 21 (29%)l 0 (0%)! 3 (4%)l 5 (7%)| 43 (60%)l 72 (100%)! (6%) I (0%) ! (6%) 1 (4%) ! (78%) I (10%) I I I ~~~~~~~~~~I I ! IIl I TOTAL | 326 (45%)| 142 (20%)! 54 (7%)l 142 (20%)l 55 (8%)| 719 (100%)| I AFTER 1(100%) 1(100%) |(100%) 1(100%) 1(100%) |(100%) I /a Non-car-pool.

Key: I I No. (Row%) (Col.%) - 124 -

The first row of cells shows that, of the 242 members of the sample who traveled to work by bus before the Area License Scheme, 208, or 86 percent, continued to use the bus, while 8 percent joined car pools and a few changed to each of the other modes. The first column shows that the 208 who remained as bus riders constituted only 64 percent of the new larger total of bus riders (326) with the Area License Scheme in effect, while 13 percent of the new total were former car drivers, 12 percent were former car passengers, and so on. These movements to and from the use of the bus are summari2ed in the diagram below, which was also presented in Chapter 5.

Before B

1 34

-208Moe

118

After Bus 326 Units = trips

The changes made by non-pool car drivers and passengers can be more readily seen in the appropriate rows of Table 5.11. Only 55 percent of pre-ALS non-pool drivers continued as non-pool drivers. A few became passengers in other non-pool cars, while 19 percent joined or formed car pools and 19 percent changed to the bus. Pre-ALS non-pool car passengers made even more drastic changes. Only 27 percent of them remained in this category, 32 percent changed to the bus, 31 percent to car pools, and 9 percent became non-pool car drivers.

The exemption from the area license requirement was a strong incentive for car pooling, and there was a large net increase in the number of car-pool trips (made both as drivers and as passengers). The total rose from 9 percent of the sample before introduction of the Area License Scheme to 20 percent afterward, as shown by the row total and column total in Table .5.11. (Roughly the same proportions are shown in Figure 5.1, based on the macro analysis.) It is interesting, then, to note that only 61 percent of pre-ALS car-pool drivers and passengers continued in that mode. Twenty-five percent switched to the bus, and the rest became non-pool drivers and passengers. Perhaps this indicates a degree of instability in car-pool arrangements. New car pools were formed, of course, in much greater numbers than the break-up of old ones. The 61 percent who continued to travel in pools constituted only 28 percent of the new total. Thirty percent of the new total were former non-pool car drivers, 25 percent were former non-pool passengers, and 13 percent formerly rode buses. - 125 -

The net results of the individual mode changes show up in the "Total Before" column and the "Total After" row in Table 5.11, which give the totals for each mode before and after. These distributions agree rather closely with those from the aggregate analysis of the larger unmatched samples shown in Figure 5.1. The most notable findings from both analyses are two major changes in overall modal usage. First, since the introduction of the Area License Scheme, the proportion of Restricted-Zone work trips by members of vehicle-owning households that were made by bus went up from about 33 percent to about 45 percent, largely, but not entirely, balanced by a decline in car trips (made by both drivers and passengers) from 56 percent to about 46 percent. Second, the proportion of Restricted-Zone work trips by members of vehicle-owning households that were made in car pools more than doubled, going from 9 percent to 20 percent. With the overall decline in car use, the proportion of all car trips that were made to work in car pools by members of this group rose from 16 percent to 43 percent.

SUPPLEMENTARY EVIDENCE ON TRIP SCHEDULING

In Chapter 5, it was reported that a considerable number of travelers were starting their trips earlier to avoid the restricted hours. To investigate this further, the distributions of times of arriving at work and times of starting work were compared with the distribution of times at which the journeys to work were started. These distributions are presented in Figure 5.18. The distributions of the times at which respondents making trips home from work-places in the Restricted Zone (a) finished work, (b) started their journeys home, and (c) arrived home, before and after the introduction of the Area License Scheme, are presented in Figure 5.19.

TRIPS TO WORK IN THE RESTRICTED ZONE

For the users of all modes as a group, either before or after introduction of the Area License Scheme, the distributions of trip starting and arrival times are less sharply peaked than that of work starting time. This result is consistent with the facts that (a) different people will tend to arrive at work at various intervals before and after the nominal starting time (partly by design and partly through random variations in travel time), and (b) people living at different distances would have to start their trips at different times for a given intended arrival time. As might be expected, the peak half hours for arrivals at the trip's destination and for starting work are 30 minutes later than the peak half hour for starting the trip to work, both before and after the introduction of the Area License Scheme.

Car Drivers. The distributions for arrival times and times of starting work support the result noted above that many car drivers are starting their trips earlier to avoid the restricted hours. It is interesting to note that after the introduction of the Area License Scheme there is a small trough in all three time distributions for car drivers. In the trip starting time distribution, it occurs between 7:31 a.m. and 8:00 a.m. and probably reflects the facts that some drivers who used to start their trips in this period (which Figure 5.18. Times of starting trip, arriving, and starting work.

OF STARTING WORK A M TRIP STARTING TIME AM ARRIVAL TIME TIME

480 - Other 480- 480 -

440 - Bus 440 - 440 - 400 - 400 - Car Passenger 400 - 360 - 360 - Car Driver 360-

3- 3_20 3 20 - 280 - 280 - 280 - 240 - 240 240 - O~BEFORE ... BEFORE ~ -BEFORE 200 6200 63200 -

160 -160 -.. **160-

120- 120 . 120 -...

80 -... 80 -.. *80-

40- 40 40 -...

0- 0 0 -M- 6 01 6 31 7 01 7 31 8 01 8 31 9.01 9 31 10.01 6.31 7:01 7:31 8.01 8 31 9.01 9 31 6 01 6131 7 01 7-31 8 01 8.31 9-01 9 31 10-01 6.01 to to to to to to to to to to to to to to to to to to to to to to to to to to 30 7 00 7 30 8 00 8 30 9:00 9.3010 0010.30 6:30 7 00 7.30 8.00 8.30 9.00 9.3010 00 6 30 7 00 7.30 8 00 8 30 9 00 9 3010 0010 30 6 Figure 5.18 Continued

A.M. TRIP STARTING TIME A M. ARRIVAL TIME TIME OF STARTING WORK 360 - 360 - 360 - 320 - 320 - 320 - 280 280 - 280 - Z 240 - 240 - 240

:; 200 , 4 {AFTER 20 AFTER 2 200 AFTER

160 - ,160 160160 120 120 -.. 120 -

80 -80 .... 80- 40 40 -44 0 0 _ 0 0 6 016.317 017:31 8 01 8 319:01 9.31 6 016 317 01 7:318 018 319 01 9 31 10 01 to to to 6 016 317 01 7 31 8 01 8 319 019 31 10 01 to to to to to to to to to to to to to to to 6307007:30800 8309.009301000 to to to to to to to to 630700730 8008:3090093010001030 630700730800 830900930100010.30

Trips from home to work in Restricted Zone from vehicle-owning households Samplesizes Before 1170 After 1026 Figure 5.19 Times of ending work, starting home, and arriving.

HOME TIME OF FINISHING WORK PM DEPARTURE TIME TIME OF ARRIVAL 600 - 600 - 600 - Other 560 - 560 - 560 520- Bus 520 - 520 - Car Passenger 480 - 480 - 480 Car Driver 440 - 440 - 440- 400 - 400 - 400- 360- 3 360 ..* $ 360 - 320 _ 320 - 320 280 d 280 - 280 _ BEFORE 240 BEFORE ¢: 240 _ BEFORE F~~~~z~ ~ 240~ - ~ ~ ~ &~240-24-BFR 200 - 200 - 200- 160 160 - 160- 120 120 - 120 - 80- 880 0-

440 0 - 4 0 6;. 00 731 531 601 631 331 401 431 501 5.31 601 631 331 401 431 501 531 601 631701 2-31301331 401431 501 to to to to to to to to to to to to to to to to to to to to to to to to to 500 530600 630700730 800 300 3304004:30500 5306006:31 700 400 430 5.00 530 600 630 700 400430 Figure 5.19 Continued

TIME OF FINISHING WORK PM DEPARTURE TIME TIME OF ARRIVAL HOME 520 - 520 - 520 - 480 - 480 - 480 -

440 - 440 - 440 -

400 - 400 - 400 -

360 - 360 - 360 -

320 -_ 320 - 320 -

Z 280 _ AFTERg , 280 AFTER 280 G AFTER : 240 240 - 240 -

; 200 - 200 - 200 -

160 -160-16 - 120 1201 - 120 -

80 8 . 80 40 - 40 _ 40 ... 0 30050 231 331 3.01 401 431 501 531601 631 331 4.01 431 501 531 601 631 331 401431 501 531 to 601 631 701 731 to to to to to to to to to to to to to to to to to to to to to to to to 3 00 3 30 4.00 4 30 5 00 5 30 6 00 6:30 7 00 4 00 4 30 5 00 5.30 6 00 6 30 7 00 4 00 4130 5 00 5 30 6 00 6 30 7 00 7 30 8 00

Trips from work in Restricted Zone to vehicle-owning households. Samplesizes Before 1170 After 1026 - 130 - was the peak half hour before the Area License Scheme) start earlier with the scheme in effect, while some have shifted to other modes. The fact that the dip also occurs for the arrival time and start-work time distributions increases the confidence that may be placed upon these data.

While there is clear evidence that people are starting their journeys and arriving at work earlier, the results on the times at which people start work are less clear. While there is some evidence of staggering, it does not appear to be as pronounced as the changes in the timing of trips.

Car Passengers. The trip starting time distributions for car passengers are paralleled in the arrival and work-starting time distributions. In each case, both the number of trips and the peak in their distribution have been diminished.

Bus Riders. As in the trip starting time distribution, the distributions for bus riders' arrival time and start-work time show the bus rider group (a) expanding and (b) increasing the peak in the distribution, thus more than compensating for the flattening of the distributions for car drivers and passengers.

TRIPS HOME FROM WORK IN THE RESTRICTED ZONE

The shapes of the distributions and the progressive change in degree of sharpness of the peak (either before or after) are mirror images of those observed in Figure 5.18 for the trip to work, and the explanation is essentially the same as before, in reverse. Nominal work stopping times are narrowly concentrated, some people leave immediately while others stay for varying periods before starting home, and the trip home takes different lengths of time for different people.

The distribution of the times when respondents arrived home can be compared with the distribution (presented in Figure 5.18) of the times when respondents started their journeys to work. The two "before" distributions are virtually mirror-images of each other; i.e., the times of arrival home from work are distributed in the same pattern (reversed) as the times of leaving home for work. The two "after" distributions, however, are different. It was reported above that, for trips to work, the "after" distribution shows (a) evidence of a small shift towards staggered hours, (b) a reduction in the number of car driver and car passenger trips, and (c) a distinct movement of the remaining car driver and passenger trips to earlier starting times. The "after" distribution for trips home from work shows only a reduction in the number of car driver and car passenger trips (with a corresponding increase in trips by bus). There is no evidence of a change in the times at which respondents by any mode arrive home. Thus, it may be concluded that the staggering of the times at which trips are started is a morning-only phenomenon. The other two distributions confirm the above observations. - 131 -

CHANGES IN TRAVEL TIME

In Chapter 5, it was reported that changes in travel time had been small. In this section, the evidence that led to that conclusion is presented.

First, an overview is presented, based on the macro-level data. This is followed by a detailed analysis of the micro-level data on individual trips. The results appear to be at variance with the estimates of travel speeds in Chapter 4, based on timing of moving vehicles. The questions raised by these differences will be analyzed in Chapter 6.

AN OVERVIEW

A global picture of changes in travel time is given by differences in the mean travel times derived from the macro-level "before" and "after" samples. For the users of different modes, the mean travel times for journeys to work in the Restricted Zone are:

Mean Reported - ravel Times (minutes)

(Home to Work Trips to Restricted Zone)

Vehicle-Owning Non-Vehicle-Owning Households Households

Car Car Bus Bus Drivers Passengers Riders Riders

Before ALS 28.8 27.9 41.1 41.2 With ALS 30.8 29.3 40.6 40.0

/a Specifically, the reported time of arrival at work minus the reported time of departure from home.

It should be noted that these figures are overall averages, i.e., the pre-ALS mean for car drivers includes all car drivers, irrespective of whether they changed their mode or starting time after the introduction of the scheme. Thus, there is no direct correspondence between the "before" and "after" samples.

In all cases, the changes in travel time are very small. Moreover, if an individual's ability to discriminate between slightly different times is limited, it is highly unlikely that changes of this magnitude are, in fact, perceived by travellers. However, the small differences in averages may come from larger differences in opposing directions for different sub- groups. This is explored in the micro-level analysis. - 132 -

Waiting and Walking Times

To test whether the Area License Scheme had affected the walking and waiting components of travel time, the following mean times were calculated:

Mean Time Spent (Minutes):

Walking at Walking at Waiting Origin Destination for Vehicle

(Before 5.8 5.0 8.4 BUS RIDERS After 5.6 4.8 6.7

(Before 1.2 2.1 n.a. CAR DRIVERS After 1.0 2.1 n.a.

Before 0.7 2.3 -- CAR PASSENGERS IAfter 1.3 2.1 -

Note: n.a. Not applicable. -- Not available.

The differences in mean times shown in the table are small, but some constitute large percentages of the base figure. It is open to debate whether a large percentage change on a small base, for example, from 0.7 to 1.3 minutes, is important either to the traveler or to society.

MICRO-LEVEL ANALYSIS

The micro-level sample of matched work-trips made to the Restricted Zone both before and after the introduction of the Area License Scheme permits a re-examination of the travel times for users of different modes before and after the Area License Scheme as well as a more detailed analysis of the changes in travel time associated with specific patterns of modal usage, e.g., a shift from car to bus. The mean travel times for bus riders and car drivers and passengers are shown in Table 5.12. As the macro-analysis showed strong similarities between bus riders from vehicle-owning and non-vehicle-owning households, only the former are analyzed here. The marginal totals from this data set confirm the findings from the macro-level data that those changes in mean travel time that have taken place are very small. However, the more interesting information is contained in the individual cells.

Certain changes are significant and expected: bus riders who changed to riding in private cars after the Area License Scheme experienced large reductions in mean travel time (except those who became car-pool passengers); conversely, those private car riders who changed to the bus experienced increases in mean travel time. For those who continued as bus - 133 -

Table 5.12: MEAN TRAVEL TIMES BEFORE AND AFTER ALS (Minutes)

POST-ALS MODE

Bus Car/a Car/a Car-Pool Car-Pool Rider Driver Pass. Driver Pass. TOTAL#

Bus Before 40.4 38.0 46.7 43.6 35.4 40.3 Rider After 41.0 29.0 20.0 29.3 36.4 - Sample (193) (5) (3) (7) (12) (229)

Car Before 30.2 26.8 29.6 31.4 30.0 28.2 Driver /a After 39.1 27.9 29.6 31.2 31.2 - Sample (37) (110) (11) (35) (8) (214)

Car Before 26.4 29.4 26.3 25.7 26.6 27.0 PRE- Passenger/a After 42.8 30.6 27.5 28.6 23.3 - Sample (38) (9) (28) (7) (4) (16) ALS Car-Pool Before 30.0 33.3 - 34.2 30.0 32.5 MODE Driver After 45.0 26.7 - 30.7 38.7 - Sample (1) (3) - (7) (4) (16)

Car-Pool Before 32.7 - 27.0 29.4 28.7 29.6 Passenger After 41.9 - 21.0 40.8 37.6 - Sample (11) - (5) (9) (19) (45)

Before - - - - - TOTAL# After 41.0 28.2 26.4 32.1 31.1 Sample (301) (129) (50) (67) (74)

/a Non-car-pool. # Includes users of "other" modes.

riders and who continued to use the private cars, in one form or another, the mean travel time changes were small. For car drivers who did not change mode, the average journey time increased by 1.1 minutes. This is a statistically significant difference, although it is debatable whether the difference is important in practical terms. For bus riders, the mean travel times are not significantly different from a statistical point of view. Be- cause a relatively constant mean travel time can conceal significant dispersion in individual observations, a more detailed analysis was carried out for two groups of travelers: car drivers (non-car-pool) and bus riders who did not change mode.

Car Drivers

The sample of respondents who drove cars (non-car-pool) both before and after the introduction of the Area License Scheme was screened - 134 -

to eliminate those respondents who had changed their destination within the Restricted Zone. It was believed that the inclusion of such observations would invalidate a detailed analysis of changes in travel time. In addition, two respondents who had changed their origins (residence) were eliminated. The changes in travel time for the remaining respondents are as follows:

Decreased by: No Change Increased by:

More than 5 mins. 1-5 mins. /a 1-5 mins. /a More than 5 mins.

17 20 27 16 19

37.37% 27.27% 35.35%

/a In fact, each observation in this category was exactly 5 minutes.

Given this pattern of changes, several factors were investigated to see whether any of them could be used to differentiate the trips with increased time from those with decreased time. These data are presented in Table 5.13.

Table 5.13: FACTORS POTENTIALLY INFLUENCING CHANGES IN TRAVEL TIME

Travel Time Changes

Percentage of each time Minus More Minus No Plus Plus More change category who: than 5 mins. 5 mins. Change 5 mins. than 5 mins.

used an ALS Sticker 43% 20% 35% 50% 23%

started Before 7:30 a.m. after ALS 59% 40% 22% 31% 32%

used part of the ring road 47% 60% 38% 65% 37%

It was thought that the necessity to use part of the ring road, which had become more congested, might have resulted in increased travel times. The data do not provide evidence that this has happened. To test whether specific sections of the ring road might have contributed to increased travel times, the routes of the respondents were examined. There is no evidence that the people who took longer used different (or more congested) sections of the ring road than people who took less time.

These two sets of people can be distinguished, however, on the basis of ALS-related factors. Area Licenses were used by twice as large a proportion of people who saved more than 5 minutes than of those who took - 135 -

more than 5 minutes longer. In addition, the starting time of the journey was before 7:30 a.m. (the start of the restricted hours) for twice as large a proportion of the time savers than of the time losers. Looked at from another point of view, 23 percent of the people who used the Area License Scheme stickers saved more than 5 minutes, whereas 14 percent of them had journeys that were more than 5 minutes longer. The results for people who changed their trip starting time from after 7:30 a.m. to before 7:30 a.m. are:

Difference in Travel Time Number Percent

Down by more than 5 minutes 7 35 Down by 5 minutes 3 15 No change 4 20 Up by 5 minutes 2 10 Up by more than 5 minutes 4 20

Total 20 100

Overall, 50 percent of people who made this change saved time, compared with 30 percent who lost time.

Additional analyses investigated whether the people saving more than 5 minutes could be distinguished from the people who took more than 5 minutes longer on the basis of their origin-destination patterns. The evidence is that neither the geographical distributions of origins nor the distributions of specific locations within the Restricted Zone are different for the different groups. The incidence of traffic "bottlenecks," such as Newton Circus, or the intersection of Whitley Road and Dunearn Road, was also examined. The groups cannot be discriminated on the basis of the frequency with which different groups travel through such areas.

The distribution of car drivers by income in the two major travel time change categories are:

Drivers for Whom Drivers for Whom Sample of Travel Time Travel Time Drivers as Income Group Increased Decreased a Whole

Low 27% 21% 20% Medium 24% 32% 42% High 49% 47% 38%

100% 100% 100%

These data indicate that people from the high income group have experienced more travel time changes -- both increases and reductions -- than members of the other income groups. The evidence is that no specific income group has either benefited or lost in terms of travel time changes. - 136 -

Bus Riders

A similar analysis was carried out for respondents who traveled into the Restricted Zone by bus both before and after the introduction of the Area License Scheme. After eliminating 19 respondents whose destination or origin had changed, the following detailed changes in travel time were observed for the remaining 174 respondents:

Decreased by: No Change Increased by: More than 5 mins. 1-5 mins. 1-5 mins. More than 5 mins.

36 21 37 27 53 21% 12% 16% 30% 33% 21% 46%

A detailed analysis of the distributions of origins and destinations, the amount of time on the ring road, corridors along which bus riders entered the central area and the trip starting times showed that the people who saved time could not be distinguished from those who lost time on the basis of these criteria. In addition, the proportions of time savers and time losers from each income group were very similar.

SUMMARY -- TRAVEL TIMES

According to the travel time data reported in the household survey, the Area License Scheme had negligible effects on travel times, except for those travelers who changed modes. At a micro-level, however, the incidence of large positive and negative changes in travel time is rather disconcerting. It is possible, but seems unlikely, that these changes form part of a normal and random pattern of changes in travel times. In the case of car drivers, there is some evidence that responding to the scheme by either purchasing a license or starting the trip before 7:30 a.m. leads to an increased likelihood of reducing travel time. No evidence was found to explain the changes experienced by bus riders. It is possible that the convoys of buses observed in the Restricted Zone after the introduction of the Area License Scheme resulted in increased time spent by buses at bus stops, which could have offset the time saved by reduced congestion in the Restricted Zone.

VARIATIONS BY INCOME GROUP

The objective of this section is to examine the extent to which the changes described above have had differential impacts on different income groups. The most frequently expressed concern regarding the distributive effects of road pricing schemes is that lower-income car-driving commuters will be unfairly discriminated against. It may be harder for them than for high-income people to change their working hours in order to reschedule their trips, and the license fee, if they do pay it, represents a larger proportion of their incomes. It was expected that car drivers in the lowest income group - 137 -

entering the Restricted Zone between 7:30-10:15 a.m. would adopt a different mode of travel -- at least recruiting passengers to take advantage of the car-pool exemption. This section concentrates on effects on work trips into the Restricted Zone by members of vehicle-owning households at different income levels. For this purpose, travelers were divided into three groups on basis of the reported household income. 1/ The categories are (a) low income (less than S$1,000 per month), (b) medium income (between S$1,000 and S$2,000 per month), and (c) high income (over S$2,000 per month). CHANGES IN MODE

For each of the three income groups, Table 5.14 shows the extent to which the users of any mode before the Area License Scheme used other modes after the scheme was implemented. The data are summarized below in terms of the bus shares. As expected, the proportion of people that ride the bus is smaller in the high income group:

Income BUS SHARE Group Before ALS After ALS % Change

Low 44% 55% +25% Medium 41% 55% +34% High 29% 37% +28%

However, the increases in the bus shares are fairly uniform between 25 and 35 percent. It is also interesting (as Table 5.14 shows) that the practice of shared riding is much less common in the low income group.

At a more detailed level, some small, but interesting, differences can be detected.

Car Drivers

It is surprising that a smaller percentage of car drivers from the lower income group changed mode than from the two higher income groups. In fact, this seems to represent a smaller shift to car pools and car passengers. As was to be expected the percentage of car drivers shifting to the bus was lower in the highest income group than in the other two. Car Passengers

For car passengers, the major shifts were to car pool and bus. For the lower two income groups the larger shift was to the bus; for the upper group, the larger shift was to the car pool.

1/ The specific question was "Would you please show me on this card how much the people in your household earn all together, on average, in a month?" - 138 -

Table 5.14: MODE CHANGES BY INCOME GROUP FOR TRIPS TO WORK IN THE RESTRICTED ZONE Part 1 of 3: Household income less than S$ 1,000 per month

Mode After ALS

Mode before Car /a Car /a Car- TOTAL ALS Bus driver passenger pool BEFORE

50 (93%) 2 (4%) 0 (0% 2 (4%) 54 (100%) Bus (75%) (5%) (0%) (15%) (44%)

Car la 12 (24%) 33 (65%) 0 (0% 6 (12%) 51 (100%) driver (18%) (87%) (0%) (46%) (42%)

Car /a 3 (27%) 2 (18%) 4 (36% 2 (18% 11 (100%) passenger (4%) (5%) 100%) (15%) (9%)

Car- 2 (33%) 1 (17%) 0 (0% 3 (50% 6 (100% Pool (3%) (3%) (0%) (23%) (5%)

TOTAL 67 (55%) 38 (31%) 4 (3% 13 (11%)122 (100% AFTER 100%) 100%) 100%) Q.00%) 00%)

(vehicle-owning households) {a Non-car-pool.

Key: No. (Row %) (Col. %) - 139 -

Table 5.14: MODE CHANGES BY INCOME GROUP FOR TRIPS TO WORK IN THE RESTRICTED ZONE Part 2 of 3: Household income between S$ 1,000 and S$ 2,000 per month

Mode after ALS

Mode before Carla Car /a Car- TOTAL ALS Bus driver passenger pool BEFORE

97 (93%) 0 (0%) 0 (0% 8 .(8% 105(100%' Bus (69%) (0%) (0%) (14%) (41%)

Car /a 17 (23%) 37 (50%) 5 (7% 15 (20X 74(100% driver (12%) (88%) (29%) (26%) (29%)

Car /a 20 (41%) 3 (6%) 11 (2a% 15 (31% 49(100% passenger (14%) (7%) (65%) (26%) (19%)

Car- 7 (23%) 2 (7%) 1 (3% 20 (67%O 30(100% Pool (5%) (5%) (6%) (34%/.) (12%)

TOTAL 141 (55%) 42 (16% 17 (7% 58 (22% 258(100% AFTER (100%) (100%) (100%) 100%) (100%)

/a Non-car-pool. (vehicle-owning households)

Key: No. (Row %) (Col. %) - 140 -

Table 5.14: MODE CHANGES BY INCOME GROUP FOR TRIPS TO WORK IN THE RESTRICTED ZONE. Part 3 of 3: Household income more than S$ 2,000 per month

Mode after ALS

Mode before Car La Car La Car- TOTAL ALS Bus driver passenger. pool BEFORE

55 (80%) 3 (4%) 2 (3% 9 (13%' 69(100% Bus (63%) (5%) (7%) (14%) (29%)

Car /a 13 (14%) 51 (57%) 6 (7% 20 (22% 90(100% driver (15%) (85%) (22%) (32%) (38%)

Car /a 14 (27%) 6 (12%) 14 (27% 18 (35% 52(100% Vassenger- (16%) (10%) (52%) (29%) (22%)

Car- 6 (22%) 0 (0%) 5 (19% 16 (59%) 27 (100% pool (7%) (0%) (19%) (25%) (11%)

TOTAL 88 (37%) 60 (25%) 27 (11% 63 (26% 238(100%, AFTER (100%) (100%) (100%) (100%) (100%)

/a Non-car-pool. (vehicle-owning households)

Key: No. (Row %) (Col. %) - 141 -

Car-Pool Riders

In general, the car-pool mode shows less change than the other car modes. No important differences between income groups are revealed.

Bus Riders

The percentage of people who remained bus riders exceeds 90 percent, except for the highest income group where more movement to auto modes, especially car pool, is observed. It is likely that this represents greater access to cars for people in this income group.

Overall, it is concluded that there were only small differences among income groups in modal response to the Area License Scheme.

CHANGES IN TIMING OF TRIPS

The proportions of travelers starting their trips before 7:30 a.m. are shown below.

Income Proportion Starting Trip Before 7:30 a.m. Group Before ALS After ALS % Change

Low 40% 45% 13% Medium 32% 35% 9% High 22% 29% 31%

As might be expected, the proportion starting before 7:30 a.m. declined with increasing income both before and after the Area License Scheme. It is interesting, however, that the largest change in this proportion occurred in the high income group. This can be seen more clearly when the data for car drivers are examined:

Income Car Drivers Starting Trip Before 7:30 a.m. Group Before ALS After ALS % Change

Low 42% 57% +36% Medium 28% 39% +39% High 20% 41% +105%

However, since these proportions reflect changes in both the number of drivers starting before 7:30 a.m. and the total number of car drivers, a more precise picture can be obtained by examining the data on those car drivers who formerly started their trips after 7:30 a.m. and changed to starting before 7:30 a.m. after the introduction of the Area License Scheme:

Income % of Car Drivers Who Changed to a Group Pre-7:30 a.m. Starting Time

Low 33.3% Medium 25.0% High 41.7%

Overall, it is clear that the largest adjustment in terms of trip starting time was made by car drivers from the high income group. - 142 -

For bus riders, the absolute number starting before 7:30 a.m. remained virtually constant, but the increase in total bus ridership resulted in a fall in the proportion of bus riders from each income group who started their trips before 7:30 a.m.

CHANGES IN TRAVEL TIME

It was noted in an earlier section that the data do not provide any evidence to indicate that the members of any specific income group saved or lost more time than the members of other groups.

SUMMARY -- VARIATIONS BY INCOME

It had been feared that the Area License Scheme would have more unfavorable effects on poor people than on rich people. In terms of having to change mode, car drivers from the low income group changed least, while drivers from the high income group changed their starting times most. Sav- ings and increases in travel time were experienced equally by the different income groups. Thus, it does not appear that low income people have suffered more than high income people as a result of the Area License Scheme.

NON-VEHICLE-OWNING HOUSEHOLDS:

In Chapter 5, the results for non-vehicle-owning households were generally presented in summary form. The complete data on trips to and from work in the Restricted Zone are shown in Figures 5.20, 5.21 and 5.22.

TRIPS TO AND FROM WORK BEYOND THE RESTRICTED ZONE

The number of people in the subsample of non-vehicle-owning households who made trans-Restricted-Zone work trips is rather small. Neverthe- less, certain distinct travel behavior patterns can be distinguished. The modal splits are shown in Figures 5.23 and 5.24. For journeys to and from work, the bus mode predominated both before and after the introduction of the Area License Scheme. The apparent shifts in other modes are not actually significant because the samples are so small.

No change could be detected in the times at which journeys to and from work were made, and no route choice analysis was peformed as bus routes remained essentially the same before and after the introduction of the traffic restraint scheme.

In summary, the data reveal no evidence that trans-Restricted-Zone commuters from non-vehicle-owning households modified their travel behavior in response to the Area License Scheme. - 143 -

Figure S.20 Mode use for trips to work in the Restricted Zone from non-vehicle-owning households.

BEFORE ALS

Bus (88%)

Taxi (3%) Walk (-2%) Bicycle (5%) Car Passenger (1%)

AFTER ALS

| ~~~~~~~~Bus(87%)

Shutile Bus (2%) Walk (2%) \ / / / Bicycle (4%)

Car Passenger (3%) Car Pool (2%)

Note Sample sizes efore ALS, 212 trips. flter ALS, 193 trips - 144 -

Figure 5.21 Mode use for trips home from work in the Restricted Zone to non-vehicle-owning households.

BEFORE ALS

Bus (86%)

Taxi (2%)

Walk (3%) , /v Bicycle (5%) Car Passenger (2%)

AFTER ALS

Bus (88%)

Shuttle Bus (2%) - Walk (3%Yo) ~ < /

Bicycle (4%)

Car Passenger (3%) -

Note Sample sizes before ALS. 190 trips, after ALS. 185 trips - 145 -

Figure 5.22 Starting times for work trips to the Restricted Zone from non-vehicle-owning households.

50 _

Key- 40 _Before ALS After ALS

30 -

2 20 -

10 _

6.01 6.31 7 01 7:31 8.01 8:31 9:01 9131 10:01 10 31 to to to to to to to to to to 6:30 7:00 7:30 8.00 8 30 9:00 9:30 10:00 10130 11.00 Time period (a.m.)

30 -

25-

20 -

0 15F

6 01 6 31 7 01 7 31 8 01 8 31 9 01 9 31 10 01 1031 to to to to to to to to to to 6 30 7.00 7:30 8 00 8.30 9:00 9 30 10 00 10.30 11 00 Time period (a.m )

J Sample 191 trps before, 177 trips after. - 146 -

Figure 5.23 Mode use for work trips beyond the Restricted Zone from non-vehicle-owning households.

BEFORE ALS

Bus (90%)

Bicycle (3%)

AFTER ALS

Bus (85%)

Other (3%) icycle

Car Pool (3%)

Note Sample sizes before ALS. 32 trips. alter ALS 33 trips - 147 -

Figure 5.24 Mode usefor trips home to non-vehicle-owning households from work across the Restricted Zone.

BEFORE ALS

Bus (94%)

Other (3%)

Bicycle (3%)

AFTER ALS

Bus (85%)

Other (3%) icycle (9X)

Car Pool (3%)

Note. Sample sizes before ALS, 32 trips. after ALS. 33 trips - 148 -

CHAPTER 6. FURTHER ANALYSIS OF TRAVEL CHANGES

The two previous chapters set out in some detail the observed traffic performance and reported travel behavior before and after institution of the Area License Scheme. Traffic performance was reported for various types of vehicles, and travel behavior for various trip purposes, origin- destination patterns, and types of household, whether they were subject to tha direct impact of the scheme or not. In this chapter, the analysis is carrie4 further for certain key questions with special regard to the most significant changes observed. Thus, the focus is primarily on automobile travel into and out of the Restricted Zone, including changes in the volumes of cars, operating speeds (of cars and other vehicles), and the response of drivers to the change in cost.

Table 5.7 summarized the principal direct impacts of the scheme, which fell upon people who made work trips crossing the future Restricted Zone boundary between 7:30 a.m. and 10:15 a.m. People making these trips by car numbered about 56,000, traveling in about 40,000 cars. The changes summarized in Table 5.7 raise a series of questions that deserve to be analyzed in greater detail. The changes in travel behavior, that is in trip rates, mode, and trip scheduling, should serve to explain the traffic flow data, which showed a large reduction in the number of cars entering the Restricted Zone during the restricted hours. It is useful to juxtapose the two ways of looking at the same phenomenon and ask what happened to all the cars that stopped driving into the Restricted Zone during the restricted hours. In the same context, it is appropriate to ask also why the Area License Scheme failed to produce the expected mirror-image reduction in congestion during the evening peak.

Similarly, the data on reported travel times and observed speeds represent different ways of looking at the same thing, and it is useful to examine the data together and to ask what has been learned from the attempts to measure changes in speeds and travel times.

The data on license sales at each of two prices reflect a behavioral response -- how motorists react to large price increases -- about which little is known. Thus, it is important to examine this response and to ask what can be said about the level at which the Area License fee was set.

The results of attempts to answer these questions are presented below.

CAR TRIPS TO THE RESTRICTED ZONE

The preceding chapters presented a wide range of information on different aspects of the changes in travel behavior and traffic that resulted from the introduction of the Area License Scheme. This section focuses on the question: What happened to the cars that drove into the central area before - 149 - the scheme was introduced? The starting point for the analysis is the March 1975 count of 42,790 cars that entered what was to become the Restricted Zone during the restricted hours. There is no direct way of identifying which of these cars were going to destinations within the zone and which were going through to points beyond. However, by applying factors derived from the evening survey of cross-town traffic it was estimated that 16,160 of these cars went through while 26,630 ended their trips within the zone. Then, using ratios from the household survey data, these trips were distributed over all the alternatives to create a global picture of changes in travel behavior. This picture allows both the relative magnitudes of different changes and the consistency between the household survey data and the traffic flow data to be examined. The results of the exercise are presented in Figure 6.1.

Consistency with the traffic flow data can first be checked in terms of the total number of cars entering the Restricted Zone during restricted hours and in terms of the proportion of car pools included in that total. The box at the bottom of Figure 6.1 shows the number of cars calculated as entering the zone during the restricted period after accounting (on the basis of proportions from the household surveys) for all of the drivers who chose other options. The total, for trips passing through the Restricted Zone and those going to destinations within it for both work and other purposes, is 14,018, representing a reduction of 67 percent from the pre-license-scheme count of 42,790. The number of cars actually counted entering the zone during these hours was 11,363 -- a 73 percent reduction. 1/ While this is not precise agreement, the reduction in the counted flow and that calculated from interview responses differ by only about ten percent.

The proportion of car pools in the total before the Area License Scheme was instituted was 7.2 percent according to the interviews and 6.6 percent from counts. With the scheme in effect, car pools entering between 7:30 a.m. and 10:15 a.m. constituted 30 percent of the total, according to the interviews, and 37 percent by observation.

In terms of the relative magnitudes of different changes, several features are worth noting. For trans-Restricted-Zone travelers, the most popular reaction was to change route and detour around the zone. Changing mode and changing time were also popular, but together they accounted for less than two-thirds as many cars as the route change. For car drivers making trips to the Restricted Zone, there was a large reduction in the number of non-work trips. Over 80 percent were either abandoned, made by a non-driving mode, or changed to another time. Of the 25,600 people who before ALS drove to work in the Restricted Zone between 7:30 a.m. and 10:15 a.m., 3,300 stopped making work trips to the Restricted Zone. 10,800 continued driving during the restricted hours, 30 percent of them in car pools. Of the rest, 4,300 changed time (most of them to an earlier time), 4,209 changed to riding the bus, and 2,896 changed to other modes, mainly riding as passengers in cars (pool or otherwise).

1/ All figures mentioned as flow counts in this section are from the tables in Chapter 4. - 150 -

It is appropriate to note that four important options -- car pooling, changing time, riding the bus, and becoming car passengers -- were utilized more or less equally. Had one alternative been very much more popular than the others, it might have resulted in some part of the transport system being seriously overloaded. For example, if most of the people had changed to the bus, severe overcrowding on buses would have resulted. The fact that different people chose different options prevented such an outcome. This result implies that it is important for planners designing area license schemes to provide a variety of alternatives.

THE EVENING PEAK

In Chapter 4 it was pointed out that -- contrary to expectations -- the reduction in the morning inbound traffic flow induced by the Area License Scheme was not noticeably reflected in the evening outbound flows. While the available before-and-after flow data give only a partial picture of the situation, there is a clear perception among planners and observers that congestion still exists in the evening peak. In this section, an attempt is made to explain the continued congestion with the help of household interview data collected before and after initiation of the Area License Scheme.

Table 6.1 shows data from the household surveys on car trips into and out of the Restricted Zone for all purposes before and after the introduction of the Area License Scheme.

Table 6.1: TIMING OF CAR TRIPS TO AND FROM RESTRICTED ZONE

Number of Car-Driver Trips To Restricted Zone From Restricted Zone Before After Before After

Before 7:30 a.m. 83 82 8 3 7:30-11:00 a.m. 378 219 21 6 11:00 a.m.-4:00 p.m. 37 17 59 38 4:00-7:00 p.m. 29 11 361 247 After 7:00 p.m. 24 10 80 45

TOTAL 551 339 529 339

WORK TRIPS ONLY 406 274 395 268

Purpose: All Purposes. O/D Pattern: To and From Restricted Zone. Subsample: Vehicle-Owning Households. Sample Size: 619 Households. - 151 -

There is a reasonably close balance between the inbound and the outbound trips, both before and after the introduction of the Area License Scheme, with the major inflow between 7:30 a.m. and 11:00 a.m. and the major outflow between 4:00 p.m. and 7:00 p.m.

The effect of the license scheme is clearly apparent in the 42 percent drop in inflow between 7:30 a.m. 11:00 a.m. (The inclusion of the period just after the end of the restricted hours makes this a smaller percentage reduction than that observed within the restricted period alone.) Also apparent is the expected mirror image in terms of a drop in outflow between 4:00 p.m. and 7:00 p.m. While the latter drop is not as pronounced as the former, it is enough, at 32 percent, so that it should have been clearly noticeable even to casual observers if it had applied to all outbound traffic flows in that period.

The fact that no appreciable reduction in evening traffic was apparent or was recorded by the mechanical counters indicates that the reduction in evening trips home from places within the Restricted Zone must have been largely offset by increases in other traffic components. Several hypotheses that have been advanced as possible explanations are examined below.

COMBINED TRIPS

One hypothesis is that people who travel to work by bus or as passengers in car pools are picked up by members of their families or friends and travel home as car passengers or drivers. It has also been hypothesized that more shopping and personal business trips are made in the late afternoon so that another family member may be taken home from work.

Although the samples for shopping and personal business trips are small, there is limited evidence to indicate that shopping trips have been rescheduled into the evening peak period. However, the micro-level household survey data for trips to work and home again (shown in Table 6.2) do not support the hypothesis that this has been done to pick up people who traveled to work by other modes. After the introduction of the Area License Scheme, 85 percent of trips from work to home used the same mode that was used to get to work. The four car-using categories constituted a smaller -- not larger -- proportion of homebound trips than of trips to work. Seventeen percent of people who were car-pool passengers in the morning went home as non-car-pool passengers (none went home as car drivers). Only three percent of people who went to work by bus went home as non-car-pool passengers. As the non-car-pool group includes people who remained in the same car but lost one fellow- passenger, and as the shifts were the same in the pre-ALS sample, the data do not support the hypothesis that mode changes were made for the journey to work but not for the journey home. - 152 -

Figure 6.1 Changes in morning peak car trips.

Car Drivers Entering Future Restricted Zone 7 30-10.15 A.M, March 1975 42,790 (3,081 Car Pool, ., ,v . ~~39,709 Non-Car Pool)

_ , ~~~~~~~~~~~~~25,587 Through Destinations in Work Trips Restricted Zone Restricted Zone 0 -I~~~~~~~~~~~~~~~~~~~I16,160 26,630 \ See next page)

O ~~ ~~All Social and Personal ui Purposes Sh6pngRecreation Bu349 ------LnI---,Trip Rate Tri Ra|t;;e > | Down 21% 11Down 40% < 12,766 210

766 5 Changed to Changed to Non-Driving Mode' J Non-Driving Mode'

3,000 1a e Changed Time Cag

l ~~~Changed 4 ~~~~Time l 6,000 0 -O Detoured

Through Trips 3,000 Personal 195 (Car Pool 900 (Car Pool 58 Non-Pool 2,100) Non-Pool 137) Car Drivers Entering Restricted Zone 7 30-10 15 A M . Post A LS - 153 -

Figure 6.1 Continued

Work Trips by Car _Drivers into Future ,Restricted Zone, (See previous page) 7 30-10. 15 A.M, March 1975 25,587X

------___- -- ______- Lu

Work Trip Rate e Down 13% 22.260 <:

4,209 Changed to Bus 4,332 Drove at i Non-RestCDcted Hours 30 A P A To Other Non-Driving Modest

(Car Pool 3,247 (Car Pool 4,206 Non-Pool 7,576) Non-Pool 9.812) Car Drivers Entering Restricted Zone 7 30-10.15 A M, Post A L.S - 154 -

Table 6.2: PERCENTAGE MODAL USAGE - A.M. AND P.M. POST-ALS Total, Work to Home (P.M.) Home Car Car Car-pool Car-pool to Percent Bus Driver/a Pass./a Driver Pass. Other Work

Bus 44 0 1 0 1 0 46 Car Driver/a 1 17 0 0 0 0 18 Home to Car Work Passenger/a 1 0 5 0 0 1 7 Car-pool (A.M.) Driver 1 3 1 5 0 0 10 Car-pool Passenger 2 0 2 0 7 0 11 Other 1 0 0 0 0 7 8

Total, Work to Home 50 20 9 5 8 8

/a Non-car pool. Purpose: Home to Work and Work to Home. O/D Pattern: To and From Restricted Zone. Subsample: Micro Level Sample. Sample Size: 640.

CHANGES IN PARKING PLACES

It has also been suggested that some travelers park just outside the Restricted Zone and continue their journeys on foot or by bus, but return later in the day to bring the cars into the Restricted Zone in order to have them available for the journey home. The level of detail at which data have been analyzed do not reveal such behavior. It seems unlikely that this is an important part of the explanation of the continued evening congestion.

TRANS-RESTRICTED-ZONE TRIPS

Another hypothesis is that the evening congestion is largely due to people making trans-Restricted-Zone trips, who use the ring road to bypass the zone going to work in the morning but travel through the zone going home in the evening.

The cross-town flow survey reported in Chapter 4 yielded an estimate of 6,357 cars entering eight selected points on one side of the Restricted Zone between 4:30 p.m. and 6:30 p.m. and leaving the zone on the other side. Applying the observed time profile of total traffic, it is estimated that the number between 4:00 p.m. and 7:00 p.m. would amount to 9,000 cars, and an - 155 -

additional factor to account for the entry and exit points that were not monitored brings the estimated total of through trips in the three-hour period up to 12,000.

Referring to Figure 6.1, it is seen that through trips during the morning restricted period were estimated at 3,000, that an additional 3,000 were estimated to have shifted to an earlier or later time to avoid the restricted hours, and that 6,000 drivers who had driven through the Restricted Zone before the Area License Scheme detoured to by-pass it. If all members of these last two groups chose to return through the zone between 4:00 p.m. and 7:00 p.m., the total through traffic in those hours would equal the 12,000 cars estimated from the cross-town count, and this would include 9,000 cars that did not enter the zone during the morning restricted period. Since that assumption is an extreme one, the number is probably somewhat less than 9,000.

STAGGERED MORNING HOURS

It is believed that some of the people who adjusted the time of their trips to work to avoid the Area License Scheme restrictions may not have been able to adjust their working hours. Thus, they may travel to work early but leave at the same time as before. As noted in Chapter 5, there is a large movement towards earlier starting times in the morning. There is no evidence of earlier starting times for the journey home in the evening. In fact, there is a small tendency towards later starting times. Additional evidence is presented in Table 6.3, which shows the proportions of travelers making trips, starting work, and finishing work at different times. The proportion of people starting their journeys to work before 7:30 a.m. rose from 26 percent to 40 percent; the proportion arriving at their destinations before 8:00 a.m. also rose -- from 26 percent to 36 percent. The proportion starting work before 8:00 a.m. rose from 25 percent to 33 percent; much of this increase apparently came out of the 8:00-8:30 a.m. period. This is clear evidence that an important number of people have moved toward earlier trip starting times, but about half of them have not actually started work earlier. At the other end of the working day, the picture is quite clear: The pattern's of times at which people finished work and left for home have virtually not changed.

The above evidence from interview data are corroborated by the Public Works Department's counts of vehicle flows. Figure 6.2 shows the time profiles of flows of cars and of all vehicles, inbound in the morning and outbound in the evening. The effect of the ALS has been to create two separate but reduced peaks in the morning. The traffic that makes up these peaks is amalgamated into a single evening peak. Moreover, the total number of cars going outward in the evening is substantially higher than the number going in the morning. The difference of 7,629 cars is within the round-number upper limit of 9,000, estimated above for trans-Restricted-Zone trips that go around the zone in the morning and through it in the evening, and are, hence, included in the outbound data but not in the inbound data. - 156 -

Table 6.3: PROPORTIONS OF CAR DRIVERS STARTING AND FINISHING WORK, AND MAKING WORK TRIPS AT VARIOUS TIMES (percent)

Start Trip to Arrive at Start Work Destination Work Morning Before After Before After Before After

6:30-7:00 10 21 1 3 1 2

7:00-7:30 16 19 7 19 4 7

7:30-8:00 30 16 18 14 20 24

8:00-8:30 25 18 26 20 24 20

8:30-9:00 10 9 23 21 35 26

9:00-9:30 5 3 9 4 6 4

9:30-10:00 4 5 5 3 4 4

10:00-10:30 1 4 2 4 1 4

Other times 1 6 9 12 5 9

Finish Work Start Trip Home Evening Before After Before After

3:30-4:00 2 4 3 3

4:00-4:30 9 9 7 8

4:30-5:00 47 42 32 32

5:00-5:30 11 12 18 19

5:30-6:00 8 10 12 9

6:00-6:30 6 3 8 5

6:30-7:00 3 4 3 4

Other times 14 16 17 20

Purpose: Home to Work and Home from Work. O/D Pattern: To and From Restricted Zone. Subsample: Car Drivers from Vehicle-Owning Households. Sample Sizes: Before = 395; After = 268.

Note: Figures may not add to 100 percent due to rounding. Figure 6.2 Vehicle flows into and out of the Restricted Zone.

INBOUND OUTBOUND 700- 11.00am. 30,000 400-700 p.m. All Vehicles 72,461 All Vehicles 73,991 Cars 28,423 A Cars 36,052

25,000 Vehicles | }Cars

<, 20,000

15,000

5,000 4 :_Ll 1 ' ] l No Data

7 8 9 10 1(1 12 T 1 2 3 4 6 7 (a m.) Time (pm) - 158 -

TAXI TRIPS

It was reported in Chapter 4 that the number of taxis operating in the Restricted Zone fell dramatically after the introduction of the Area License Scheme. The traffic flow data (for September/October 1975) indicate that just over 7,000 taxi trips into the Restricted Zone have been suppressed. (Of course, this includes repeat trips by the same taxis and trips by empty taxis.) This left about 8,000 taxi trips into the Restricted Zone during the morning peak (7:00-11:00 a.m.), which may be compared with about 12,000 taxi trips observed leaving the Restricted Zone in September 1975 during the evening peak (4:00-7:00 p.m.). While the data do not permit a detailed analysis, it is clear that the renewed taxi activity in the evening peak is an important factor contributing to continued evening congestion.

SUMMARY -- THE EVENING PEAK

It is unfortunate that the data do not permit before-and-after comparison of the evening traffic flows or the detailed accounting of inbound and outbound trips that would be required for a precise understanding of the causes of the continued evening congestion. Nonetheless, the existing data do reveal some helpful insights. There is substantial evidence that the continued evening congestion is the result of three main factors: trans- Restricted-Zone trips that bypass the zone in the morning, but travel through it in the evening, trips by drivers who stagger their travel times in the morning but not in the evening, and taxi trips.

The first category comprises between 7,600 and 9,000 trips. The second has an upper bound of about 6,000 trips (the number that shifted to a pre-restricted-hours starting time). The third comprises about 4,000 trips. These numbers indicate in a crude fashion the relative magnitudes of the contributions of the different factors to the continued evening congestion.

TRAVEL TIME AND SPEED MEASUREMENTS

Chapter 5 included travel times for trips to work calculated from the starting and arrival times reported in the household interviews. Rather large changes in time in both directions appeared after the introduction of the Area License Scheme. The increases and decreases largely offset each other for trip-makers who did not change mode, yielding relatively small changes in the mean values. For example, the reported travel time increased by more than five minutes for 19 percent of car drivers, and went down by more than five minutes for 17 percent of drivers, while the mean trip time for this group increased by 1.1 minutes. Similar results were found for other travelers who did not change mode. - 159 -

In Chapter 4, average speeds were presented from moving-car observations on various roads at different times of day. From these observations -- all made after the Area License Scheme was introduced -- estimates were made of the mean speeds before and after the Area License Scheme on roads with different orientations and locations (Table 4.18). The differences in speeds did not appear large enough to account for any very large savings or losses in travel times.

In this section two main questions are addressed, as follows:

1. Are the data from the interviews and from vehicle observations consistent?

2. Do they provide a suitable basis for evaluating benefits and losses to different groups in the form of time savings and losses?

To investigate these questions, the mean speeds of Table 4.18 were used to calculate "before" and "after" travel times for a subsample of eight specific car trips corresponding to those reported in individual interviews. In one case the calculated time both before and after were close to those indicated by the reported starting and arrival times, and the before-and- after change was calculated as a 2.2 minute saving compared with a 5 minute saving based on the interview data. Given that times were almost universally reported in five-minute steps, this agreement is as close as could be hoped for in an individual case. In one other case, the calculated before-and-after difference was close, in this sense, to the difference from interview data -- 2.7 minutes compared with 5 minutes -- but the absolute magnitudes were widely different. In the other six cases there was no resemblance between times based on the interviews and those calculated from the mean speeds of Table 4.18. The sign of the difference was frequently reversed, and the magnitudes based on the interviews ranged from a fifteen minute saving to a twenty minute loss, while the calculated figures ranged from a 2.7 minute saving to a 2.3 minute loss.

The lack of agreement between the results of the two approaches can be attributed to several factors concerning each of them. The time savings or losses are, for most travelers, on the order of only a few minutes, and neither method is suitable for measuring such differences accurately.

For the interview approach, part of the problem stems from the fact that people are not aware of precise times; they report time, at best, rounded to the nearest five or ten minutes, and in many cases to the nearest quarter hour or even half hour.

While these roundings would not affect the mean values if the actual distribution were symmetrical and the analysis covered several hours, they would, even in that case, increase the variance of the differences between starting and arrival times. If the actual distribution is skewed, then the rounding may introduce a bias as well as increase the variance. The important point here is that interviews, while good for many purposes, do - 160 - not provide accurate enough estimates of travel time to justify using the difference between before and after estimates as a measure of time savings or losses that are only a few minutes in magnitude.

The other method, using vehicle speed observations, also has limitations, especially in the way it was used in this study. Because of the technical difficulties in the early part of the study, data on speeds before the Area License Scheme were not available, and estimates thereof had to be based on observations afterward combined with assumptions, as explained in Chapter 4. Because of the uncertainties in this approach, it was judged insufficiently precise to permit estimating before and after speeds on individual road links. Instead, only averages for certain categories of road were used, and these averages conceal a considerable degree of variation. Had it been possible to estimate speeds before and after on specific links, the calculated differences for individual trips would undoubtedly have been somewhat more widely dispersed. In fact, the use of average speeds has clearly resulted in smaller time differences than actually exist. In short, it must be concluded that the available speed measurements are not sensitive enough to give a reliable measure of travel time changes on the order of a few minutes.

If neither of these approaches is accurate enough to measure the changes in question, should some other method have been used? Apart from the fact that better results would have been achieved if the initial "before" observations of vehicle speeds and flows had been carried out successfully, there are no other methods that could have been used that were judged practical from a cost-effectiveness standpoint. Moreover, the judgment was made that if time savings (or losses) were too small to measure accurately by the methods used, they would not be perceived by the beneficiaries (or losers) and would not be important in the overall evaluation. Overall, it must be concluded that, since changes in travel time for any particular mode were too small to be measured accurately, the data do not constitute an adequate base for a meaningful evaluation of benefits or losses in the form of changes in travel time except for those people who made changes in their modes of travel.

MOTORISTS' RESPONSES TO PRICE INCREASES

Before the introduction of the Area License Scheme in Singapore, the World Bank convened a panel of experts to examine and discuss the available evidence on responses of both private and public transport users to price changes. The panel concluded that, while some scattered evidence existed on responses to small changes in bus fares and parking charges, there was nothing to indicate how motorists in Singapore might react to a package made up of (a) a completely new type of charge at a high level and (b) a relatively large increase in parking charges. Thus, the Singapore government found itself in the situation of having to set the license fee and parking rates without the benefit of previous experience. - 161 -

The analysis of the demand changes stimulated by the introduction of the Area License Scheme is in two parts: the first is limited to the change in the volume of license sales that resulted from the increase in the area license fee that took place on December 31, 1975; the second examines changes in the demand for trips by car to the Restricted Zone.

THE DEMAND FOR AREA LICENSES

Evidence on demand elasticities can be derived from the change in area licenses sales that occurred after the increase in area license fees that took place on December 31, 1975. 1/ Price elasticities of demand for area licenses have been calculated for both daily and monthly licenses, and separately for private cars, company cars and taxis (Table 6.4).

The most interesting feature of these results is the difference between the elasticities for private cars, company cars, and taxis. The fact that, overall, the private car elasticity is twice the company car elasticity reflects the fact that the incidence of a price increase falls more directly on a private owner who has to pay it out of his after-tax income than on a company user for whom it is tax deductible. The very high elasticity for taxis underlines the doubts of taxi drivers that it would be worth the cost to purchase a license so that they could operate in the Restricted Zone.

Table 6.4: PRICE ELASTICITIES OF DEMAND FOR AREA LICENSES

Monthly Daily Total

Private Cars 0.39 0.79 0.48 Company Cars 0.26 0.36 0.27 Taxis 1.43 0.98 1.20

Total 0.31 0.61 0.35

Notes: 1. The changes in the quantities of licenses sold were based on (a) the average of the three months from October to December 1975 (the months for which private and company car licenses could be distinguished) and (b) the average from February 1976 (the first stable month after the price increase) to November, 1976.

2. The "total monthly", "total daily," and "grand total" elasticities were computed on the basis of an average percentage price increase weighted by the volumes of private, company car, and taxi licenses sold in October-December, 1975.

3. The elasticity was calculated as the percentage change in the number of licenses sold divided by the percentage change in the price of an area license. No other trip costs were considered.

1/ It cannot be done for the introduction of the Area License Scheme, as that change constitutes an infinite increase in price. - 162 -

THE DEMAND FOR CAR TRIPS TO THE RESTRICTED ZONE

The approach taken in the previous section has two drawbacks: first, it does not permit an examination of motorists' responses to the introduction of the Area License Scheme; second, it does not reflect the idea that motorists take into account more than just the Area License fee, which is only a part of the perceived cost of a trip. Thus, the analysis in this section develops an approach in which the number of car trips into the Restricted Zone is analyzed as a function of the perceived cost of the trip.

This analysis is based on the following assumptions:

1. the measure of demand is the number of cars observed entering the Restricted Zone during the restricted hours;

2. the measure of price includes the Area License fee, parking charges, and operating costs, all expressed as costs per month;

3. the parking charges used are S$50.00 and S$65.00 per month 1/ for the periods before and after the introduction of the scheme respectively;

4. the operating cost data are based on the average reported cost per mile (S$0.22) from the household survey, an average trip length of 9 miles, and 26 trips to work and home again per month;

5. other factors that influence demand remained constant.

The time period for which data exist covers two price changes: the initial introduction of the Area License Scheme, with the complementary increases in parking charges, and the subsequent increase in the Area License fee. As the changes in the Singapore Consumer Price Index over the period in question were very small (up 0.1 percent from March to October and down 1.2 percent from October to May), prices were not corrected for inflation. The flow data are for March 1975, September/October 1975 (average), and February/ May 1976 (average). The results are presented in Table 6.5:

1/ Source: Lim Leong Geok, "Case Study on Singapore" in Better Towns with Less Traffic, OECD, Paris, 1975. - 163 -

Table 6.5: TRAFFIC FLOWS, MONTHLY TRAVEL COSTS, AND PRICE ELASTICITIES OF DEMAND

Before ALS After ALS After Dec. 31, 1975

Area License Fee S$ 0 S$ 60 S$120 /a

Parking Charges S$ 50 S$ 65 S$ 65

Operating Costs S$103 S$103 S$103

TOTAL MONTHLY COST S$153 S$228 S$288

% Change from previous column - +49.02% +20.83%

Traffic Flows /b 42,790 11,363 10,515

% Change from previous column - -73.44% -7.46%

Price Elasticity of Demand /c 1.50 0.36

Price Elasticity (Arc) of Demand /d 2.95 0.33

/a Weighting the different rates charged company cars and private cars by the number of licenses of each type sold from October to December, 1975. If the weights used were the relative numbers of licenses sold from February to November, 1976, the elasticity would be 0.33. /b Number of cars (including company cars and car pools) entering the Restricted Zone during restricted hours (cf. page 142). /c Percentage change in quantity demanded (number of trips) divided by percentage change in price (cost of a trip). /d (Change in quantity t mean quantity)/(Change in price t mean price).

The data for the first change shows that the demand was price elastic; that is, a given percentage price increase led to a reduction in demand that was proportionately larger. However, a further increase in cost from the higher level led to a proportionately smaller decrease in trips; that is, the demand was price inelastic.

This result is in accordance with traditional economic evidence that for most goods and services, the demand becomes less price elastic as the price rises. - 164 -

The elasticity figures presented above are based on very large price and quantity changes and the elasticities from different initial price levels are far apart. Thus, they do not provide much guidance to planners in other cities who are investigating road pricing. A different approach was therefore employed, based on the graphical representation of the data presented in Figure 6.3.

The straight line represents the demand curve that results directly from connecting the observed data points. There is, of course, no evidence to indicate that the demand curve is, in fact, a straight line. The true demand curve might equally well be represented by a curve, such as the ones shown. It should arso be noted that, if the assumption that other factors remained constant is false, the demand curve might have shifted and the three points might not even lie on the same demand curve.

Assuming that the data do come from the same curve, a new series of point elasticities have been calculated using the lower hand-fitted demand curve from Figure 6.3. This curve has been used to estimate the traffic flow changes that would result from successive S$10.00 increases in monthly travel cost. The resulting elasticities were then calculated. 1/ In addition, the cost has been expressed as the percentage of the mean reported household income of a car-owning household. The results are shown in Table 6.6. The elasticities decline continuously as the cost rises.

Assuming that the current cost of car travel as a percentage of household income can be calculated or estimated, a transport planner can determine the part of the Singapore demand curve that corresponds most closely to local conditions and obtain an estimate of the reduction in car traffic that might be expected to result from the price increases under consideration. Although the estimate will be crude, it is based on the only available empirical evidence on the response of motorists to a large price increase.

1/ In Chapter 5 it was noted that work trips to the Restricted Zone had fallen by 13 percent. If the demand curve has shifted by this amount, the price elasticity of demand would be 1.42 and not 1.50 as stated above. Moreover, the elasticities shown in Table 11.3 are overstated by about 7 percent in the S$170-210 price range. Outside that range, the changes are negligible. - 165 -

Table 6.6: ELASTICITIES FROM ESTIMATED DEMAND CURVE

Monthly % of Reported Price Elasticity Cost (S$) Household Income /a of Demand

150 12.0 4.89 160 12.8 3.46 170 13.6 3.15 180 14.4 2.95 190 15.2 2.70 200 16.0 2.11 210 16.8 1.57 220 17.6 0.94 230 18.4 0.72 240 19.2 0.54 250 20.0 0.23 260 20.8 0.12 270 21.6 0.12 280 22.4

/a The mean reported monthly household income for vehicle-owning-households is S$1,250.

THE LEVEL OF THE AREA LICENSE FEE

Observation in Singapore during the morning rush hour reveals extremely light traffic in the Restricted Zone contrasting strongly with heavy traffic on parts of the ring road. This contrast, which is considerably higher than that which could be expected to result from discontinuity at the cordon, and the underutilization of roads within the zone has led observers to conclude that the price for entry to the Restricted Zone was set too high. With the benefit of hindsight, this question may also be examined using the graph in Figure 6.3. Figure 6.3 Demand curve for car trips to the Restricted Zone during restricted hours.

280 (Feb.-May 1976 average

260 . ______

Monthly cost of travel by car 240 J l to the Government Restricted Sep/Oct, 1975 target Zone average (50% reduction)

S$ 220

200 .\

180 S$ 176--

| \ ~~~~~~~~March, 1975> 160 -I __

140 0 10,000 20,000 30,000 40,000 Number of cars entering the Restricted Zone, 7:30 to 10:15 a.m. - 167 -

Assuming that the true demand curve lies somewhere between the straight line and the lower curved line, it appears that the government could have achieved its goal of a 50 percent reduction in car commuting traffic by raising the monthly trip cost from S$153 to between S$176 and S$205. A middle figure of S$190 implies that an increase in monthly parking rates of S$15 and an Area License fee of S$22 per month might well have been sufficient to achieve the target. Given that increased hourly parking fees are seen by businessmen as detrimental to their businesses, it might not have been necessary to increase hourly parking rates.

This discussion does not imply any criticism of the initial choice of license fee by the Singapore government, which had no previous experience to guide it. A reduction in the license fee after it was found to be higher than necessary might have improved short-run economic efficiency, but at the cost of endangering the long-run objective of promoting a long-run change in the attitudes of the public to the ownership and use of cars. However, the December 31 increase in license fees was not justified in terms of technical or economic efficiency. - 168 -

CHAPTER 7. EFFECTS ON PEDESTRIANS

Studies in the UK 1/, 2/ have shown that pedestrians are more concerned about the difficulties of crossing streets than about any other aspect of the walking environment. In Singapore, a survey conducted for the World Bank 3/ revealed the same attitude.

Measures which reduce traffic congestion may have complex conse- quences for pedestrians. Inducing a shift from private cars to public transport will result in longer walking distances for some people. More people will be walking along footways and more will be crossing streets, thus being exposed to possible conflict with vehicles. The reduction in the number of vehicles might make the situation either better or worse as flows become smoother and speeds increase. The net result is not obvious. Therefore, a study was undertaken in Singapore using time-lapse photography to investigate whether pedestrian delays and conflicts with vehicles had increased or decreased as a result of the Area License Scheme. 4/

Until quite recently, little had been done to facilitate travel on foot in Singapore. Recent exceptions include the provision of very wide footways under building overhangs in the redevelopment area of Shenton Way and the covering of a wide monsoon drain by a promenade along Orchard Road. Some new pedestrian overpasses have also been provided, two of them being tied into shopping complexes with escalators as an added incentive for people to use them. Here and there also, footways have been widened, but in much of the city footways are either absent or too narrow. The old style five-foot-ways 5/, while providing both shade and shelter from monsoon rains, are frequently obstructed by traders' goods, bicycles, and other impedi- menta. Moreover, walking along them often involves frequent negotiation of steps and open storm gutters. These problems together cause many people to walk in the street rather than on footways.

1/ TEST (Transport and Environment Studies): Improving the Pedestrian's Environment, a Report for the Department of the Environment, 4 vols., 1976.

2/ Social and Community Planning Research (Barry Hodges): Road Traffic and the Environment. SCPR Centre for Sample Surveys, Preliminary Report, May 1973.

3/ Reported in Chapter 10 below.

4/ The study was carried out for the World Bank by Transport and Environment Studies (TEST) of London. This chapter is based on their data and calculations.

5/ "Five-foot-ways" is the term used to describe the arcaded footways found particularly in the older parts of the city where buildings are mainly "shop houses". They are so called because five feet was the width between the building front wall and the inside of pillars sup- porting the upper floors. - 169 -

Road crossing in Singapore is not very easy. A variety of crossing facilities exist, but they are widely spaced and generally only located at street intersections. Within the Restricted Zone there are currently eleven footbridges, two subways, and several signal controls. However, the latter generally do not afford pedestrians protection from turning traffic, and some of them do not even warn the pedestrians before the motor traffic gets a greer light. Under these conditions, there is not much incentive to use signalized crossings, and jaywalking is common practice.

DESIGN OF THE STUDY

Methods of Observation

Times required to cross streets and data on pedestrian-vehicle conflicts were derived from analysis of 16 mm time-lapse films at representative locations within the Restricted Zone, before and after the Area License Schem was inaugurated. Additional information obtained from the films, which were supplemented by manual counts, included, hour by hour through the day, flows of different types of vehicles, volumes of pedestrians crossing the street, and the densities of pedestrians on the footways and walking in the street. The principal advantages of using time-lapse photography are:

- Without disturbing the activities observed, it makes a permanent record of all visible aspects of the situation and their changes through time, which may be analyzed later for factors that may not have been originally thought important, and which may be replayed forward or backward to observe the sequence of events leading up to any particular situation.

- The action can be stopped to count people and vehicles.

- Times elapsed during street crossings and delays can be accurately determined.

Some limitations inherent in the photographic method are:

- It is sometimes very difficult to find a vantage point for locating the camera where it will have an unobstructed view of all parts of the area to be studied, yet will be unobtrusive enough to avoid attracting the attention of the drivers or pedestrians being observed.

- Both the photography and the analysis of the films requir' expensive equipment, the film and its processing are costly, and a competent photographer is needed.

- The film must be processed before one can be sure that usable pictures have been obtained. - 170 -

On balance it was judged that the advantages outweighed the disadvantages of using time-lapse photography, and the observation sites were selected to minimize the locational limitation mentioned above.

At each of the sites, one weekday was spent making observations before the Area License Scheme was introduced (March/April 1975) and one day several months afterward (February/March 1976). In Stage 1 (before the Area License Scheme), films were taken at one frame per second for 15 minutes during each of six selected hours, three during the morning peak period, one around mid-day, which tends to be a peak for pedestrian activity, one during the afternoon just before the evening peak, and one during the evening peak period. In hours when no film was taken, traffic flow, numbers of pedestrians on footways in the study section, and numbers of pedestrians crossing the street were recorded manually using hand tally counters. Manual counts of pedestrians on footways were also carried out during hours when films were taken because of the difficulty of seeing people on five-foot-ways in the photographs.

In Stage 2 (with the Area License Scheme in effect) it was decided to reduce the risk of manual observation error by relying completely on 15-minute film periods to provide traffic and pedestrian crossing flow data. Manual recording was, however, still used throughout to record numbers of people on footways, some of whom were hidden from the camera. Observation of pedestrians on footways from analysis of still photographs was rejected after initial trials had shown it to be fairly expensive and no more reliable than manual observation.

Site Characteristics

The choice of sites for the observations was guided by the requirements that the conditions at each site should be typical of those experienced by pedestrians at many locations in the city and that they should be places used for crossing by many pedestrians, where conflict with vehicles is likely to occur. In addition, the necessity for a practical camera position with good visibility constrained the choices. After extensive investigation, ten sites within the Restricted Zone were chosen and were observed in Stage 1.

Poor photographs at two of the sites in Stage I and at one other site in Stage 2 eliminated three sites from the comparative analysis, and a fourth site (Site 6) was dropped because the traffic flow there was markedly increased by vehicles avoiding construction on a parallel street. Before- and-after comparisons as measures of the effects of traffic restraint policies are therefore limited to sites 1, 2, 3, 4, 7, and 10.

Figure 7.1 shows the locations on the map, and Figure 7.2 shows the view from the camera position at each of the six sites. The photographs show the numbers of traffic lanes, presence of bus lanes, and location and type of crossing facilities. The limits of the study area are marked on the photographs. The streets at all six sites are one-way streets and carry four or five lanes of traffic, generally with a reserved bus lane. Direction of flow is northeast at sites 1, 2, and 4 and southwest at sites 3, 7, and 10. All six sites are on main routes through the Restricted Zone and carry large through-traffic flows as well as vehicles with central area destinations. IBRD 13185 I / wrThismap has been prepared by the ./ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~WorldBank'sstaff eoclos,vey for . , the co-vmenre oJf te eal'deF of° .- the report to which It is attached. 4.. I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. Thedesomotations osed an,dthe _ / boundariesb sh so- ti map,, -~~~~~~ .~~~~~~~~~~~~~~.~~~~~~. x ~~~~~~~~~dop sot 'mply. on the pan of the -. b World Bak an its affd,ates, -sy odg t -meitothe legal statAs of

;, anyt-rrt-ry or -ayendoroemet , rccepta e of ssch boundaries.

2' -' *,S V ~ nIANUAY;97 4 ~~~~~~~~4

*1 .. I ' .* . ~~~~~~~~~N,1.0ii 2.0

A dtj~~~~~~J'~~~s t A ~ ~ ~~~~- ~KILOMETERS

I-!Aandra R.,,,-. '*1,'

N 'g. - Moioin-ie .0 , ... ; . ,,- ,,, ;a '. N 1 '"'Odcor' 2 ~ 5 jCENTRAL Rsa SINGAPORE :~~~~~ ~~uf ~ ~~~~~~~~~~~~~~~~~_s;.,,-- \ ,e

I ~~~~~Air Quality Monitoring Sites

/ -'~~~~~~~~Roads ( ~~~~~~~~~~Rivers 7' %,One / 7 way streets .... Restricted Zone boundary / ~~~~~~~~~~I Monitoring Sites 4,~~~~~~~~

Figure 8.1 JANUARY 1978

- 173 -

The sites cover footways of different types and varying qualities. There are traditional five-foot ways at sites 3 and 7 and parts of sites 1, 2, and 4. The remainder of site 1 has a reasonably wide, unobstructed footway, part of site 2 lacks any footway at all, and part of site 4 has only a very narrow one. Site 10 has ample, unobstructed footways on both sides of the street.

A range of land-use types is also covered by the sites. At site 1, land use is mixed between offices, shopping, and educational facilities. Sites 2, 3, 4, and 7 are all in traditional shopping areas, while site 10 is located in the rapidly developing heart of the CBD where offices predominate. The different land uses generate different patterns of pedestrian movement, which are reflected in both the magnitudes of flows and their diurnal variation.

Between the stage I and 2 surveys, changes took place which may have influenced pedestrian activity at two sites. First, at site 10 pedestrian crossing flows were almost certainly reduced by the relocation of bus stops from the mid-point of the study section to immediately south of it. Secondly, the demolition of a block of shop houses on the west side at site 4 removed the five-foot-way; this probably accounts for the reduction in numbers of pedestrians using and crossing the street. Presumably these changes affected only numbers of pedestrians and not the delays or hazards faced by those who did cross at these locations. Thus, observations at sites 4 and 10 have been included in the analysis.

TRAFFIC AND PEDESTRIAN VOLUMES

Vehicle Flows

All the streets studied carried fairly heavy traffic during most of the day. The motor vehicle flows, 1/ before and with the Area License Scheme, are shown in Figure 7.3 for the six sites with comparable before and after conditions (1, 2, 3, 4, 7, and 10; see above). The vertical bars, in these charts, show hourly flow rates corresponding to observations made in a 15-minute period within each hour. The flows shown between 7:00-8:00 a.m. are from observations that started at 7:30; those between 10:00-11:00 a.m. from observations after 10:15. Thus, the first three periods correspond to the hours of the Area License Scheme restrictions, and the rest to unrestricted hours.

All the traffic counts in Stage 2 were made from the time lapse photographs, but in Stage 1, photography was used in only six of the time periods and the counts in the other periods were done manually, with results that are considered less reliable. In Figure 7.3, the periods when counts were made from films in both stages are indicated by the letter "F" (for "film").

1/ Excluding motorcycles. - 174 -

Figure 7.2: Views from the Camera

1: Victoria Street 4 _ _ -mm Site

I!-

Shopping Complex - 175 -

Figure 7.2 (continued)

,J I~.. Site 4: New Bridge Road

A ~- Site 7: North Bridge Road

)'F iz

Fi _ | Site 10: Shenton Way - 176 -

Figure 7.3. Hourly variations in traffic flows at each site. SITE I (Victoria Street) F F F F F F 2800

2400 -

2000- ~~~~~~~~~~~~~Key:

2,:1.1 600 _-_ | _ r | r I r 1 TTotal vehicles S. 1600 not including 1200- ~~~~~~~~~~~~~~~motorcycles& bicycles

800 ~~~~~~~~~~~~~~~~~~Busesand trucks

400 Cars, light vans 0 7 8 9 10 11 12 1 2 3 4 5 6 (a.m.) Time (p.m.) H Stage I

I Stage 2 SITE 2 (Hill Street) F F F F F F 0 Stage 1 figures 2800 suspect or not filmed 2400 -

1600

0 7 8 9 10 11 12 1 2 3 4 5 6 (a. m.) Time (p.m.)

SITE 3 (Colombo Court Shopping Complex) F F F F F F 2800 2400-

2000 -

1600-

(a.m.) Time (P.M.) - 177 -

Figure 7.3 Continued

SITE 4 (New Bridge Road) F F F F F F 3600 3200 2800- Key 4 2400- Total vehicles k; _not including ' 2000- motorcycles & bicycles

Z 1600 7_ Buses and trucks 1200 [J Cars, light vans 800

400L T%F

7 8 9 10 11 12 1 2 3 4 5 6 |]Stage2 (a in.) Time (pin.) * Stage I figures suspect SITE 7 (North Bridge Road) or not filmed. F F F F F F 3200 2800 - 2400 -

- 2000 -

X 1600 1200-

400-

7 8 9 10 11 12 1 2 3 4 5 6 (a m.) Time (pm.)

SITE 10 (Shenton Way)

F F F F F F 3200 2800

2400

-E 2000- 1600-

400 ~

7 8 9 10 11 12 1 2 3 4 5 6 (a.m ) Time (p.m

Noic F inJjjitat periods liinimd mi Si.ge I IAll %%crcflmcd in S.ige 21 - 178 -

At all six locations there was a large reduction in traffic during the first three observation periods, with the exception of the earliest period at site 1. The higher level observed in Stage 2 at site I (just after 7:30 a.m.) may well have included a group of cars that had entered the Restricted Zone just before the start of the restricted period. During the unrestricted hours of the day (from 10:15 a.m. onwards), traffic at sites 1, 2, 3, and 10 in Stage 2 was either roughly equal to or lower than that in Stage 1, with reductions especially prevalent between 3:00 p.m. and 6:00 p.m. At the other two sites (4 and 7), Stage 2 traffic exceeded that of Stage I for one or two periods after 3:00 p.m., and at Site 7 it was higher than in Stage 1 during most of the day after 10:15 a.m. Most of the differences were attributabld to automobile and light van traffic. No systematic large differences are to be seen in the flows of heavy vehicles (trucks and buses) either overall or at particular times.

At site 7 a complete analysis of traffic flow was carried out, in order to examine the proportions of "other modes" (motorcycles, bicycles, trishaws, and hand pushed vehicles), which have not been included in the data presented thus far, as well as to determine the proportions of heavy vehicles. Table 7.1 shows the flow volumes at site 7, before (Stage 1) and with the Area License Scheme (Stage 2), for seven categories of vehicles, including two- and three-wheelers and pushcarts. These "other modes" clearly increased after the Area License Scheme restrictions, not only during the restricted hours when other traffic was reduced, but even in the afternoon, when the volume of cars and larger vehicles ("main modes") was slightly higher in Stage 2 than in Stage 1.

Pedestrian Crossing Flows

The patterns of pedestrian crossing volumes generally conformed with expectations, in that peaks occurred around lunch time. (See Figure 7.4). At several sites, second peaks were observed during the evening "rush" period as office workers started home. This was particularly noticeable at sites 2, 4, and 10. At sites 4, 7, and 10 relatively high crossing volumes were also observed during the morning when pedestrians were on their way to work. At most sites, the flow variation profile over the day was similar in Stage 1 and Stage 2 surveys. Pedestrian crossing volumes tend to be higher where high density uses, such as office buildings and shopping centers, predominate and lower where uses are of lower density, such as traditonal shop houses. Thus, the largest crossing volumes were observed at sites 3 and 4 (both high shopping activity) and 10 (in the heart of the office employment area). - 179 -

Table 7.1: ANALYSIS OF "OTHER" TRAFFIC MODES FOR SITE 7

7:00 8:00 9:00 1:00 4:00 5:00 to to to to to to 8:00 9:00 10:00 2:00 5:00 6:00 a.m. a.m. a.m. p.m. p.m. p.m.

STAGE 1 ~ Cars and light vans 747 1 1302 1 1491 1406 1524- 11491 z Goods vehicles 118 93 149 130 8! 57 _ Public service vehicles 223 ! 262 218 1 149 184 1 74 M Motorcycies | 280 387 248 318 392 I 220 Cycles 146 153 1178 92 152 I 129 u Trisnaws 45j 981 931 57 72 4Z o Hand pushed o 4 0! 0 0| 0 Total (all vehicles) 1559 12299 2377 2152 !2404 12319

Total main -lodes 1088 11657 1858 1685 1788 11722 Total otnerrmodes 471 642 519 | 467 616 597

_%. othe`r modes of total 30.2 127.9 21.8 121.7 125.6 125.7

STAGE 2 - Cars and light vans 248 1 668 I 540 11376 11508 1396 c Goods vehicles 140 116 228 124 128___ 100 E Public service vehicles 244 304 j 200 1 184 ! 180 1 le5 M-Motorcycles 283 -380 436 [ 356 i356 3&4 Cycles 256 | 312 260 | 160 212 228

Trishaws__ 40 92 1144 72 80 56 - Hand pushed 12 16 | 0 4 8 4 Total (all vehicles) 1228 [1888 1808 2276 [2472 |2356 Total maJiT modes 632 1088 i 68 1684 1816 11684 Total other m.odes - 596 | 800 840 592 656 67 % other modes of total 148.5 142.4 146.5 26.0 126.5 128.5

Note: Taxis were not separately counted and are included with cars. - 180 -

Figure 7.4. Total pedestrians crossing at each site.

3600 I I ~~~~~~~~~~Key: 3200 SITES I & 10 = erore A L.S - - With A.L S. 2800

-~2400

-~2000

1600 SIT 10

1200 I"A

800

400 S 10

7-8 8-9 9-10 10-11 11-12 12-1 1-2 2-3 3-4 4-5 5-6 (a.m.) Time (p.m.) - 181 -

Figure 7.4 Continued

3600 l l _ l l 3200 _ SITES2& 3 Key,: Before A.L.S.1 2800 With A. S

2400 SITE 3

E 2000 -

' 1600 -

; 1200 - ITE 3

800 -

0 7-8 8-9 9-10 10-11 11-12 12-1 1-2 2-3 3-4 4-5 5-6 (a.m.) Time (P.m.)

3600 1 -

3200 SITES 4 &7

2800

1400_|SIE7 L_l -~2400 SITE 4 2000 I

1600

1200 -SITE 4 SITE 7-

400 - - - - -

7-8 8-9 9-10 10-11 11-12 12-1 1-2 2-3 3-4 4-5 5-6 (a.M.) Time (p.ml.) - 182 -

Data from Figure 7.4 are summarized in Table 7.2, in terms of average and peak flows in each stage. Both the average and peak flows were lower in Stage 2 than in Stage 1 at three sites (1, 4, and 10) and higher at three sites (2, 3, and 7). Extraneous events, mentioned earlier, explain the materially reduced pedestrian crossing volume at site 4 (demolition of the block between Sago Street and Smith Street) and site 10 (relocation of bus stops). No explanation is apparent for the reduction at site 1.

At the other three sites (2, 3, and 7), Figure 7.4 shows that the numbers of people crossing increased in some time periods and fell in others, but -- as seen in Table 7.2 -- both peaks and averages are considerably higher in Seage 2 than in Stage I at these sites. Thus, in three of the four locations where extraneous factors were not identified, more people were observed crossing the streets.

Pedestrian Numbers on Roadsides

At each site, pedestrians walking on footways and in the carriage- way were counted separately and the numbers combined for each side of the road. The results were expressed as pedestrians per 100 meters of footway as a measure of pedestrian density. As would be expected, the pattern of variation over the day was similar to that for crossing, peaks tending to occur at around midday or in the evening, though at some sites high densities were

Table 7.2: PEDESTRIAN CROSSING FLOWS

Site 1 2 3 4 7 10

Average ped/hr Stage 1 475 468 913 2,226/a 567/b ,51/a Average ped/hr Stage 2 354 517 983 1 ,5 69,a 715 868 % Change -25% +10% +8% -30% +26% -25%

Peak Number ped/hr Stage 1 1,164 684 2,100 3,568 712 1,888 Peak Number ped/hr Stage 2 748 760 2952 1,960 856 1,472 % Change -36% +11% +41% -45% +20% -22%

/a One observation of pedestrian counts excluded as invalid. Average based on 10 observations for both Stage I and 2. /b Manual counts abandoned; averages based on six corresponding filmed observations for both Stage 1 and Stage 2. also observed during the morning "rush" period. The three highest observations for each side at each site before and with the Area License Scheme, have been plotted in Figure 7.5. It can be seen that peaks did not always occur at the same time for opposite sides of the street nor were they always at the same time in Stages I and 2. In most cases, each of the three highest pedestrian densities observed at a given site in Stage 2 was higher than its counterpart in Stage 1. Of the 36 pairs, there were only six exceptions, and three of those were at sites (4 and 10) where physical changes have probably inhibited walking. - 183 -

Figure 7.5. Peak pedestrian volumes near each site.

Number per 100 meters WEST I EAST Site 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 . . '~~~~~~15301 1531 1 14131°111 030 H"Before" 1 ~~~~~ ~~13151 |130 Rl ""After" 1730 | 1630 1315 Time at 1230 | _ 1315 beginning hourly obser-of _1315 |1-7 10 vation period 12|15/1630 (24-hour clock)

2~~~~~~~~~~~~~13

2 1 1315 1315

1315 -~~1 830 1630 16731

3 1730 1 173 7 _ 17301[--11

302 1230 1671 . ~~~~~~~~~~~~~~~~1730E X 1,3

l2t.183 135 10 1230 17301230 16301131 _16393 1315 1315~~~~~~~~~~~11 6 1 1230 1630 =X730 1230 1730I230 310 17301735 _ 0 I tg1730 1630~~~~~~123031,

17301~ [ f~ '2/7= _112301121315830

16301~~531 11 1230 ,|30 17230 11 3'1-= 60 123 J14530

11 031 13 15 . _ ~~~~~~~~~~~~~~17,3011263,0 - 184 -

This finding, along with that of the previous section, suggests that there were more pedestrians in Central Singapore after the introduction of the Area License Scheme than there were before, but the magnitude of the increase varied widely from location to location within the Central Area, and the small number of sites successfully monitored was not a representative sample of the whole downtown area. It is possible that reductions in private car traffic and improvements to the public transport system played a part in this increased pedestrian activity. Other factors, such as the gradual occupation of new office buildings, may also have been important.

PEDESTRIAN DELAY

CONCEPTS

Measures of Pedestrian Delay

In assessing the overall impact of traffic policies on pedestrians, the main issue is whether it is easier or harder to cross the street. Pre- vious studies in the UK had successfully used changes in pedestrian delays to measure improvement or deterioration in pedestrian conditions. The same approach was tried in analyzing the Singapore data, in two different forms. Both measures defined delay relative to the "unimpeded crossing time" for each location, which was determined by averaging the crossing times of a sample of people who were seen to cross without being delayed in each of the filming periods.

The first measure, "mean delay to all" in a given period, was defined simply as the mean of all crossing times measured in the period (including time spent waiting on the curb) less the unimpeded time. After extensive attempts to apply it, this measure was rejected as unsatisfactory, because of the wide variations in crossing times of different people who were obviously not delayed by traffic. These variations were apparently affecting the mean crossing times to an extent that obscured the differences in actual delays.

The second delay concept used was "delay to those delayed". The results in these terms were judged to be more meaningful, although some supplemental measures (described below) were required to complete the picture of pedestrian problems. In measuring this concept of delay, crossing times were separately recorded for people who were observed, on the films, to have had to wait for a break in traffic, either before starting to cross or at some point in crossing. Subtraction of the unimpeded crossing time from the mean for those delayed yielded "delay to those delayed".

Proportion of Pedestrians Delayed

To be meaningful, the measure of delay to those delayed must be complemented by a measure of the proportion of all those crossing who experienced delays. If, for example, only two or three out of a hundred - 185 -

people were delayed, it would not matter much whether they were delayed for 10 seconds or 30 seconds. Assuming that the proportion delayed is equal to the probability of being delayed, it might be combined with the duration of delay to those delayed to obtain (in statistical language) the "expected delay". This procedure, however, would not do away with the need for considering the probability of delay separately, as it is probably important from the pedestrian's viewpoint independently of the length of time involved. OBSERVED DELAYS

At Traffic Signals

At three of the six sites (1, 2, and 10) there were traffic signals and crosswalks; some pedestrians used them and some crossed elsewhere within the area observed. In the analysis those crossing at signals were analyzed separately. It was expected that they would not experience any change in conditions unless the signal timings were changed, since the probability and duration of any delays would depend mainly on the signal settings. For those who cross elsewhere, however, changes in traffic volumes and speeds due to the traffic restraints could be expected to have more important effects.

Figure 7.6 shows the mean delays to people delayed in crossing at traffic signals before and after institution of the traffic restraint measures. The mean delay decreased in six out of the eight observation periods that were within the restricted hours, and increased (in some cases markedly) in seven out of the nine "non-restricted hours" periods.

Figure 7.7 shows the proportion who were delayed out of all those crossing at each of the traffic signals. Considerable variation throughout the day existed before the introduction of the Area License Scheme; with the Area License Scheme the variation did not diminish. No consistent pattern can be detected either in the results for each stage separately or in the differences between the two stages.

Thus, the hypothesis that the proportions delayed and the mean length of delays at any signalized site would be stable is not supported by the results. There is some evidence that the Area License Scheme reduced delays to those crossing at signals during the restricted hours.

Crossings Not at Signals

At the three sites without traffic signals, pedestrians often crossed one traffic lane at a time as gaps appeared in the moving traffic, waiting between lanes as they proceeded (a somewhat hazardous practice, as drivers often disregard the lane markings). Some would walk upstream between lanes while waiting for an opening, thus following a zig-zag diagonal path. Even at the sites with traffic signals, many people crossed outside of the crosswalk, without waiting for the signal, and behaved much like those crossing at the sites without signals. Figure 7.6. Mean delays to hose delayed, at signals.

SITE I SITE 2 42.1 SITE 10 40

32 -

28-

~24-

~20

q h~~~~~~~~%~ ~16 N C LaLe afiernoon 14 30, 5 301

A B C A B C A B C

Key Before ALS (Stage I) Unimpeded crossing time- With A LS (Stage 2) Site I- 13 5 seconds A During ALS hours (730.8.30,9.30) Site 2 17 3 seconds B Midday (I j5) Site 3 I15.0 seconds C Late afternoon (4130, 5 30) AL Data not available Figure 7.7. Proportionof pedestrians delayed, at signals.

SITE I SITE 2 SITE 10 lOG 8090 - 60-

640 -*0

*' 350-

20-

A B C A B C A B C

Key Before ALS (Stage I) With ALS (Stage 2) A During ALS hours (7130,8 30,9:30) B Midday(I115) C Late afternoon (4 30, 5:30) A. Data missing 0 Zero - 188 -

The mean delays to those delayed (not at signals) are shown in Figure 7.8 for five of the six sites, including two that had signals. (Site 10 had to be omitted from this analysis because the view of pedestrians outside the crosswalk was obstructed in the second stage.)

The differences between Stage 1 and Stage 2 delays were not consistently in the same direction in different time periods, nor did they follow similar patterns at different sites. At sites 1, 4, and 7, the Stage 2 delays were longer than those of Stage I in all but one time period. At site 2, all but one were shorter. At other sites, there were mixtures of increases and decreases. During the restricted hours, the mean delay increased in 9 and decreased in 5 out of 14 observation periods. During unrestricted hours, the mean delay increased in 11 and decreased in 4 out of 15 periods. Thus, there is no clear pattern related to the restricted period. For all sites and time periods the increases in delay outnumber and outweigh the decreases, and variations exist from site to site that are not explained either by the site characteristics or by traffic volumes. (An attempt to find a statistical relationship between delays and traffic volumes yielded an R-squared of 0.14, indicating practically no relation.)

Further consideration of the factors contributing to delays has led to the belief that mean delay is a poor indication of pedestrians' difficulties in crossing streets. Studies in England 1/ have indicated that people accept greater risks when traffic is heavy and try to cross quickly through smaller gaps in the traffic flow. Thus, a reduction in traffic may permit people to cross more easily and safely but with less need to hurry, so that the time taken to cross the road may be greater than before.

The proportion of all people crossing (not at signals) who were delayed is shown for each time period at each site in Figure 7.9. Here, a definite pattern is seen. In 26 out of 30 observation periods, the proportion delayed was reduced in Stage 2. In most cases, the reductions were appreciable and the increases were generally small -- only one exceeded five percentage points.

SUMMARY -- PEDESTRIAN DELAY

Although there was some indication of reduced delay during the restricted hours and increased delay at other times for people crossing the street at traffic signals, no pattern was found in the delays of those crossing elsewhere. Reconsideration suggested that delay is not a good indication of pedestrian problems.

Taking the proportion delayed as a criterion did not reveal any systematic pattern of change for those crossing at signals, but, for people crossing elsewhere, the proportion delayed was reduced in almost all observation periods at all sites.

1/ Summarized in Elkington, McGlynn, and Roberts, The Pedestrian: Planning and Research, London, TEST 1976, pp. 25-27. Figure 7.8. Mean delays to those delayed, not at signals.

SITE I SITE 2 SITE 3 SITE 4 SITE 7 42 7 41.9 40

36-

32-

28-

24-

20 - I 16 -

N~~~ N

N le 10eod WithALS (StAge 2) Sate I I1.Oseconds~~~~~~~~~~N 7 codyl15

Cat ~reNn (43,:0 St 410seod ~mising D.i~ ate 7120 scond KeAData msBcfore Umed72 r0secnd m Figure 7.9. Proportionof pedestriansdelayed, not at signals.

% SITE I SITE 2 SITE 3 SITE 4 SITE 7 100 90 _ 80 - N 70-

50~ -~ ~ ~ ~~ ~ ~~ ~ ~~ ~ ~ ~ ~ ~ ~ ~~~~~

A B C A B C A B C A B C A B C

Kcy Belore ALS (Stage I) With ALS (Stage 2) A 4 ungALShours(7 30,-8 30, 9 30) B Mtdday (I 15) C Latc allernoon 1430, 5 30) - 191 -

CONFLICTS BETWEEN PEDESTRIANS AND VEHICLES

Challenges

A further indicator of conflict between pedestrians and vehicles is the occurence of "challenges". A challenge is defined as an event when a pedestrian crossing the road is forced to hurry or slow down or reverse direction to avoid being hit by a vehicle. The concept was not judged applicable for crossing at traffic signals, where most challenges occur to people violating the signal. For other locations, however, where crossing involves watching for and taking advantage of breaks in the flow of traffic, it seems to be a meaningful measure of street-crossing difficulty.

Figure 7.10 shows the proportions of street-crossers challenged, before and after traffic restraints were implemented, while crossing at places other than signalized crosswalks, at various times of day. The results show a consistent reduction in the proportion of pedestrians ex- periencing challenges. Both during and outside the restricted hours, the proportion challenged is lower in Stage 2 than in Stage I in 26 out of 30 observation periods. Three out of the four exceptions are at site 4, and the increases are small. Altogether, the general picture is of a significant reduction in the probability of a pedestrian being challenged while crossing the street.

Vehicle-Pedestrian Accidents

Yet another indicator of improvement or worsening of conditions for pedestrians might be a change in the incidence and severity of accidents in which they are injured. In Chapter 4 of this report, accident statistics from the Singapore police were reported; no significant difference in the frequency of vehicle-pedestrian accidents was detectable.

CONCLUSIONS

Since the inauguration of the Area License Scheme in Singapore, there seems to have been some increase in the number of pedestrians in the downtown area, while the volume of vehicle traffic has declined, especially, of course, during the restricted hours.

For people crossing streets at traffic signals the analysis of pedestrian conditions was inconclusive. For those crossing outside of crosswalks in the vicinity of traffic signals, and for those crossing at locations without signals, the situation had clearly improved. While the duration of delays was not found useful as an indication of pedestrian problems, the probability of any delay occurring to people crossing streets outside of signalized crosswalks was very much reduced at all times of day, and the net effect is judged to be an improvement in pedestrian crossing conditions. - 192 -

Improved conditions are also indicated by a marked reduction in the probability of being challenged by a vehicle while crossing at any time of day outside of signalized crosswalks and at locations without signals. Figure 7.10. Proportion of pedestrianschallenged while crossing, notat signals.

SITE I SITE 2 SITE 3 SITE 4 900 _ SITE 7

80-

70 -

60-

40 - .

30 Kes~ Ilelore~ ~ ALS (51.sgc ~ ~11~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~ C~~~~~~~~~~ R

Wiih ALS (Sidge 2) A I)urtng AlI S hours (7 30.8 30. 9 301 li Midd.lN (I 151 ( L.iic dtlCrn(K)n (4 3(1.5 3(JI - 194 -

CHAPTER 8. AIR POLLUTION

One of the beneficial effects expected to result from reducing congestion was a reduction in pollutants in the air. Pollutant levels in Singapore before the Area License Scheme were below limits established in the United States and European countries, and pollution was not considered a serious problem. Nevertheless, it was judged useful to collect data that would show to what extent the restraint policies were affecting the quality of the air. This would help in making judgments about the possible air quality effects of such policies under future conditions in Singapore or in applications in other cities.

THE AIR QUALITY MONITORING PROGRAM

To ascertain whether a reduction took place and, if so, to what degree, the Anti-Pollution Unit (APU) of the Singapore Government carried out an air monitoring program. 1/ The program started in April 1975, about two months before the Area License Scheme went into effect. Results reported here cover this pre-ALS period and the first four months after the License Scheme was instituted, i.e. June and July 1975, when the area restrictions were in effect from 7:30 a.m. to 9:30 a.m., and August and September 1975, when the hours of restriction were from 7:30 a.m. to 10:15 a.m. Although monitoring continued until February 1976, the results after September are not included here, because a distinct change in climatic conditions occurring at about the end of September would invalidate comparisons between the period after that and the earlier period.

The pollutants monitored were carbon monoxide, oxides of nitrogen, and smoke. Carbon monoxide and oxides of nitrogen are the principal injurious products given off by the gasoline engines used in most automobiles. Diesel engines, which power buses, taxis, and most of the large trucks in Singapore, produce hardly any carbon monoxide and generally less nitrogen oxides than gasoline engines; their principal pollutant is smoke, which, although far less injurious than the gases mentioned, is a nuisance.

A preliminary study by consultants 2/ and discussion with several experts led the Bank researchers to recommend focusing monitoring efforts, if possible, on measurements of carbon monoxide (CO) as a potential health hazard and as a reliably measurable indicator of the rest of the complex mixture of pollutants produced by gasoline engines. The APU agreed to conduct

1/ Results reported in this chapter are taken from a report on that program by the Anti-Pollution Unit, Office of the Prime Minister, Government of Singapore, 1975.

2/ Environmental Effects of Traffic Restraint in Singapore, TEST, London, 1975, Chapter 3. IBRD 13184 Th,sm,ap has beernprepared by the World Bank's staff eciussicely for the conueshenceof the readersof ,4;;t; J tts i.),'A'a \zs I the reportntowahich it is attached. Thedenominations usedand the !losdv; _ |5d .,SgoRoo u 0 boundariesdo not imply, shown on the onpart this map 0 F of the _~> vr a>>Z | ,o< World Bank and its affiliates, any a~~~~~~~~~~~~~~.~~~0 jd-e ntelegal status o

=t7 ~~~~~~CENTRAL SINGAPORE Jt Sites for Public OpinionanSurvey Study

70 | r P +**^^*-Restricted Zone boundaryol

X 1.0C2n J Figure 10.1

JANUARY 1978

- 197 -

such a program if enough measuring instruments of a suitable type could be obtained for them, since the Unit had only one such instrument. In addition they decided to measure oxides of nitrogen and smoke, for which they were able to make their own instruments. The methods of making these last measurements were such that only 24-hour cumulative readings would be obtained, whereas the CO measurements, if they could be made, would have the advantage of giving a continuous record, showing any variations that took place in the course of a day. Measurement of hydrocarbons was omitted from the program after experiments showed that the level was too low to give meaningful readings with the available instruments.

For monitoring carbon monoxide, the United States Environmental Protection Agency (EPA), at the Bank's request, lent four measuring instruments 1/ to the APU. By moving two of the instruments back and forth between locations, the APU planned to monitor six sites, as described below. Of the four instruments available, however, only one gave reliable service. The APU experts, in spite of putting considerable effort into overcoming the problems of the other three, had to do most of the monitoring with a single instrument, moved each week to a different site. For this reason, it was only possible to monitor four sites in the first and third stages of the program, and only three in the intermediate stage. Thus the results do not include a comparison of CO levels at the ring road with those in the city center.

The monitoring sites were chosen to be close to substantial traffic flows but away from roadworks, traffic signals, or other inter- ferences with the flow. The locations are shown in Figure 8.1 and listed in Table 8.1. Three of them are in the inner core of the Restricted Zone, two are elsewhere in the Restricted Zone, and one is outside the zone, near the ring-road and a fringe car park. The two where carbon monoxide was not monitored are in the outer part of the Restricted Zone (Site 1) and outside the zone (Site 6).

1/ Beckman non-dispersive infrared carbon monoxide analyzers. - 198 -

Table 8.1: POLLUTANTS MONITORED AND LOCATIONS

NO Smoke CO x

Restricted Zone, Outer 1. St. Joseph's Institution, Bras Basah Road * * 2. Hill Street Police Station, Hill Street * * *

Inner Core of Restricted Zone 3. Waterboat Office, Fullerton Road * * * 4. Telecommunication Authority of Singapore, Robinson Road * * * 5. Customs Headquarters, Maxwell Road * * *

Fringe Car Park 6. Newton Circus Post Office * *

RESULTS

CARBON MONOXIDE

The measurement of carbon monoxide levels was done with instruments that recorded continuously, so that the variations through the day were obtained. In spite of the malfunctioning of three of the four instruments, observations were recorded throughout several days per month at four locations by moving the good instrument once a week.

The units of measurement are parts per million (ppm), i.e.: the number of units of carbon monoxide per million units of air. For each hour of the day at each of the four locations, the carbon monoxide (CO) levels from different days were averaged to yield the time profiles shown in Figure 8.2. Results are shown for three different phases: before June 1975 (when there was no Area License Scheme); June 2 to July 31, 1975 (when the Area License Scheme restricted hours were 7:30 a.m. to 9:30 a.m.), and after August 1, 1975 (when the hours were 7:30 a.m. to 10:15 a.m.).

The pre-ALS time profiles in Figure 8.2 show a morning peak and an afternoon peak at each location with a reduced level in the middle of the day. There are noticeable differences in the timing and breadth of peaks at different sites, in addition to the level at two of the sites being about twice as high as at the other two. While three of the four sites exhibit a peak during the 7:30-10:00 a.m. period, the morning peak at Robinson Road does not occur until the 10:30-11:00 a.m. period. Regardless of the timing of this first peak, the CO levels at all four locations declined during the middle of the day and rose again to a second peak between 5:00 p.m. and 6:00 p.m. - 199 -

In order to look at the results without the minor hour by hour variations, average readings for different periods of the day, such as the morning peak, middle of the day, afternoon peak, etc., were calculated for the same locations and months shown in the graphs. The average levels are shown in Table 8.2. The periods of the day were chosen to isolate times during which specific changes were expected. Thus, one period is only a half hour long in order to capture the effect of the short pre-7:30 a.m. traffic peak that developed in some places after the Area License Scheme was instituted; the next period, 7:30-10:00 a.m., includes the morning peak before the Area License Scheme and corresponds roughly to the Area License Scheme restricted hours. Longer periods -- 5 or 6 hours -- were used for the off- peak hours in the middle of the day and at night.

The data in Table 8.2 for "Before June 1975" show peaks and low periods corresponding to those in the graphs -- morning and afternoon peaks with lower concentrations in the middle of the day, except at Robinson Road where, as observed in the graph, the first peak occurred after 10:30 a.m., in the "Middle of Day" period of the table rather than in the period labeled "Morning Peak".

The most significant impact of the Area License Scheme was, of course, expected to be observed during the restricted hours. However, it was not clear in advance to what extent the reduction in morning peak traffic might improve the air quality over a longer time span. The expected reduction in CO in the 7:30-10:00 a.m. period is clearly evident at three of the four locations, as seen both in Figure 8.2 and Table 8.2 for both the June-July stage and that after August first. At Fullerton Road, Hill Street, and Maxwell Road, the morning peak was cut down by about 30 percent, to a level below that of the inter-peak hours. At the Robinson Road site, the graph shows a reduction of the peak, even though the peak was after the end of the restricted period. It also shows some redistribution during the restricted period, with a small increase in the 7:30-10:00 a.m. average, as also shown in the table.

Beyond the morning peak period, the results are mixed. At two of the three sites for which data are available in June and July (Hill Street and Maxwell Road) there was clearly a reduction at nearly all hours of the day in those months, but a return to pre-ALS levels in August and September. These changes do not correspond to changes in overall traffic volumes, and serve as a reminder that the CO levels are affected by a number of factors extraneous to the traffic restraints, including wind, air temperature gradients, and local traffic tie-ups, which may well have varied from one month's observations to another. Thus, small changes in absolute values cannot be unequivocally attributed to traffic policies. The suppression of the morning peak, however, is consistently observed as-a change in the pattern at all four locations, so that there can be no doubt that it resulted from the reduction in morning peak traffic. - 200 -

Figure 8.2. Carbon monoxide variations through the day.

20 ROAD Key STATION 3 FULLERTON Before iune 1975 2 June-31 July 1975 ------After 1 August 1975

ISH 10 :*

0 l l l l l l l l l l 2 4 6 8 10 12 2 4 6 8 10 12 (a.m ) Time (p m.)

20 STATION 4~ ROBINSON ROAD

0.~~~~.

0 S Note Datu not wailalble , ~~~~~~~~~~~~~forthis locbation in ,,' inermediate periud

2 4 6 8 10 12 2 4 6 8 10 1 (a.mM Tunte ( p rm) - 201 -

Figure 8.2 Continued

STATION 2 HILL STREET

Key:

15 Before June 1975 c_ _ _ _ 2 June- 31 July 1975 ...... After I August 1975

0 o~~~~~~~

2 4 6 8 10 12 2 4 6 8 10 12 (a.m.) Time (p.m.)

STATION S MAXWELL ROAD

'-~~~...10. 5

O I I I I I I I I I I I I 2 4 6 8 10 12 2 4 . 6 8 10 12 (a.m.) Time (p/.m) Table 8.2: CARBON MONOXIDE LEVELS--parts per million

Time Early o14orning Middle IAfternoon Before After 24-Hour Interval Morning - Peak of Day Peak Midnight Midnight Average Station 0600 0700 0730 -- 1000 1600 1900 2400 -0700 -0730 -1000 -1600 -1900 -2400 -0600

Waterboat Office, Fullerton Road

Before June 1975 6.7 10.9 15.8 11.7 13.0 7.9 1.8 8.9 2 June - 31 July 1975 9.11 13.4 12.2 11.4 10.5 6.5 1.4 8.0 After 1 August 1975 6.6 13.0 10.4 11.7 12.8 5.9 1.6 7.9

Telecommunication Authority of Singapore, Robinson Road'

Before June 1975 8.6 9.3 13.1 15.0 15.3 11.3 6.8 11.7 2 June - 31 July 1975 la ------After 1 August 1975 7.4 15.1 13.6 11.3 13.7 10.5 2.7 10.3

Hill Street Police Station, Hill St.

Before June 1975 5.2 6.7 7.8 6.3 7.7 5.5 3.1 5.7 4.2 2 June - 31 July 1975 t1.8 7.0 4.h 5.0 5.5 b.4 2.0 After 1 August 1975 5.0 6.1 5.h 6.8 8.5 h.8 2.9 5.5

Customs Headquarters, Maxwell Road

Before June 1975 6.2 6.5 7.6 6.4 7.5 )1.2 3.0 5.4 2 June - 31 July 1975 5.9 7.5 5.2 5.3 4.2 3.3 2.0 b.0 ter 1 August 1975 5.0 R.o 5.5 6.1 6.6 3.L 2.5 b.7

/a Data not available for this period at this location. - 203 -

It is interesting that after suppression of the large morning peak a lower and very narrow peak in CO levels can be noticed at all four locations between 7:00-7:30 a.m., corresponding to the increased influx of traffic just before the Area License Scheme restrictions go into operation. This peak is small in both height and duration and scarcely distinguishable from other fluctuations later in the day. It is not important except as an indication of the quick response of the CO level to changes in traffic flow. Indications of another peak at the end of the restricted hours are much less clear.

Comparison of the CO levels and patterns on the ring-road with those in the central area would have been useful. It is unfortunate that instrument problems precluded gathering data at the ring-road location. OXIDES OF NITROGEN

Figure 8.3 shows the levels of oxides of nitrogen, averaged over each month and averaged over five locations in the Restricted Zone. The level of nitric oxide (NO) is graphed in terms of the mass of nitrogen dioxide produced by oxidizing it, and the nitrogen dioxide (NO 2) is shown on the same scale and added to get a combined total. Since NO is converted to NO2 in the ambient air at a rate varying with atmospheric conditions, the combined total is a more stable index of the polluting effects of traffic than either component alone.

In the same terms as Figure 8.3, Figure 8.4 shows the nitrogen oxide levels at Newton Circus Post Office, which is at an important junction on the ring-road, just outside of the Restricted Zone.

There was clearly a reduction in the oxides of nitrogen, both within the Restricted Zone and at the ring-road location. The patterns are somewhat similar, with the highest level in April and the lowest in August. It could not be said, however, that the downward trend was a direct result of the traffic policies. With the Area License Scheme starting June 2nd, one would have expected a significant change between May and June, but only a 5 percent difference is seen within the Restricted Zone, and almost no difference at all (in the combined total) at the ring-road location. In contrast, there is a 10 percent to 11 percent drop from April to May, when no drastic change in the traffic situation took place. The biggest reduction took place from July to August, but half of that drop was offset by an unexplained rise in September within the Restricted Zone, and at the ring-road station the September level was fully back to the July level.

The fact that only all day cumulative readings were available made it impossible to see whether peak concentrations during rush hours had been flattened off as in the carbon monoxide observations. Although there was, indeed, a reduction in levels of nitrogen oxides from the two months before the Area License Scheme to the four months afterward, the patterns of change leave doubts about ascribing the decline to the changes in traffic that resulted from the Area License Scheme. - 204 -

Figure 8.3. Oxides of nitrogen in the Restricted Zone.

200 -

150

April May June July Aug. Sept 1975

Key - Total NOx (as NO2) NO (as NO2) -____ - 205 -

Figure 8.4. Oxides of nitrogen near a ring roadjunction.

250 -

200-

150

100 - - -

1S0 - X

April May June July Aug. Sept. 1975

Key: - Total NOx NO (as N02) - 206 -

SMOKE AND HAZE

Two different methods were used to measure smoke. One method measured the mass of particulate matter suspended in the air, per normalized cubic meter. The other measured its effect on visibility. Both methods yielded all day cumulative readings with no indication of peak levels.

Figure 8.5 presents the monthly average data on particles in suspension, averaged over the five sites within the Restricted Zone and for the ring-road location (dotted line). There is some sign of a downward trend, but not clearly related to the traffic restraints. Recalling that the smoke-producing vehicles are the diesel-powered taxis, buses, and trucks, one might have expected an increase in smoke due to the increase in commercial vehicle and taxi activity in the Restricted Zone, starting in June with the beginning of the Area License Scheme, but this would be at least partially offset by the reduced number of taxi trips after restrictions were extended to taxis on June 22. Whatever the net result, it should show up in June or July, but Figure 8.5 shows minimal changes from May to June to July. The relatively large steps from April to May and from July to August simply cannot be explained in terms of traffic phenomena. For the ring-road area, the outstanding feature of the graph is an unexplained peak in July, with a drop-off in the two months following.

The coefficient of haze, which measures the visual effects of smoke and haze, is shown in Figure 8.6. After the April-May drop, there is no important change or trend within the Restricted Zone, and a slight upward trend at the ring-road.

With weather and industrial sources having possible large effects on the all day averages, and with no data on changes in diurnal patterns, conclusions about the impacts of the Area License Scheme could not be validly inferred from the data on smoke in either form.

CONCLUSIONS

It is clear that the morning peak concentration of carbon monoxide in the central area of Singapore was reduced by the Area License Scheme to a level below that prevailing in the middle of the day. This conclusion is supported as much by the definite change in diurnal pattern at each of four locations as it is by the reductions in absolute values. Thus it is not weakened by the knowledge that extraneous factors may have significantly affected the absolute measurements.

Other pollutants generated by gasoline engines are presumed to have been reduced in approximately the same pattern and degree as carbon monoxide. Measurements of two such pollutants -- the two oxides of nitrogen -- showed a decline in monthly average values, but it cannot be unequivocally attributed to the traffic policies, since the changes did not all occur in the same months as the inauguration or the extension of the Area License Scheme, and since there is no information on diurnal variations. - 207 -

Figure 8.5. Smoke in the Restricted Zone and near a ring roadjunction.

250-

r---- ~ ~ ~ - I ~ ~ … I

200 M J J A S

~150

100

April May June July Aug Sept. 1975

Key: - Average of 5 stations in Restricted Zone a ring……Near road junction - 208 -

Figure 8.6. Haze in the Restricted Zone and near a ring roadjunction.

2.0

1 5

1.0

0 - - - - -

-. 0.5

April May June July Aug Sept 1975

Key - Average of 5 stations in Restricted Zone Near a ring road Junction - 209 -

Measurements of smoke and haze, which are produced by diesel engines as well as by factories and other non-vehicular sources, also gave results that could not be specifically related to traffic restraint policies. - 210 -

CHAPTER 9. IMPACTS ON THE BUSINESS COMMUNITY

A special study was undertaken to ascertain the views of the business community regarding the measures being used to reduce congestion in Central Singapore -- specifically the traffic restraint package comprising the Area License Scheme and increased parking charges in the central area. 1/ After extensive discussions in Singapore, it was decided that a quantitative study would not be an effective method for examination of the changes resulting from the traffic restraint scheme. The availability of data on land values, rents, sales, labor force, etc. was judged to be extremely limited. Moreover, even if adequate data could be assembled, extraneous factors, such as the economic recession, would seriously impede any objective interpretation of the statistics. Thus, it was considered better to use a qualitative approach to obtain insights into the general attitudes and opinions of business managers concerning (a) their expectations and apprehensions before, and (b) their evaluations afterward, regarding the impacts of the traffic restraint scheme on their own businesses and on business in general.

The approach made use of in-depth interviews with representatives of specific sectors of business. The results reported in this chapter in- corporate impressions and observations from the interviews (most of which were taperecorded), together with a thorough content analysis and interpretation of the transcripts. -Where appropriate, comments are reported verbatim to support or illustrate points made by several respondents.

Eight respondents were interviewed before the introduction of the traffic restraint measures, twenty-four afterwards. The companies and business activities represented by the respondents are set out in Table 9.1. The "before" interviews were carried out in late April and early May 1975 and the "after" interviews took place in December 1975 and January 1976. The latter dates were chosen so that the respondents would have had about six months exposure to the traffic restraint measures, thus ensuring sufficient familiarity to permit an accurate assessment of the business impact. The five respondents interviewed in January 1976 were also able to make known their opinions on the increase in road taxes and the area license fee.

It should be noted that findings based on in-depth interviews of the kind used in this study are qualitative and should be interpreted with caution. The extent to which the findings are "representative" is not quan- titatively or statistically established. Nevertheless, it is believed that the findings indicate the "typical" beliefs, attitudes, and opinions of business managers. This belief is strengthened by the fact that the interviews were carried out by executives and managers from a Singaporean company with considerable experience in conducting and interpreting such interviews and with intimate knowledge of the city and its business community.

This chapter reports: first, the reactions of businessmen to the traffic restraint scheme, and second, their assessment of its impact on the business community.

1/ The study was carried out for the World Bank by Survey Research Singapore. This chapter is largely based on their report. - 211 -

Table 9.1: RESPONDENTS' BUSINESS AFFILIATIONS

A. Before:

American International Assurance (AIA) Design Partnership (Architects) Faber Union Ltd. (Property Developers) Far East (Property Developers) Nestle (Marketer of Consumer Goods) Overseas Chinese Banking Corporation Ltd. (OCBC) Robina Departmental Store Chinese Friendly Association (Provision Shopkeeper in the CBD)

B. After:

American International Assurance (AIA) Prudential Insurance Company Design Partnership (Architects) Faber Union Ltd. (Property Developers) Far East (Property Developers) Nestle (Marketer of Consumer Goods) Robina Departmental Store Metro Departmental Store Chinese Friendly Provision Shop Association (an association of provision shopkeepers; two of its members with shops in the CBD were interviewed) United Overseas Banking Hong Kong and Shanghai Bank (replacement to Overseas Chinese Banking Corporation interviewed in Stage I) Taxi Driver (plying in the CBD) Taxi Fleet Owner (operating in the CBD) Shuttle Bus Service Operator (a driver of the supplementary bus service, known as 'City Shuttle Bus') Comfort Bus Operator (a driver of another supplementary bus service known as 'Comfort' Bus) Company that re-located its office outside the CBD since June 1975 (two) Truck Hire Service in the CBD Wholesaler/Distributor with warehouse in the CBD (two) Fraser & Neave (Soft Drinks Manufacturer) Company in the CBD with 30-50 employees Real Estate Agent - 212 -

REACTIONS TO THE TRAFFIC RESTRAINT SCHEME

PRIOR EXPECTATIONS

Attitudes toward government. The interviews carried out before the introduction of the traffic restraint measures showed very clearly that the respondents' attitudes to the measures were rooted in their general attitudes towards the way in which the country is governed and that their remarks were based largely on their personal observations of how Singaporeans had reacted in the past to the implementation of new legislation. The general consensus was that Singaporeans are disciplined and normally react positively to changes that benefit the masses. The view is explained by the respondents' belief that the government is reasonable, flexible, and pragmatic. They acknowledged that the government can be forceful in implementing changes, but they realized that change was for the benefit of the public:

this is the nature of our Government. It is very flexible." 1/

"Singaporeans understand the role of government."

"The government is pretty reasonable . . . . What it brings in is usually considered for the public's good."

The business managers interviewed have confidence in their government, based on the government's ability to develop Singapore "into a modern city, comparable to Tokyo," and to promote fast economic growth.

In the light of these attitudes towards government, the respondents felt that the implementation of traffic restraint measures would be viewed as another national project likely to benefit the public. It was generally agreed that the scheme was a social necessity.

Perceived Objectives. All respondents indicated a clear understanding of the measures that were about to be implemented, and the majority showed a positive attitude to the scheme. They believed that Singaporeans were aware of the mounting traffic problems in the heart of the city, resulting from increasing use of private cars for travel to work in the Central Business District (CBD). The scheme was perceived as having been designed to ease traffic congestion, by curbing the tendency toward this particular use of private cars.

1/ All quotations are verbatim; it should be remembered that English is not the native language of many respondents. - 213 -

"It is quite clear that the traffic is getting worse and something has to be done. The principle behind implementing restrictions to traffic in CBD is perfectly acceptable."

"It is to discourage people going to town by their private cars. The government hopes to cut down the number of cars entering the CBD to improve the traffic problems."

One respondent perceived that the scheme had a preventive, as well as a curative, side:

"Tackling the traffic problem before it becomes insoluble is far better than reaching a stage when the whole lot just comes grinding to a halt."

Two respondents (termed "discontents" in the original study report) felt that the traffic restraint scheme was part of a government plan to make money out of the public, by increasing the revenues of the bus companies (erroneously perceived as owned by the government):

"Improve the bus system, the roads, but, for heaven's sake, don't think in terms of making money from them."

Overall, respondents felt that the success of the scheme would depend on the broadness or narrowness of the individual's perception of the government's objectives:

"The government sees [the problem] from a broad spectrum. You, as the motorist, see it from a motorist's point of view. If you see it from a broad way, the objective is good, from a narrow view, it is bad."

Nevertheless, respondents were confident that most Singaporeans would understand the need for such a scheme and adjust their travel patterns to achieve the objective of reduced congestion. The social objectives of the scheme were perceived as going beyond reducing congestion. It would also help to reduce air pollution:

"I would say . . . . the scheme would be regarded as a positive gain, because the city will run better. The people in town will gain at least from less polluted air."

Finally, from the social point of view, respondents expressed expectations of improvements in bus services. Attitudes to public transport were in general negative, but respondents pointed out the necessity for improvement and expressed hopes that it would occur: - 214 -

"The demands will be on the buses. It is . . . . essential that the buses are well managed."

At a more personal level, respondents felt that, although there would be some inconvenience, there would also be gains; including money saved and even exercise gained by those who gave up driving and rode buses, and easier access for those who paid the price to continue driving:

"Ironically, we have become slaves to our cars; now we have to adjust and learn to live with some discom- fort."

o. . . . motorists will use their legs more often; it is healthier this way."

"I envisage less frustration in driving to the city area; less frustration also in trying to find a parking space."

Assessment of the Design of the Scheme. There was general agreement that the restricted hours were appropriately scheduled and that the two-hour restriction would not cause any real problem to motorists. 1/ Top managers were expected to pay the fee or practice "Stagatime." 2/ The consultants report a perception that "it is inappropriate for people wearing neckties to travel by bus." Others would try to enter the area just before or just after the restricted hours, perhaps leading to congestion at these times. Others might by-pass the center, perhaps leading to congestion on the ring-road.

There was doubt as to whether the alternative of car pooling would be used more extensively, due to problems of compatibility and scheduling. There was agreement that the Park-and-Ride Scheme would probably not be popular: first, time would be lost in the transfer process; second, the cars would be left exposed and unused in the fringe parking lots; and third, the cost did not compare favorably with other forms of public transport.

The determination of the government to reduce congestion was plainly perceived:

"Presumably the government will increase the charges until it becomes unbearable to drive your car into the CBD."

1/ The restricted period was originally from 7:30 a.m. to 9:30 a.m. After two months of operation it was extended to 10:15 a.m.

2/ The term used in promoting staggered work hours. - 215 -

"It is very likely that the government will extend the restricted hours, maybe from 9:30 a.m. to 10:30 a.m., or even earlier than 7:30 a.m.; this will leave no choice to the motorist."

Expected Compliance with the Scheme. Respondents believed that the restrictions would be observed by motorists without fuss. This acceptance could be perceived as a manifestation of their social consciousness -- a will- ingness to cooperate with the government in achieving its goals. Overall, it was believed that, if the primary objective is to alleviate growing peak-hour congestion in the city, the scheme would have no difficulty in achieving its goals.

EX POST EVALUATIONS

Six to seven months after the introduction of the Area License Scheme, respondents reported their assessment of the success or failure of the following components:

(i) Area License Scheme (ii) Parking Policy (iii) Park-and-Ride Scheme (iv) Promotion of Car pooling (v) Promotion of Stagatime

The Area License Scheme was perceived as the measure that was most effective in solving the traffic congestion problem in the CBD. The scheme was regarded as a success in that it reduced peak-hour congestion by 40 percent, which is above the target of 20-30 percent:

"Personally, I think the ALS is the most effective scheme. It has cut down the traffic congestion tremendously by about 40 percent. Motorists have learned to adjust to this scheme without difficulty."

"I would say it is a brave bold step taken by the Singapore Government."

There appear to be no reservations about the necessity of creating the Restricted Zone. High intensity development was blamed for escalating air pollution and environmental deterioration in the CBD. Mounting traffic congestion was felt by the general public. The delineation of the Restricted Zone was seen as a "justifiable and wise move":

"The root of the traffic problem lies within the now-restricted zone."

The restricted hours -- 7:30 a.m. to 10:15 a.m. -- did not receive adverse comments, except from the two taxi drivers. In general, the extension of the restricted hours to 10:15 a.m. was regarded as necessary, but a further extension would be regarded as a problem: - 216 -

"I think 10:15 is the limit. If they extend it further, it is going to affect the business sector. 10:15 a.m. is just nice [sic]."

It should be noted, however, that similar comments were made during the first phase of interviews regarding the extension of the restricted hours beyond 9:30 a.m.

The fact that the morning restrictions have not produced a mirror- image effect on the evening peak was noted by respondents. They believed that it was due to (a) motorists who bypass the zone in the morning returning through the.zone in the evening, (b) motorists who use car pools or public transport in the morning being picked up in cars by their families in the evening, (c) motorists who park outside the zone during the restricted hours and move their cars into the zone later in the day. Most respondents felt that some form of evening restriction might be necessary to ease the traffic flow. Some had heard rumors that the government was considering such a plan. Taxi drivers did not like the idea.

The five respondents who were interviewed in January 1976 gave their views on the increase in the area license fee from S$60 to S$80. I/ These respondents believed that the increase would have an additional negative effect on taxi drivers, but that the number of cars entering the Re- stricted Zone during the restricted hours would not fall by a large amount.

"Motorists who have no choice but to pay S$60 to get to work in the CBD will continue to pay S$80. They would pay because their jobs demand use of the car."

It is worth noting that the increase in the fee did give the impression that the scheme was aimed at making money. Since the ALS had been successful in reducing congestion by more than the target, there would seem to have been no necessity to increase the fee.

While the scheme successfully controlled congestion in the CBD, some increased congestion was reported in the fringe areas. Motorists making detours to avoid the Restricted Zone cause congestion on the ring-road. Respondents felt that the authorities should look into this problem, which causes frustration to those whose offices are located on the fringe of the CBD. Some respondents perceived this as a symptom of a more fundamental problem.

"So it is a very inefficient way of utilizing our roads while we are jammed on certain roads on the fringe and having no load in the city."

1/ On December 31, 1975 all taxes and fees related to private cars were increased, including the area license fee. - 217 -

Parking Policy. The respondents appeared to accept the increased parking charges with resignation, as a cost to be met if their business obliges them to drive into the CBD for business purposes. Some did note, however, that it is now more advantageous to travel by taxi in the CBD, thus saving time in looking for a parking place:

"Business contacts are not normally short and you might find it more expensive paying the parking fees in the CBD, than hopping into a taxi."

Park-and-Ride Scheme. This scheme was without doubt regarded by respondents as ineffective. They noted that the public's response to it had been poor. The low popularity of the scheme was attributed mainly to the inconvenience incurred in the process of shuttling (transferring) and the high cost involved: time is wasted at the fringe car parks, people do not like leaving their cars exposed (to the weather or to criminal elements), and the high prices are not justified by the service offered.

"The fringe car park [park-and-ride scheme] is a failure. People prefer to park their car at home rather than drive to a fringe car park and then take a city shuttle service bus from there to work. Taking a [shuttle bus] involves inconvenience coupled with the car park fees and bus fares."

The shuttle bus operators interviewed were thankful that the government acted quickly to extend the shuttle bus routes and to reduce fares, thus increasing their earnings. Respondents generally believed that inte- grating the shuttle buses into the general bus system was a wise move. Some respondents felt that the government could have done more to find other uses for the fringe car parks, which were regarded as an economic waste.

"They spent about S$7 million and I think it was a serious mistake. At the same time, knowing our Singapore Government, they would surely not wish to see those white elephants go to waste."

It seems that some of the fringe car parks are put to good use by people trying to form car pools.

"Some motorists drive to the Kallang fringe car park with no intention of using the shuttle bus, but to form car pools to go into the city."

Promotion of Car Pools. The businessmen interviewed before the Area License Scheme doubted that car pooling would be widely used despite the exemption of car pools from the license fee. They cited inconvenience -- especially if a pool member was sick -- and possible personality conflicts. - 218 -

In the second series of interviews, these negative attitudes were still evident with respect to the respondents' own participation in pools, although many others less able to pay for licenses had formed car pools, and several of the managers had organized car pools among their staff members in support of the government traffic policies. The managers interviewed in January (after the fee increase) anticipated increasing dependence on car pools.

"....if the Government keeps up its increase of the road tax ..... the average motorist would have no choice but to form car pools."

Some respondents recognized that car pooling was less effective than it might be to the extent that car pool members were drawn from public transport rather than from driving separate cars.

Promotion of Stagatime. The majority of respondents were in favor of staggered work hours. Several had staggered their own hours so that their business activities would not be affected by the traffic restraint scheme. Wholesalers, salesmen, and departmental stores had all adjusted their operating hours.

It was noted that, while in favor of Stagatime, most respondents felt that they could not allow staff to start after 10:15 a.m. Thus, any extension of the restricted hours could seriously affect business activity.

Public Transport. While most respondents held positive attitudes towards the Area License Scheme, they were also concerned about the quality of public transport -- especially bus services. Most reported an unacceptable standard of service -- notably from the Singapore Bus Service. Buses were said to be dirty, overcrowded, over-susceptible to breakdowns, and unreliable. Bus drivers were said to be reckless and conductors rude. While it is unlikely that many of the respondents were really familiar with conditions on buses through personal experience, their perceptions highlight an area in which there is clearly room for improvement. Moreover, the point is well taken that, if efforts were made to improve services, the general public would respond more readily to travel by public transport.

"Although the bus services have improved after the introduction of ALS, it is not good enough; not up to the standard of what it should be."

According to the respondents, the requirement for taxis to buy area licenses has caused inconvenience to the public. It has resulted in bargain- ing over who should pay the extra charge (usually the customer). For some trips, the taxi could circumvent the Restricted Zone, but this also results in the passenger paying more for the trip. The businessmen were concerned that tourists might be affected by this and get an unfavorable impression of Singapore. - 219 -

Taxi drivers were also displeased at being included in the scheme. Their earnings -- especially those who normally work the morning shift -- have been badly affected.

"Taxi driving business is poor nowadays since ALS was introduced."

There was general agreement among respondents that, because taxis are an integral part of the public transport system, they should be exempt from payment of the Area License fee.

M,ass Rapid Transit (MRT) System. There was enthusiasm among respondents to see the introduction of a mass rapid transit system (subway) in Singapore. Clearly, it is seen as a higher-grade public transport alternative that would be an improvement over the inefficient bus services. It is also clear that the respondents were impressed with the high-technology image of a mass rapid transit system:

"The implementation of the MRT system would be a manifestation of progress and development in Singapore."

SUMMARY

Overall, the respondents, on the basis of positive experiences with government, expected the scheme to be successful and the people to respond to it well. Their ex post evaluations confirm that their expectations were justified.

ASSESSMENT OF IMPACTS ON THE BUSINESS COMMUNITY

In both the "before" and "after" interviews, the respondents' assessments of the impacts of the Area License Scheme on business were structured into six categories:

(i) Sales (ii) Labor Availability (iii) Availability and Cost of Commercial Property (iv) Commercial Location (v) Land Values (vi) Climate for Business

The respondents were asked primarily for their assessment of the business situation in the Central Business District. Thus, it is useful to provide some background on the nature of the CBD in Singapore.

BACKGROUND INFORMATION ON THE CBD

The central business district of Singapore is commonly referred to as "the city" or "the heart of the city" by Singaporeans. - 220 -

The CBD is located around the mouth of the Singapore River and close to the sea front. This development was historical and had the objective of facilitating transportation of goods by sea and river. For similar reasons, the Government offices were also set up along the sea/river front. Then, when private entrepreneurs wanted to set up their business and offices, they followed suit and hence private offices mushroomed around the Govern- ment offices and the sea/river. This was how the "heart of the city" began. As time went by, the "heart" grew to cover the areas which, today, are bound by the perimeter outlining the Restricted Zone. Within the "heart of the city" is a particular stretch of road called Shenton Way that is lined on both sides with banks and financial houses. This setting was deliberately planned a few years ago under the Urban Renewal Scheme -- Shenton Way was to become Singapore's equivalent of the famous Wall Street of New York.

IMPACT ON SALES

Before. In the first series of interviews, the shopkeepers, in particular, and all the respondents in general, could not see how the Area License Scheme would affect their retail business. Most shops normally open around 10:00 a.m. and shoppers can drive into the city after 10:15 a.m. without restriction. Business in the larger shops -- especially department stores -- peaks around lunch time. For smaller shops, sales are predominantly to people in the neighborhood. The only potential problem seen was a loss of sales just after office hours, i.e., between 5:00-6:30 p.m. Most of the shoppers at that time were motorists, who would be likely to find it too inconvenient to shop after work and carry packages home if they switched to traveling by bus. However, the department store respondent felt that this problem could be overcome by opening later in the evenings or on Sunday.

The retailers interviewed saw their business problems more in terms of competition from other traders. Provision outlets in the CBD faced strong competition from the shops and supermarkets that were pro- liferating in the outlying housing estates and other peripheral areas. This decentralization had been going on before the traffic policy package was announced, and there was some concern that the policies might accelerate it.

The wholesaler interviewed felt that the scheme would not have any adverse effect on his business, although he believed that he might have to re-schedule the times at which his salesmen called on customers. He was pleased that the restrictions would not apply to commercial vehicles. Other respondents expressed fears that distribution costs, and hence retail prices, might rise if the restrictions were extended.

Respondents believed that the business of financial institutions -- especially banks and insurance companies -- would not be affected. Either they open for business after the restricted hours, or they work largely during the lunch hour and in the evening.

The only adverse effect anticipated was on the motor car industry, which was expected to continue to be hurt by government restrictions on the ownership and use of private cars. - 221 -

After. The retail sector of CBD business underwent significant changes during 1975. Several small shops (including one of our Phase One respondents) went out of business; one large chain store closed two branches; and all retailers were obliged to re-think the viability of their operations in the CBD. Respondents were asked to assess the contribution of the traffic restraint scheme to this situation.

Retail sales were significantly affected by the lower purchasing power of the consumers (Singapore was in an economic recession in 1975), but retailers in the CBD also faced keen competition from smaller shops, supermarkets, and shopping complexes set up on the outskirts of the city.

"Retail sales in CBD are affected in the sense that shops have decentralized."

In addition, the relocation of several large firms to the outskirts of the city has meant a reduction in the potential CBD market. While the chain stores experience transfers from one (CBD) branch to another (in the suburbs), other retailers have been seriously affected by these changes.

Given that the major factors affecting retail sales were recession and decentralization, respondents were pressed to explain the role played by the Area License Scheme and associated traffic policies. It was reiterated that the Area License Scheme had not affected retail sales: most stores now open between 10:00-10:30 a.m. and they do not expect shoppers until around 11:00 a.m. However, practically all respondents believed that the increase in parking charges had affected retail sales. They explained that the increase had "dampened shoppers' spirits."

"The parking fee is one of the main problems affecting retail sales in the CBD. It has now become too expensive for shoppers to drive their cars to shop here."

It is believed that, if shoppers have to pay more for parking, they will spend less time shopping in the area, and hence spend less:

"The longer one [a shopper] stays, the more one will buy."

The increased parking charges affected small retail outlets more than large ones. A few large shopping complexes offer free parking if purchases are made; others offer free parking for the first hour. Smaller retailers who cannot provide such facilities suffer.

Wholesalers were reported to have made some adjustments to their distribution schedules, but the general consensus was that the scheme had had no adverse effect on wholesale trade.

In terms of sales of services to the public, banking and insurance respondents reported that their business had not been affected at all by the scheme. - 222 -

Summary. Overall, it is clear that the policies to reduce congestion had an impact on sales in the CBD that was minor compared with the and effects of economic recession and the decentralization of housing, shops, business. Moreover, it was only one component of the scheme -- increased parking charges -- that was seen as detrimental to sales. The Area License Scheme itself was not perceived to have negatively affected sales.

IMPACT ON LABOR AVAILABILITY

Before. Considering the fact that the traffic policies were aimed at easing traffic flow and improving the public transport system, most respondents were positive that the policies would not adversely affect the mobility of labor and many felt that they might lead to an improvement.

It was estimated by respondents that around 80 percent of workers do not use cars and that these "junior" workers would not be induced to change jobs. "Top management" (5 percent) either could adjust their hours or have the fee paid by their company, so they would not be induced to change jobs. "Middle management" (15 percent) might be affected in the sense that organizations outside the CBD might exploit "the inconvenience of working in CBD" to attract middle managers to join them.

Overall, it was felt that decentralization would have a stronger impact on labor availability than the traffic policies:

"As more business managers decide to set up offices outside CBD and if the job offers are good, people will move to work with them regardless of traffic restraints."

After. In the second series of interviews, much the same views were expressed. In addition, it was stressed that "job satisfaction" -- the nature of the work, the salary, the hours, etc. -- was more important than the location of the place of work. The level of unemployment was also reported to be a factor:

"Definitely people are willing to work in CBD if there are jobs for them. There are so many people jobless nowadays and they cannot be so choosy just because of ALS."

Although the fear that middle managers might be attracted to jobs outside the CBD was reiterated, the respondents did not report that this was happening on a large scale. Again, they believed that to the limited extent that this was a problem, it was decentralization, not restrictive traffic policies, that was to blame.

IMPACTS ON AVAILABILITY AND COST OF COMMERCIAL PROPERTY

Before. Early in 1975 (before the introduction of the traffic restraint measures), the supply of commercial property in central Singapore exceeded the demand. Respondents mentioned decentralization and recession as causes: - 223 -

"....in the CBD at the moment, you will find the supply of office space far exceeds the demand. People tend to move out of CBD into the outskirts."

"It is not the traffic scheme that affects office space supply. I would say that overall business has been poor. Less new offices are opened."

Given this excess of supply over demand, the developers interviewed speculated that rents within the CBD would either remain the same or fall slightly. However, the business managers interviewed argued that the market would probably stabilize after initial fluctuations, as lower rents inside and increasing rents outside the CBD would restore the relative competitive position of CBD locations.

After. In the second series of interviews, respondents indicated a belief that the excess of CBD office space supply over demand has increased as a result of several companies relocating outside the CBD. There is also a belief that the increase in vacancies in the CBD is linked with the introduction of the Area License Scheme.

"With the introduction of ALS, quite a number of head offices and offices have shifted to ... just outside the CBD zone."

In reality, the major moves cited had been planned before the policies to reduce congestion were first brought up for public discussion in 1974. Also, the total supply of office space was significantly and rather abruptly increased by the completion of several large new buildings in 1975, some before and some after the restraints were implemented. One respondent mentioned this among other factors:

"Nowadays in CBD office space is in excess supply over demand. Partly it is because they have overbuilt too many offices, and partly also because of ALS. Parking is a problem and the fees are high in CBD."

Others did not identify overbuilding as a cause, but vacancies in some of the new buildings were mentioned in describing the slack market even while it was being attributed to the Area License Scheme:

"I noticed that there are many offices moving out after the introduction of this traffic scheme. Shenton House is quite empty too; so is International Plaza. I don't know what Hong Leong Building would be like in future although it is not completed yet but we don't expect high occupancy rate." - 224 -

The result of the excess supply was perceived to be lower rents, although there was some disagreement as to the extent, with estimates ranging from simple stagnation to a 20 percent fall in rents.

Some respondents believed that CBD rents were likely to rise again in the future as office rents increased in the fringes of the CBD. It was also believed that improved public transport and reduced parking costs could help to restore the attractiveness of the CBD.

Summary. None of the managers interviewed had made a systematic comparison of the property market and rental rates inside and outside the CBD before and after the introduction of the traffic limitation measures, yet they all reported excess supply and stagnant or falling rents. The causes are perceived as recession, decentralization and the traffic restraint measures. Although these perceptions may be valid in varying degrees, it is also clear that (a) central Singapore was overbuilt before inauguration of the Area License Scheme, and even more so soon afterwards, (b) the major companies that relocated planned their moves before ALS, and (c) it is unlikely that rental demand could have adjusted so rapidly as to produce excess supply and falling rents in six months as a direct response to ALS. Thus, the evidence tends to indicate that the traffic measures have had only a minor effect on the availability and cost of commercial property. However, the impacts of the measures are generally perceived to be significant, and this may have an important effect on the future decisions of businessmen.

IMPACTS ON COMMERCIAL LOCATION

Before. Decentralization is not new to Singapore. Businessmen had long been looking for and finding opportunities outside the CBD. Re- spondents believed that, if managers had a choice, they would try to operate outside the Restricted Zone. However, for those whose business involves contacts with government, banks, or export/import activities, there are advantages to being in the CBD.

Overall, respondents believed that the Area License Scheme would be an additional incentive to decentralize. It should be noted, however, that this decentralization is not seen as a bad thing. Rather, it is a means of expanding opportunities.

After. The respondents' views on location had not changed since the first series of interviews. Many still reported that, given a choice, they would set up new offices outside the CBD. However, given the new concentrations of population and employment in the suburbs, the advantages of further development outside the CBD seem to be independent of the traffic policies.

Some respondents said it was still more convenient to operate in the Restricted Zone. - 225 -

"It depends on what business you are in. Like us, we need to be near PSA [Port of Singapore Authority] for our import and export business. Our warehouse has to be near to the harbour for loading and unloading."

"For banks it is better that their head office is in Shenton Way for faster communication."

Overall, decentralization in Singapore is not extreme, and it may be concluded that businessmen are still in a position to take advantage of progress towards a more hierarchical pattern of land uses. IMPACTS ON LAND VALUES

Before. In the first series of interviews, respondents made a clear distinction between the rental price of commercial property and land value. They believed strongly that the value of land used for both residential and commercial purposes in the CBD will remain high.

"Rental in Singapore may go down, but not the land value. Singapore is a small city faced with shortage of land."

"About 60 percent of business is in the CBD. It will continue to be important. The land value cannot drop even after the traffic restraint scheme is introduced."

After. In the second series of interviews, two points of view emerged. The property developers and the real estate agent were confident that land values would remain high even though rents had fallen. The other businessmen, however, argued that the fall in rental rates would lead to a fall in land values.

"Land value is only affected marginally because the main thing that has affected the property market is recession and nothing else." (Real Estate Agent.)

"I think land values would be affected as the demand for land within the CBD has dropped." (Taxi Fleet Owner.)

Since land values depend on long-run expectations and not just on current supply and demand, the developers may be right in believing that land values will not fall, given the continuing pressure on land in Singapore. It also seems likely that other factors -- economic recovery, increasing rents outside the CBD -- will stimulate the demand for office space in the - 226 -

CBD, and as the CBD continues to be the busiest zone for business activities in Singapore, land values are unlikely to fall. On the other hand, they are unlikely to grow again at the boom rates of the past decade.

IMPACTS ON THE "CLIMATE FOR BUSINESS"

The term "business climate" was explained to respondents as including the whole range of economic, financial, and government activities in the CBD.

Before. Respondents believed that the business climate would continue at the same level with or without the introduction of the Area License Scheme and increased parking charges. They were inclined to blame business failures on recession.

"Any loss in business is due to competition and recession, but certainly not the introduction [of the restrictions]."

Other policies resulting in demographic changes and population location were seen as affecting the CBD business climate.

"If sales drop off, it is not because of the traffic restriction in the CBD but merely due to the social aspects of the urban renewal policies -- tearing down the slums and re-allogating people. Maybe 1 percent is due to the traffic restraints, but 99 percent is due to government policies in urban renewal."

The only group who might see a change in the business climate in the CBD are the developers. They felt that it would become less lucrative to pursue projects in the CBD once the scheme is in effect. However, their business must already have been adversely affected by decentralization and the over-building of office space.

After. Apart from comments on the retail sector discussed in detail above, comments were limited to the identification of specific businesses that had been adversely affected:

(i) the motor car industry was said to have been crippled by government policies -- including but by no means limited to the Area License Scheme -- designed to discourage car ownership and use;

(ii) small retail shops, restaurants, private clinics, etc. were viewed as having a bleak business climate, given the growing preference for obtaining these services nearer to home and the high parking charges in the CBD;

(iii) taxi drivers were viewed as having been seriously and adversely affected. Business was poor during the restricted hours and earnings had fallen; - 227 -

(iv) the climate for bus operators was thought to have improved, given reduced congestion and a larger number of passengers;

(v) Finally, it was thought that tourists might be given a bad impression of Singapore if they could not get taxis or if they had to haggle over payment for a daily license.

CONCLUSIONS

Overall, the businessmen interviewed responded positively to the scheme. They believed that the Government had taken actions that were necessary and, on the whole effective. In particular, the Area License Scheme was viewed as highly successful.

Regarding business impacts, two responses stood out. First, it was believed that retail sales in the CBD, already falling as a result of recession and competition from stores farther from the center, had been further depressed by the increase in CBD parking charges that accompanied the Area License Scheme. Second, the decentralization of businesses, which was taking place in response to both population dispersion and land pressure in the CBD, was believed to be being accelerated by the restraints on private travel to the CBD.

In general, it is clear that the policies to reduce central area congestion did not, of themselves, initiate changes in either business conditions or location patterns. Rather, they added marginally to the effects of the business recession and to the existing trend toward decentralization. - 228 -

CHAPTER 10. PUBLIC OPINION AND ATTITUDES

The monitoring efforts described in previous chapters were designed to provide objective measures of the changes that had resulted from the introduction of the Area License Scheme. However, the existence of an objectively measured improvement (or deterioration) is not conclusive evidence that the consumers of transport services have actually perceived that change. Nor does the objective measure indicate how the consumers value the improvement or deterioration. Therefore, a supplementary study was undertaken to find out what people subjectively thought of the changes that had taken place. 1/ The approach taken was to use an attitudinal survey to investigate perceptions and opinions about the impacts of the traffic policies. In particular, the study investigated four areas of public opinion:

1. Importance, i.e., the degree to which different features of the transport system or the environment (such as congestion, noise, etc.) are considered important by the respondents;

2. Perception of Change, i.e., what changes in the features noted above have been perceived by respondents since the introduction of the Area License Scheme;

3. Effectiveness, i.e., the degree to which the various traffic policies were perceived by respondents as effective in reducing congestion;

4. Distribution of Impacts, i.e., which consumer sub-groups were perceived as being favorably or negatively affected by the traffic policies.

In each of these areas, respondents were asked to report opinions on a five or seven point scale. The results were processed to yield composite scores for the sample as a whole and for several sub-samples.

THE SURVEY

Respondents were selected on a quota basis to ensure representation of (a) users of different modes, and (b) respondents from both inside and outside the Restricted Zone. Interviews were carried out at eight sites inside and five sites outside the Restricted Zone (see Figure 10.1 for specific locations). The intra-zone sites varied in land use, traffic conditions, and pedestrian activities, and interviews were spread over pedestrian areas near traffic signals, car parks, and bus and taxi stops. Extra-zone sites

1/ Transport and Environment Studies (TEST) collaborated with World Bank staff in the study and questionnaire design, supervised the field work, and carried out the basic data analysis. The interviews were carried out by Survey Research Signapore (SRS), who also assisted in the questionnaire design. IBRD 13181 Thi ,ap has bee prepaed by the World Bank'sstaff eaclusuely for the con-enience of the readersof the reportto which it is attached. '"tah ~~~~~~~~ .n~~~~~ / ~~oaboundar,esc0 GzThe denmiatonsshow sonusedthis and moap the tt ~ ~~ ~ ~~ ~.,AsO5'~~~~~~~~~~~~~~~~' - '1.~~~~~~~~~~~~~ do not usply.on thepart of the -~~~~~~~~~~o"-' ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~a World Bashand its affiliates, any - ~ judg met on the legal stars of ar t 1 5.W any estritoedor anyendorseent 55 0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~oracceptance of sath boundafes.

~~C ~9St , _~ Roads

.nge ~~~~~~~~~~~~~~~~~~~KILOMETERS

CENTRAL SINGAPORE

R-ad _.S;ip

t iCeppel ~~~~~~~~~~Roads,^¢a1..-One way streets Road ...... ~~~~Restricted Zone boundary

... di-~~~ r~~ rn~ Survey Sites

Figur . JANUARY 1978

- 231 -

represented busy shopping areas near the boundary of the Restricted Zone. The extra-zone respondents were asked to respond to the questions with respect to conditions in the area in which the interviews took place.

A total of 1,200 interviews were conducted representing the following groups:

Intra-zone Respondents: Residents - 100 Pedestrians - 100 Car Users (Drivers and Passengers) - 250 Taxi Riders - 200 Bus Riders - 250 Motorcyclists - 100

Extra-zone Respondents: Residents of and Visitors to the Areas being surveyed - 200

The questionnaire elicited basic travel data, subjective opinions on the areas noted above, and basic socio-economic characteristics.

BASIC SAMPLE CHARACTERISTICS

Certain characteristics -- race, sex, age, occupation, income, etc. -- were examined.

Race. The race distribution of respondents -- 75.7 percent Chinese, 14.9 percent Malay, 6.7 percent Indian, 2.7 percent other -- corresponds closely with that of Singapore as a whole. The proportions varied slightly across mode user groups: Chinese were more heavily represented in the Car User, Intra-Zone Resident, and Pedestrian groups; Malays were more heavily represented in the Bus Rider and Motorcyclist groups.

Sex. 66.4 percent of the respondents were male. This corresponds with observations during the pedestrian study that about 60 percent of pedestrians were male. Certain groups were even more male dominated -- motorcyclists 98 percent; car users 71.2 percent.

Age. The age distribution for the sample reflects the pattern in Singapore: 73.4 percent of respondents were under 30 years old.

Occupation. Although 60 percent of the sample described themselves as "professional and executive," "businessman," or "other white collar," the proportion varied considerably among travel groups, ranging from a high of 75.2 percent for car users to a low of 37 percent for pedestrians. These proportions reflect the fact that Singapore's central area is largely devoted to business and commercial activities. - 232 -

Income. The distributions of reported monthly household income are shown in Table 10.1.

Table 10.1: REPORTED MONTHLY HOUSEHOLD INCOME

% Reporting % Reporting % Reporting Respondent Monthly Income Monthly Income Monthly Income Category S$2,000

Intra-Zone Residents 69.5 25.4 5.1 Pedestrians 74.0 22.0 4.0 Car Drivers + Passengers 28.1 43.1 29.8 Taxi Riders 40.2 43.9 15.9 Bus Passengers 63.2 29.1 7.7 Motorcyclists 76.8 19.2 4.0 Extra-Zone Respondents 51.0 38.2 10.8 All Groups Combined 52.4 34.2 13.4

The mean income was about S$1,100 1/ per month, but the variation was considerable. 74.0 percent of pedestrians earned less than S$1,000 per month; by contrast, 72.9 percent of car users earned more than S$1,000 per month.

Car Availability. Overall, 58.2 percent of the sample belonged to non-car-owning households, but for intra-zone residents, pedestrians, bus riders, and motorcyclists, the proportion was much higher. (71-82 percent).

Travel Characteristics. Of the sample of intra-zone respondents, 17 percent lived in the Restricted Zone and 69 percent of the remainder visited the zone daily. Of extra-zone respondents 15 percent lived in the area where the interview took place and 51 percent visited it daily. Fifty- five percent of trips by intra-zone respondents were work-trips, compared with 41.5 percent for extra-zone respondents. The next most reported trip purposes were shopping (15.1 percent) and personal business (11.3 percent). Considerable variation existed between the sub-groups. Motorcyclists and car users made more work trips (and hence, more daily trips); taxi riders made fewer daily trips. Motorcyclists made virtually no shopping trips, but pedestrians and taxi riders made a higher than average number.

About half of the intra-zone respondents usually arrived in the Restricted Zone during the hours of operation of the Area License Scheme (7:30-10:15 a.m.).

Overall, 89.6 percent of intra-zone respondents reported that they had habitually visited the central area for the same purpose before the introduction of the Area License Scheme.

1/ S$2.40 = US$1.00 - 233 -

SUMMARY -- SAMPLE CHARACTERISTICS

The above data confirm (a) that the sample was representative of the appropriate socio-economic and travel groups, and (b) that the respondents haD experience of the conditions that they were being asked to judge and evaluate.

IMPORTANCES

Public opinion and attitudes towards changes in the transport system or in the environment, and, hence, public acceptance of the means used to bring about such changes, depend to a considerable extent on the importance that people attach to the elements or features of the system that change. It is important, therefore, to know whether the Area License Scheme and associated traffic policies changed parts of the system that people consider to be important. Thus, the first stage of the attitudinal survey was to investigate the relative importances placed on different features of the transport system (e.g., travel time) and of the environment (e.g., air pollution). After pilot testing of a tentative list, the following list of features was presented to respondents, who were asked to indicate how important each feature was to them:

1. - having a pleasant and attractive environment in Central Singapore 1/; 2. - having a reduction in traffic fumes and dust; 3. - having a reduction in traffic noise; 4. - in general, feeling safe when crossing a road; 5. - having less crowding in pedestrian areas, that is, five-foot ways; 6. - having less traffic congestion; 7. - being less crowded on buses; 8. - being able to get around more easily in Central Singapore 1/; 9. - being able to get a taxi easily; 10. - the time you take to travel from home to where you want to go; 11. - how much you have to pay to travel from home to where you want to go; 12. - being able to find a parking space; 13. - how much you have to pay for parking; 14. - the time you have to wait for buses; 15. - being comfortable while traveling; 16. - having fewer heavy lorries on the road.

1/ For extra-zone residents, the words "around here" were substituted for "in Central Singapore". - 234 -

The resulting mean importance scores for the sample as a whole are shown in Figure 10.2. While this scale does not necessarily reflect absolute importance, the findings clearly reflect relative importances. The ranking exhibits a characteristic frequently found in responses of this type: features that physically affect the respondent tend to be reported as more important. Thus, personal safety, noise, and fumes all scored highly, while travel times, waiting times, and costs scored below them.

Table 10.2 shows the scores (and rank orders) for each of the travel and environment factors as rated by the different travel groups that made up the intra-zone sample and also by the extra-zone sample.

DIFFERENCES AMONG TRAVEL GROUPS (INTRA-ZONE SAMPLE).

Since this type of study illustrates the relative importances of the features considered, the rank orders of the mean importance scores for each travel group were compared.

Three features -- feeling safe when crossing roads, less traffic congestion, and less fumes and dust -- were ranked most important by the sample as a whole. They all appeared among the top four features for each travel group. The major differences are to be found in the responses of car users, motorcyclists, and taxi users.

Car users ranked ease and cost of parking 8th and 10th (relative to 15th and 16th for the sample as a whole) and they ranked less crowding on and waiting time for buses 11th and 16th, (7th and 9th overall).

Motorcyclists ranked ease of parking very highly -- 5th (15th overall), and considered a pleasant environment to be most important (5th overall). Conversely, they ranked reduction in traffic noise (8th), and waiting time in buses (12th) lower than the average (4th and 9th, respectively).

Taxi riders also differed considerably, most notably in ranking ease of getting a taxi 5th (15th overall). They ranked waiting time for buses (6th) higher than average (9th overall), but ease of getting around (9th) was ranked lower than average (6th overall).

Bus riders ranked less crowding on buses and waiting time for buses equally in 4th place, compared with their overall rankings of 7th and 9th.

Pedestrians and Residents showed no major departures from the overall rankings although residents did rank ease of getting around lower than the sample as a whole (9th compared with 6th).

Among the features under consideration, the scores for waiting time for buses showed most variation, ranging from 4th for bus riders to 16th for car users. Ease of parking, cost of parking, ease of getting a taxi, and less crowding on buses also varied considerably. - 235 -

Figure 10.2 Mean Importance Scores for the Whole Sample

5.0 -- V3Cf rTORTANT

4.5 -7--feeling safe while crossing roads (4.44) (4.30) less traffic congestion -- --reduction in fumes and dust (4.29)

(4.11) pleasant environment-- -- reduced traffic noise (4.12) (4.01)ease of getting around 4.0 - less crowdtdabuee &.00) (3.92) waiting for buses-- --comfort while travelin(3.37;

--cost of travel (3.69) 3.5 (3.42) fewrer heavy lorries-- --less croviding in pedestrian areas(3.43) ease of getting a taxi (3.39)

1.o _.-ease of parlcing (3.03)

--cost of parkinZ (2.78)

2.5

2.0

1.0 UNIFORTAfT Table 10.2: MEAN IMPORTANCE SCORES FOR TRAVEL AND ENVIRONMENTAL FACTORS

_ I [ k---~r. *' I 1/) __ ATJ j Ic4t L. Ii 4oj(A V AI J3. . 1

;- *- 0 o>o | 'U.- C o c s-.z v ; i l 1 - l ' ') Ct- C L

_' LO. nD i ( dk.- _ ., 0 .: CC'

-' C bC: '_ . C C 4- .' tY)0 0 tA 0 CA G< 1i V,__ I)__ _-. CL __Li I-_l L~it 'A 4.^sO f~e1ir.g s3fe when crcssin, roads 1 4.41 4.57 '4.60 4.3? .1.59 4t,7 4.35 1 4.46 1 - - -~~~.54- ri -- 1 _ - . _ - - - - - a)-I: 2 4.26 less traffic congestioa 2 4.30 4.27)4.31 4.26 4.32 4.38 4-.28- 4.32 2 red<:tion in fumes and dust 3 4.29 4.34 4.25 4.27 4.36 4,28 4.33 3 4.30 314.21 5 3.98 reduction in traffir noise 4 4.12 4.33 4.40 4.15 4.17 4.00 4.0 5| 4.15 4 4.17 10) 3.81 pleasarlt environment | 5 4.11 4.11 . 4.14 4.10 4.43 3.97 4.44 -___H_ - _ ._1----- _ _ _ __ -___ _ _ 6 4.01 3 874 0 4.1 3 4.15 3.97 4.09 6 4.05| 9 3.83 easP of g2tting around __t ______I--___ I- I- _- _-- ___ 4.00 less crowding on buses 7 4.00 4.13] 4.06 3.64 4.16 4.21 3.88 7 4.001 4 4.00 7 j3.97| titie to travel froin home 8 3.99 3.91i 3,92 4.C3 4.07 3.88 4.09 7 _------}------.- - i -- ______. ______I____ 9 3.91 G 3. 3 I waiting tnime fo)' buses 9 3.92 4.09 3.94 3.41 4.19 4.21 3.70

3.84 10 i 3.85 5 3. comnfort vihil2 travelling 10 3.871 3.76| 3.68 3.90 4.05 3.77 11 of travel from hlome 11 3.69 3.62' 3.71 3 60 3.92 3.70 3.65 11 1 3 71 361 cost _ I __ __ I H____ j _ _-I t-r~~~~~~r 12 3.36 7 less crowaing in pedestrian areas 12 3.43 3.65 3.45 3.50 3.461 3.39 3.22 14 3.451 | 3.46j 13 3.26 reduction in vo1ume of heavy lorries 13 3.42 3.47 3.22 3.60 3.53 3.251 3.74 12 12 3.z51 14 3.07 ease of getting a taxi 14 3,39 3.42 3.21 3.50 4,21 3.06 3.14 15 2.85 ease of parking 15 3.03 3.00| 2.40 4.04 2.85 2.121 4.19 15 I 307 I I I12I.79, 1 2.79 16 12.73 cost of parking . 16 2.78 Z.781 2,23 3.67 2.56 20.00 3.61 I( _ _L ______I______- J - 237 -

DIFFERENCES BETWEEN INTRA-ZONE AND EXTRA-ZONE SAMPLES

Two main differences may be observed: first, extra-zone respondents ranked a pleasant environment much lower (10th) than intra-zone respondents (5th), but ranked comfort while travelling higher (5th compared with 10th). In addition, less crowding on buses was ranked higher (4th vs. 7th) and ease of getting around lower (9th vs. 6th).

SUMMARY -- IMPORTANCES

In general, the features whose scores differed are those that were expected to be important only to one or two travel groups, e.g., cost of parking to car users. The fact that these results conform to a priori expectations means that they may be used with confidence as a background against which the perception of change results may be examined.

PERCEPTIONS OF CHANGE

A list of characteristics of the transport system and the environment (e.g., traffic congestion, noise) was developed based on the results of the "importance" pilot tests and on expectations about what was likely to change following the introduction of the Area License Scheme. The list is as follows:

1. the amount of traffic fumes and dust in Central Singapore; 1/ 2. traffic noise in Central Singapore; 3. ease and safety in crossing roads in Central Singapore; 4. crowding of pedestrians on five-foot-ways (foot paths) in Central Singapore; 5. general shopping conditions in Central Singapore; 6. traffic congestion during the morning restricted hours in Central Singapore; 7. traffic congestion in Central Singapore during the rest of the day, after the restricted hours; 8. volume of heavy lorries in Central Singapore; 9. the time it usually takes you to get to Central Singapore; 10. number of road accidents in Central Singapore; 11. ease of getting a taxi to go into Central Singapore during the morning restricted hours; 12. ease of getting a taxi during the rest of the day, after the morning restrictions; 13. how much you have to pay for parking in Central Singapore; 14. ease of getting a parking space in Central Singapore; 15. the time you usually take to wait for a bus; 16. standard of bus service in general;

1/ Where necessary, the words "around here" were substituted for the words "in Central Singapore" in interviews outside the Restricted Zone. - 238 -

17. how crowded it usually is in buses; and 18. speed of traffic flow in general.

Respondents were asked whether each of these characteristics had changed compared with the situation before the introduction of the Area License Scheme. If a change was reported, the direction (better/ worse) was ascertained, together with the magnitude (better/much better; worse/much worse). Responses were coded on a scale which ranged from -2 for much worse to +2 for much better. The results are shown in Table 10.3 and the mean scores for the sample as a whole are presented in Figure 10.3.

Overall, the respondents reported that the introduction of the Area License Scheme and associated transport policies had resulted in a general improvement in transport and environmental conditions in Singapore.

On the negative side, it was reported that the ease of getting a taxi during the restricted hours and the cost of parking had become worse. Although the scores for ease of finding a parking space and traffic congestion after the restricted hours are significantly different from zero in a statistical sense, they are so close as to indicate that only very minor changes were perceived.

On the positive side, the largest improvement was reported for general shopping conditions. This confirms the feelings of businessmen (reported in Chapter 9) that the Area License Scheme had not hurt downtown retail business. Relatively large improvements were reported for traffic congestion in the restricted hours, time it takes to travel, and speed of traffic flow, indicating that respondents perceived a real increase in mobility. Ease and safety in crossing roads and the amount of fumes show relatively large improvements, indicating environmental gains, and the standard of bus services and waiting time for buses are also among the higher scores, indicating that public transport also improved.

The significance of these overall results is that respondents reported perceptions of a balanced improvement in conditions -- traffic conditions, public transport, the environment, and shopping -- with no elements being improved at the expense of making others worse.

DIFFERENCES AMONG TRAVEL GROUPS (INTRA-ZONE RESPONDENTS).

Perhaps the clearest indications of travel group differences may be seen at the lower end of the scale, where considerable variation exists on perceptions of no change and changes for the worse.

Ease of Parking. For each group, the score for change in the ease of parking is not statistically significantly different from zero; i.e., there is no clear evidence that any change was perceived.

Traffic Congestion after Restricted Hours. Residents, bus passengers, and motorcyclists reported no change, while pedestrians, car users, and taxi users reported that conditions had got worse. Table 10.3: MEAN SCORES FOR PERCEPTIONS OF CHANGE

M 41~~~~~~~~~~~~~~~~~~~~.

IU wO 1 o 0 S h 04h VARIABLE @ £ o8 la 0 4 2 4-~~~~0 Co cise 14 S4 'd °4 h h0 0 , t h0 40 h 0 h ri r- " h t " 0 9 09 A d .4 cw it E3 E4' 14 0 0 general shopping conditions 1 0.76 o.69 0.78 0.72 0.79 0.92 0.88 2 0.80 1 o.57 traffic congestion in ALS period 2 0.74 0.80 0.72 0.87 0.81 0.76 0.95 1 0.82 8 0.34 ease and safety of crossing roads 3 0.72 0.70 0.79 0.67 0.74 0.79 0.94 3 0.76 2 0.55 standard of bus service 4 0.69 0.74 0.82 0.51 0.78 0.82 0.74 4 0.72 3 0.54 amount of fumes 5 0.60 0.57 0.60 0.56 0.63 0.65 0.85 5 0.63 6 o.43

Time it takes to travel 6 0.55 0.44 0.47 0.52 o.60 o.68 o.68 6 0.58 7 o.42 speed of traffic flow 7 0.53 o.45 o.58 0.57 0.57 o.46 0.69 7 0.54 5 0.45 waiting time for buses 8 0.51 0.30 0.55 0.35 0.54 0.71 0.53 8 0.51 4 0.52 crowding of pedestrian ways 9 0.36 0.34 0.28 0.32 0.33 o.46 0.42 9 0.37 10 0.33 volume of heavy lorries 10 0.33 0.39 0.32 0.35 0.32 0.30 0.4i 10 0.34 12 0.29 amount of traffic noise 11 0.30 0.13 0.39 0.28 0.31 0.23 0.52 12 0.29 8 0.34 crowding on buses 12 0.29 0.34 0.48 0.19 0.35 0.24 0.37 11 0.30 12 0.29 number of road accidents 13 0.27 0.16 0.26 0.30 0.18 0.29 o.46 13 0.27 14 0.26 ease of getting a taxi after ALS period 13 0.27 0.29 0.24 0.13 0.40 0.26 0.29 14 0.26 11 0.31 ease of parking 15 0.06 0.06 0.04 0.09 0.00 -0.04 0.15 15 0.04 15 0.17 traffic congestion after AIS period 16 -o.1o 0.08 -0.16 -0.18 -0.25 -o.o4 -o.11 16 -0.12 16 0.03 cost of parking 17 -0.28 -0.18 -0.10 -0.18 -.043 -0.22 -0.03 17 -0.31 17 -0.12 ease of getting a taxi in ALS period 18 -0.36 -0.03 -0.33 -0.45 -0.56 -0.28 -0.38 18 -o.40 18 -0.20 - 240 -

Figure 10.3 Mean Perception-of-Change Scores for Whole Sample

MUCH BETTER +2.0

+1.5

BETTER +1.0-

(0.76) general shopping conditions, rt 2t9fceRgTVI7Yin (0.72) ease and safety of crossing - standard of bus service (0.69) (0.60) amount of fumes - (0.53) speed of traffic flow -- time it takes to trivel (0.55) ' waiting tide for buses (0.51) (0.36) crowding of pedestrian ways _volume- of heavy lorries (0.33) (number of road accidents =l- crowding on buses (0.29) (0.27) ease of getting a taxi after ALScb ,pperiod - ease of parking (0.06) NO CHANGE 0- -rtraffic con elstion after l_ ALS period 0.10) cost of parking (-0.28) (-0.36) ease of getting a taxi in the ALS- period -0.5 -

WORSE -1.0 _

-1.52-_

MUCH WORSE -2.0_ - 241 -

Cost of Parking. Residents, car users, taxi users, and bus passengers all reported that the cost of parking had got worse. Pedestrians and motorcyclists reported no change.

Ease of Getting a Taxi in Restricted Hours. All groups reported that this had become worse.

Although some differences exist, all groups agreed that the four characteristics noted above had not improved.

At the upper end of the scale, the similarities are more pronounced than the differences (see Figure 10.4). Congestion during the restricted hours was consistently perceived to have undergone a large improvement -- the biggest reported by all groups except pedestrians and bus passengers. Shop- ping conditions and ease and safety in crossing roads were reported among the largest improvements by all groups. Bus service was also among the largest improvements for all groups except car users. The following characteristics showed relatively large improvements for most groups: amount of fares (all groups); time it takes to travel (except residents and pedestrians); speed of traffic flow (except residents and bus passengers); and the time spent waiting for a bus (except residents and car users). The exceptions are not at all surprising; they indicate failures to perceive changes in characteristics that the travel groups in question do not experience.

DIFFERENCES BETWEEN INTRA-ZONE AND EXTRA-ZONE RESPONDENTS

At the bottom end of the scale, the perceptions of change of the extra-zone respondents were slightly less negative than those of the intra- zone respondents. Both groups considered that the cost of parking and ease of getting a taxi during the restricted hours had got worse, but extra-zone respondents reported that traffic congestion after the restricted hours had not changed (worse according to intra-zone respondents) and that ease of parking had improved (no change as perceived by intra-zone respondents).

On the positive side, extra-zone respondents are considerably different from intra-zone respondents in two main ways (see Figure 10.5). First, the scores for the extra-zone group are distributed over a much smaller range than those for intra-zone respondents. This indicates that although both groups perceived positive changes in essentially the same sets of characteristics, the extra-zone group reported smaller changes in most characteristics. Second, while the rank orders of the characteristics are similar (i.e., what changed most according to one group also changed most according to the other), three important differences can be seen (they are indicated on Figure 10.5):

1. The reduction in congestion during the restricted hours -- the largest improvement reported by the intra-zone group -- was (a) perceived by the extra-zone group as a much smaller change, and (b) ranked lower than the changes in many other factors (eighth largest change). Figure 10.4 Largest perceptions of change by travel group

Degree of change perceived by:

intra-zone residents bus motor- respondents of R.Z. pedestrians car users taxi users passengers cyclists +1.0- kALShours congestion

shopping roaas

ALS hotufs cshopping - fumes > hous _ o husseic bu~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~L 0 ALShours hours service X ALcongestion . bus service ALS GJ'-I I ~ ~ ~ ~ AL ho congesbus shopping - congestion crossing rds croassing 0.8- shopping roas a i crogsing Ishopping 1 bus icIAShor V roa s service congeson > - bus service roasing [ service p _ bus service _ALS hours shopping _ rssrvice ~ ~ ~ ~ ~~Icongestion Ibus waits traf ic -H 0.Al - crossing rds. shopping c travel time tael i 44 time o - crossing rds travel -fumes fumes fumes

0.6- , _ fumes c travel time travelu time trafvic _ - ~~~fumes spaeeal - fmspe- spaeeXi 0. ;pe;eui m f bue waits fumes bus waits bus waits _bus waits F travel time noise +0. 5 - .

Note: The relevant part of the scale has been expanded. The complete scale runs from -2 (much worse) to +2 (much better). - 243 -

Figure 10.5 'Perceptions of change by intra- and extra-zone respondents

INTRA-ZONE EXTRA-ZONE

+0.9

(ELSconges hotum on 0.8 shopping

crossing roads

bus service 0.7

- fumes \ 0.6 - travel tihe shopping traffc -crossing roads speed bus service - bus waits- - bus waits 0.5

traffic speed fumes travel time 0.4 pedestrian crowding \ volume of lorries _ _ noise/ALS hour congestion _- pedestrian crowding 0.3on_ro buses n getting a taxi(post restricted hours) 0.3 -_crowdllng on buses _noise crowding on buses

Note: Only part of scale is shown. The complete scale runs from -2 to +2. - 244 -

2. The improvement in the time spent waiting for a bus was reported as greater relative to other characteristics by the extra-zone group (4th largest change compared with eighth largest).

3. The improvement in the amount of traffic noise was also reported as relatively greater by the extra-zone group (eighth largest change compared with twelfth largest).

Clearly, the first difference indicates that, although the traffic policies improved perceptions of the level of congestion in areas peripheral to the Restricted Zone, the reported improvement was not as dramatic as that reported within the zone. The "noise" result is also interesting, since it tends to refute the hypothesis that the Area License Scheme might lead to increased noise in peripheral areas. The extra-zone respondents reported a greater improvement in noise levels (relative to other improvements) than did the intra-zone respondents. It is not clear why extra-zone respondents ranked the change in bus-waiting time differently from intra- zone respondents.

SUMMARY -- PERCEPTIONS OF CHANGE

Overall, respondents reported that most of the characteristics discussed had improved following the introduction of the Area License Scheme and associated transport policies. The improvements covered traffic flow, public transport, the environment, and shopping, leading to the conclusion that the improvements were balanced, if not dramatically large.

For intra-zone respondents, ease of parking, traffic congestion after the restricted hours, cost of parking, and ease of getting a taxi during the restricted hours were reported not to have changed or to have got worse. Little difference could be detected among the responses from different travel groups.

Extra-zone residents also reported an overall improvement, with only congestion after the restricted hours, cost of parking, and ease of getting a taxi during restricted hours reported as the same or worse. The magnitudes of the improvements reported were smaller for extra-zone residents. In particular, congestion during the restricted hours was seen to have im- tproved, but not by very much; it was perceived as only the eighth largest improvement compared with the first as perceived by intra-zone respondents.

RELATIONS BETWEEN IMPORTANCES AND PERCEPTIONS OF CHANGE

The basic hypothesis of attitudinal measurement is that a person's attitude regarding a given change is a function of (a) his perception of that change, and (b) the importance that he attaches to the item or element that has changed. Thus, he may respond more positively to a small change in an element that he considers important than to a large change in an element - 245 -

that he considers unimportant. For example, a motorist may have a neutral attitude towards a large decrease in bus waiting times, but a strong positive attitude towards a small decrease in parking fees. In order to take account of this, the perception of change and the importance scores were multiplied together to produce an attitude score. 1/ These scores are presented in Table 10.4. The mean scores for the sample as a whole, and separately for extra- and intra-zone respondents are shown in Figure 10.6.

The weighting has left the ranking of the elements by the sample as a whole virtually unchanged. Within the individual user groups, the main change has been the higher scores for "ease of crossing the road" for all groups. *This is probably because personal safety almost always elicits a livery important" importance score as all modal users at some point are pedestrians. At the other end of the scale, the higher importances attached to "cost of parking" and "ease of getting a taxi (Area License Scheme hours)" by car users and taxi users, respectively, resulted in stronger negative scores on each element from the corresponding group.

The pattern of differences in intra- and extra-zone respondents' perceptions of change that were reported above can also be seen in the weighted perception-of-change scores.

Thus, overall, it may be concluded that weighting the perception- of-change scores by importances has not greatly changed the picture, based entirely on perception, of the way in which respondents view the changes due to the Area License Scheme.

OVERALL ATTITUDE SCORE

The final step in this analysis was to integrate the attitude scores for each element into an overall index. This index is the average of the individual attitude scores for each user group and is presented in Figure 10.7. This figure shows that each of the respondent subgroups viewed the overall change in conditions brought about by the Area License Scheme as positive. Motorcyclists perceived by far the greatest improvement. The next strongest perception of improvement was felt by bus riders and pedestrians, closely followed by taxi riders and intra-zone residents. The positive change reported by car drivers and extra-zone residents was rather weak. Overall, the sample as a whole produces a mildly positive attitude index. Since none of these scores is strongly positive, it may be concluded that users of Singapore's Central Area believe that the Area License Scheme has, on balance, improved conditions, but by a rather small amount.

1/ The range of the attitude score is from -10 to +10. Table 10.4: MEAN WEIGHTED PERCEPTION-OF-CHANGE SCORES

v0)Ltsli: U 2 +i 3 e 3 to 02 go S0 c0 32 to 2 2 VARIABLE T6 2 1.83 1 22 2 s w 0 co 02 0os .2 0. 1.70 15 1.35 lo 23 Ease ofroad crssing 1 3.22 3.9 3.220)) 2. 4 3,3 3.5 4.00 .1 2.733

Ease of crossing roads 1 3.22 3.19 3.22 2.9 3.36 3.55 1.06 1 2.37 2 3.39

Traffic conditions (A3 Peak) 2 3.21 3.381 3.11 3.821 3.65 3.28 1.058 8 l I 1 3.57

Shopping conditions 3 3.20 2.8tie3.69 2.97 3.52 3.71 3.96 2 2.31 13 3.38

Quality of bus service 4 2.68 2.691 3.13 2.02 3.16 3.20 2.77 2.10 4 2.80 Crowingin aeasedetrin 1.8 126 .171.2 1.1 1.5 1-.5 11 l.0 9 1.31 Fumes 5 2.54 2.h3 2.52 2.32 2.79 2.76 3.63 5= 1.78 5 2.69

Travel Time 6 2.21 1.83 1.82 2.121 2.43 2.62 2.85 7 1.69 6 2.32

traffic flow (Seed) 7 2.12 1.77 2.38 2.24 2.38 1.86 2.70 5=. 1.78 7 2.19

Bus waiting times 8 2.10 1.251 2.20 1.26 2.29' 3.00 2.13 3 2.21 8 2.08

Noise 9= 1.28 0.591 1.70 1.15 1.35 1.01 2.33 9 1.30 10 1.27

Crowding in pedestrian areas 9. 1.28 1.261 1.17 1.20. 1.18 1.52. 1.53 11 1.10 9 1.31

Accidents 11 1.20 o.6 hj 1.19 1.3hi 0.83 1.301 2.03 10 1.13 11= 1.21

Crowding on buses 12 1.19 1.18i82.01 0.751 1.33 1.071 1.36 12 1.09 11= 1.21

Volume of heavy lorries 13 1.15 1.ho0 1.11 1.21 11. 2 0.91 1.58 14 0.93 13 1.19

Ease of getting a taxi(rest of day) 14 0.97 1.19 0.81 0.46 1.59 0.92 1.08 .13 0.99 14 0.97

Ease of parking 15 0.19 0.16 0.05 0.26 -0.02 -0.16 0.66 15 0.63 15 0.11

Traffic conditions(rest of day) 16 -0.51 0.39 -0.79 -0.88 -1.18 -0.24 -0.73 16 0.07 16 -0.63

Cst of parking 17 -1.11 -0.67 -0.38 -2.418 -1.419 -0.73 -0,26 17l -0.57 17 -1.22 Ease of getting a taxi(x' peak) 18 -1.36 -0.91 -1.07 -1.81 -2.25 -0.93 -0.94 18 -0-.7 18 -1.43 - 247 -

Figure 10.6 Weighted Perception-of-Change Scores

AM peak traffic (3.57) ,crossing roads (3.39) crossing roads (3.22) 'shopping cond. (3.38) =rAM peak traffic (3.21)

'shopping conditions(3.20 + 3 _~~~aualjtZ~Tuajtsk~ gy~ -sUeravtevec tzfbo8JS - fumes (2.69) fumes (2.54) Scrossing roads(2.37) _ travel time (2.32) - shopping cond.(2.31) _travel time (2.21) - traffic flow (2.19) - bus wait time (2.21) traffic flow (2.12) _ bus waiting time (2.08) gqualtc ° + 2 72bus waiting time (2.10) - u atn tm 20)-ervict TMU traffic flow(l.78) Sfumes (1.78) travel time (1.69)

- OVERALL INDEX 5pedestrian crowds (1.31) - AM pk traffic (1.45) -pees AiRP?rowds(1.28)8 =no n (1 noise (1.30) - accidnts/bus crowd(l.20) ,acc s u crowd (1.21 acidents (1 13) -heavy lorries (1.15) 'heavy lorries (1.19) =aC9 e(strn r wd(10) + 1 + 1 bt a ~tB~ ~)non-etting ~~ ~ ~ ~h- ~haxio ~ a ~ ~ non-~ ~~%ustaxi,non-AL crowd i S(1.9 ) oA~tflra IhE)onAS hour 0.9) heavy lorries(.93)

ease of parkin (.63) rz~~~~~~~ ~~INTRA-ZONE EXTRA-ZONE 4 IALLRESPONDENTS RESPONDENTS RESPONDENTS

-ease of parking (0.19) ease of parking (0.19) ease of parking (0.11) 'traffic, o ______~~~jion-ALS hours (0.07)

gftE c,non-ALS hrs. a o cost of pkg(-.57) _. _ t~~~~rafilc,non-ALS hrs. .estting a ta -JLS hours Ili,4

cost of parking (-1.11)

- cost of parking (-1.22) ,'getting a taxi, ALS _hours (-1.36) _ o Au_

Note: Only part of scale is shown. The complete scale runs from -10 to +10. - 248 -

Figure 10.7 Overall Index of Perceived Change Weighted by Importance

-- Motorcyclists

1.9 -

1.8 --

1.7 --

106 -- -- Bus Riders

-- Pedestrians

1.5 --

-- Taxi Riders

-- ALL RESPONDENTS 1.4 -- -- Intra-zone Residents

1.3 --

1.2 -- -- Extra-zone Residents

-- Car Users (Drivers and Passengers)

1.1

Note: Only part of scale is shown. The complete scale runs from -10 to +10. - 249 -

PERCEIVED EFFECTIVENESS OF TRANSPORT POLICIES

As the Area License Scheme was not introduced in isolation, the basic public opinion survey was designed to ascertain respondents' reactions to a package of transport policy actions. Part of the investigation, however, probed perceptions of the relative effectiveness of the various policies in reducing congestion. The policies were presented in the questionnaire in two sets:

A. Area License Scheme

(i) in general, the morning restricted peak hours (7:30-10:15 a.m.) of the Area License Scheme;

(ii) required payment of area license fee by private motorists, who pay out of their own pocket;

(iii) required payment of area license fee by company cars whereby the company pays the fee;

(iv) required payment of area license fee by taxis entering Central Singapore with less than three passengers;

(v) exemption of area license fee for car (or taxi) pools;

(vi) ban on very heavy lorries (three or more axles) in Central Singapore during the restricted hours. B. Other Actions

(i) increased road taxes;

(ii) increased parking charges in Central Singapore; (iii) increased cost of petrol;

(iv) Park-and-Ride Scheme with fringe car parks and shuttle buses;

(v) extension of Shuttle Bus Service to supplement Singapore Bus Service (SBS) buses;

(vi) efforts to improve SBS bus services;

(vii) introduction of air-conditioned buses on selected routes;

(viii) introduction of express Blue Arrow buses on selected routes; and

(ix) promotion of staggered work hours. - 250 -

Respondents were asked to rate on a scale from one (no effect) to five (very large effect) the effectiveness of each policy in reducing traffic congestion in Central Singapore. The results are presented in Table 10.5.

Overall, respondents reported that the Area License Scheme in general (i.e., as a package) was the most effective of the measures introduced, closely followed by efforts to improve the SBS services. As the various elements of the Area License Scheme appeared separately, it is not surprising that these elements also scored highly, although none of them alone was rated as high as the complete Area Licensing Scheme package. It is interesting that most of them were ranked as less effective than the extensions of shuttle bus operations and the practice of staggered work hours. There is a significant gap between the scores for the measures mentioned above and the remainder, as shown below:

Policies Judged:

Effective (Score > 3.23) 1/ Ineffective (Score < 2.67) 1/

Area License Scheme in general Payment of fee for company cars

Improvements to SBS service Increased Road Tax

Ban of heavy lorries Increased parking charges

Extension of shuttle buses Park-and-Ride Scheme

Staggered work hours Increased petrol costs

Exemption for car pools Express Blue Arrow buses

Payment of fee by private Air-conditioned buses motorists

The score for the Area License Scheme in general (3.74) is higher than that for any of its components, and significantly greater than the mean component score (3.19). Among the components, the order of perceived effectiveness is a little unexpected:

1. ban on heavy lorries in Restricted Zone during restricted hours (3.46); 2. exemptions of fee for car pools (3.32); 3. payment of fee by taxis (3.25); 4. payment of fee by private motorists (3.23); and 5. payment of fee by company cars (2.67).

1/ Differences of less than 0.23 from the scale mid-point -- 3.0 -- are not statistically significant. Table 10.5: MEAN SCORES FOR ATTITUDES TO EFFECTIVENESS OF TRAFFIC POLICIES

4.) 4:) _ _ _ c- I

I C <]JC a tN UbCZ S t ZDS ) , 4d) a,o 0) 0 (L 0) * (a A r VARIABLE O o- I L- 0in a-J o(> S- f 0 o a)( ai QJ 4J tA CY:: .:J b' .n c_c I |C -_ *,- V ) .- 2 LO .- i.s -u L X VIC 4- C4-) C C4 4J Vf C co 0 0) af3 co O c ,-W IO )GJ xc) AArea 1.icensing Scheme in yereral 1 3.74 3.59 3.55 3.90 3.76 3.69 4.13 1 3.78 2 3.56 [Improvements to SBS bus seevices 2 3.64 3.71 3.55 3.43 3.76 3.78 3 64 2 3.64 1 3.61 Ban of very heevy lorries ir Restricted 3 3.46 3.42 3.10 3.62 3.63 3.17 7 3 3.44 3 3.52 ______.__ Zone ___ I _.___ F

The last Area License Scheme component has a score that places it in a class with other relatively less effective measures that consistently score lower than the top set.

As a group, the mean score for the Area License Scheme components (3.19) is higher than the mean score for all the other policies (2.83).

DIFFERENCES BETWEEN TRAVEL GROUPS (INTRA-ZONE RESPONDENTS)

All the individual travel groups reported the same dichotomy in effectiveness noted above, but important differences appear among groups with respect to the policy measure judged most effective.

Only car users reported the Area License Scheme by itself to be most effective. Bus passengers and residents thought that efforts to improve SBS services were more effective. Pedestrians and taxi users reported the two policy measures to be equally effective, and motorcyclists reported that the Area License Scheme in general together with the ban on heavy lorries were the most effective. It should also be noted that the score given to the Area License Scheme in general by motorcyclists was significantly higher than that reported by any other group.

Certain clear distinctions emerged in the perceptions of effectiveness of the different groups. Relative to the intra-zone group as a whole, residents rated the effectiveness of the payment of Area License Scheme fee by taxis more highly; pedestrians rated the effectiveness of exemptions for car pools lower; car users rated the payment of the fee by private motorists higher, 1/ together with the payment of the fee by taxi, but they rated the improvements to SBS services and the extension of the shuttle buses lower; taxi riders and bus passengers rated the extension of shuttle buses higher; and motorcyclists rated the payment of the Area License Scheme fee by private motorists higher, but the practice of staggered work hours lower.

DIFFERENCES BETWEEN INTRA- AND EXTRA-ZONE RESPONDENTS

Only minor differences can be seen between these two groups. Both groups rated the same two measures at the top of the effectiveness list, but the intra-zone respondents thought that the Area License Scheme in general was a little more effective than improvements to SBS services, while the extra- zone residents thought the opposite.

A comparison of the average scores for "ALS" and "other" policies reveals that the two groups are virtually identical:

1/ Note that the car user group is made up of people using cars after the introduction of the policy measures. - 253 -

Intra-zone Extra-zone group group mean of 5 ALS policies 3.20 3.13

mean of 9 other policies 2.84 2.80

SUMMARY -- PERCEIVED EFFECTIVENESS

It is interesting that the "top ten" measures, in terms of effectiveness, include the Area License Scheme components, improvements to public transport, increased road taxes, and bans on heavy lorries, demon- call upon a broad spectrum of measures that complement each other.

PERCEPTIONS OF IMPACTS ON DIFFERENT GROUPS

To investigate the perceived incidence of the impacts from the traffic policies, respondents were asked to report whether they themselves, different travel groups, and certain other groups had been positively or adversely affected by the introduction of the traffic restraint measures. Specifically, they were asked to consider the effects of (a) the Area License Scheme, and (b) increased parking charges. The response was obtained on a seven point scale ranging from one (very unfavorable effect) through four (neutral) to seven (very favorable effect) and effects on the following groups were considered:.

1. overall effect on yourself; 2. upper income motorists; 3. middle income motorists; 4. company car users; 5. people in car pools; 6. taxi users; 7. taxi drivers; 8. bus passengers; 9. bus companies; 10. pedestrians; 11. residents within the Restricted Zone; 12. residents outside the Restricted Zone; 13. shops and businesses; 14. the Government; and 15. Singapore as a city.

Mean scores for each travel group are presented in Table 10.6; they are also shown graphically in Figure 10.8. Several interesting results emerge from these scores.

First, the overall range of the scores is quite restricted. Although the scale runs from one to seven, the lowest score is 2.43 and the highest 5.57. This indicates that respondents believe that the traffic restraint measures have had no effects that could be considered either highly favorable or highly unfavorable to particular groups. Table 10.6: MEAN SCORES FOR IMPACT OF AREA LICENSE SCHEME ON DIFFERENT GROUPS

-I°u #,nfo In In I*^ I

VI, VARIABLE .nS-- ) 4>e 08 a) cu 4J, zin 'X , c *.- 0) 0~~~~~~~ i- 0. CIL 'fu < n: - In O *iDS.. .xs DIn ~~~~o .. ~~4.)wct1004-) (A e::'Y (A W-11 aJ '0 0 )e 0) a-) 10 co i L.) - - UJw < e- 1 Effect on Singapore Covernment 5.47 5.57 5.40 5.57 5.51 5.39 5.52 5.38 5.49

2 Effect on Singapore as a City 5.20 5.40 5.18 5.10 5.16 5.27 5.41 5.13 5.21

3 Effect on bus con-panics 5.18 5.15 4.98 5.28 5.24 5.02 5.37 5.23 5.17

4 Effect on pedestrians 4.65 4.81 4.49 4.63 4.65 4.63 4.94 4.57 4.67

5 Effect on Central Area residents 4.50 4.54 4.33 4.32 4.50 4.59 4.65 4.48 4.51 6 Effect on bus passengers 4.48 4.24 4.44 4.55 4.48 4.49 4.55 4.48 4.48

7 Effect on ircmbers of car pools 4.40 4.48 4.09 4.35 4.45 4.41 4.74 4.37 4.41

8 Overall effect on interviewqee 4.23 4.42 4.16 3.78 4.17 4.59 4.61 4.15 4.Z4

9 Effect on company car users 4.01 3.96 3.62 4.O0 4.05 4.05 4.24 4.06 4.01 10 Effect on upper income motorists 3.76 3.73 3.67 3.70 3.77 3.74 3.84 3.90 3.74

11 Effect on shops and businesses 3.63 3.SI 3.39 3.22 3.74 3.86 4.02 3.59 3.64

12 Effect on extra-Zone residents 3.61 3.50 3.64 3.34 3.63 3.81 3.99 3.54 3.63

13 Effect on middle incenie motorists 2.89 3.21 2.88 2.63 2.76 3.10 2.96 2.89 2.89

14 Effect on taxi passengers 2.86 2.87 2.98 2.73 2.70 3.07 2.96 2.76 2.87

15 Effect on taxi drivers 2.70 2.73 2.89 2.58 _2.43 2.9:' 2.73 2.72 2.70 Figure 10.8 Impacts of Policies on Different Groups

as rated by:

all residents of bus motor- respondents zone pedestrians car users taxi users passengers cyclists impact on: ~~~~-government . - government 5.5 - government - government -the city -government - government the city the city -bus company bus companies the city -~~~the bustcompanycompanies- the city =bus companies-bus companies -the city the city O5.0 - ,-bus companies. bus companies-

pedestriacpo s pedestrians pedestrians edestrians - car pools - esidntsL-R.Z. pedestrians -pedestriansspedesdrinn _ - R.Z. residnts 4.5 - *-caresidnt=RZ.residnts eestrans - passngrs pa.rssdntrs edestrianbus bus _'RZ residnasatsr ngrs INERVIEWEER.Z. resdnts- busINTERVIEWEE passngrs n

car pools INTE_ VIEWEE -R.Z. residnts= car pools - car pools -INTERVIEWEE~~ ~ _-~~mp. bus passngrsk residnts~~~R.ZaresdtR.Z . residnts o car INTERVIEEE buspassng- INTERVIEWEE - INTERVIEWEE Lcomeanyuses coarn_ - _ny. - car pools L figW ny car _ompany car Jcomrany car_Lshop &bus. users u _(Ucar s - shops zer income _shops & bus.e upper income upper e& -1uper- ss b . minotorists - INTERVTEWEE |- toristpincome - sh esrlts bus. | ugerincmoutside R.Z motorists -pper_residents income emnc - uMRZrin R Z o 50 motoristseidnt o tsNdnt mfris ts~ res n otsdeRZoutside R.Z ~~utsde R.. outsident R.Z mpany car m oiss -rsdnsS e.icoe m ost 3 I id outside R.Z cruse s I-shops & bus. residents middle incomel outside R.Z e § l~motorists F shops & bus. riddleincomt 2 3.0 incom l motorists& jm ists1 o p3 middle inc0meS - > taxi *users - tmiddle income

otorists - taxi users |=,iddle incomE middle incomE axi users "'taxi users I .- ITotorists taxi users I_.motyrists [4 f-taxi drivers -taxi drivers lmiiddlie incomE- taxi users taxi drivers 2.5 taxi drivers

- taxi drivers - 256 -

Second, responses for the sample as a whole may be interpreted as revealing that most people felt that the traffic restraint measures had been the somewhat beneficial to them personally. Government, Singapore as a city, bus companies, pedestrians, Restricted Zone residents, bus passengers, and people in car pools are all seen as having been affected favorably. Motorists, taxi users and drivers, residents outside the Restricted Zone, and shops and businesses 1/ are seen as having been affected unfavorably.

Third, no great differences can be detected among the responses of the different travel groups, when the impacts on the various groups (other than the interviewee personally) are being considered. Indeed, very consistent patterns of response may be seen. The effects on different groups fall into five distinct categories:

Quite favorable effect - government - city - bus companies

Small favorable effect - pedestrians - Restricted Zone residents - bus passengers - car pools

Neutral effect - company car users

Small unfavorable effect - upper income motorists - shops and businesses - residents outside Restricted Zone

Quite unfavorable effect - middle income motorists - taxi users - taxi drivers

The responses of groups of actual travellers reveal a clear perception that pedestrians, bus passengers, car pools, and Restricted Zone residents have benefited at the expense of motorists and taxi users. Among motorists the effect on upper income motorists is seen as only slightly unfavorable compared with the more significant effect perceived to fall on middle income motorists.

Fourth, the perceptions of respondents from different groups of the effects on themselves personally were mostly favorable. Bus passengers, motorcyclists, and Restricted Zone residents reported significantly favorable effects on themselves; pedestrians, taxi users, and extra-zone respondents reported scores that were greater than the neutral value (4.0) but not by a statistically significant margin. Only car users reported an unfavorable impact on themselves, and that was only a small one.

1/ This perception is in conflict with that of businessmen (see Chapter 9). - 257 -

In most groups, the response with respect to the impact on the respondent personally is very close to the response of that group with respect to the group itself. For example, the bus passenger group reported a score of 4.49 for the effects on bus passengers as a class and a score of 4.59 for the effect on themselves personally. The clear exception to this pattern is the taxi user group whose members reported a mean score of 4.17 for the effects on themselves personally, but a mean score of 2.70 for the effects on "taxi riders".

This may reflect the fact that it is difficult to identify regular taxi users, and may indicate that this sub-sample contains respondents who do not think of themselves exclusively as taxi users. Thus, they personally may not experience unfavorable effects, although they perceive that people who use taxis regularly are more unfavorably affected. It is also interesting that the car user group's "interviewee" response is very close to its response for "upper income motorists", and significantly different from that for "middle income motorists". Thus, it may be presumed that the sample of car users (post-ALS) is made up largely of upper income motorists, who perceive themselves to be slightly worse off, but who perceive middle income motorists (many of whom have had to stop being car users) as being much worse off. - 258 -

CHAPTER 11. DESIGNING AN AREA PRICING SCHEME

The World Bank conducted this extensive research program only partly for the purpose of helping the Singapore authorities measure the impacts and repercussions of their traffic restraint measures. A broader objective was to gain knowledge that would be useful to officials of other cities who are also seeking ways to limit congestion and use their streets more efficiently. Accordingly, this chapter summarizes what the authors believe they have learned about the design of traffic restraint schemes, partly from the research findings and partly from personal observation, discussion with officials involved, reading, and discussion with many others concerned with the subject over a period of two-and-a-half years.

It is not intended to suggest that every city should have road pricing or that the Singapore scheme is what should be adopted wherever pricing is judged desirable. Rather, this chapter highlights the factors that should be considered in deciding whether to use road pricing in any specific city with its particular conditions and goals and in designing the details of a pricing scheme if that option is chosen.

SPECIFICATION OF OBJECTIVES

The decision on road pricing and the design of a specific scheme must be based on local objectives and circumstances. The immediate goal of the Singapore scheme was specified in terms of a reduction in the volume of commuter traffic entering the central area in the morning. The accompanying discussion made it clear that such a reduction was expected to achieve the more general objective of changing people's travel habits so as to prevent congestion from becoming a severe problem in the future. It was emphasized that the object was to change people's choices of travel mode and the timing of their trips without discouraging them from coming into or moving around within the city center. Negative effects on commercial transport were to be avoided as much as possible. Such considerations led to the selection for Singapore of a combination of measures that would: (a) impinge mainly on people driving cars into or through the central area during the period of heaviest traffic, (b) provide several alternative means of transport for people who gave up the use of cars, (c) give special dispensation to cars with high occupancy, and (d) avoid restraining commercial traffic.

In other cities the objectives will probably be somewhat different. The makers of policy in one city might wish to limit the number of people going downtown, rather than just influence their choice of mode and timing. Those in another city might want to exert less impact on trips to downtown destinations but discourage people from driving through to destinations beyond the central area. In another, the concern for leaving commercial traffic alone might be absent. Since different objectives require different measures, it is important to have the objectives defined and agreed upon before going very far in working out the design and operating details. - 259 -

CHOICES IN THE DESIGN OF THE SCHEME

Once the objectives have been specified, a series of decisions have to be made about the general approach and specific means of implementing it. The choices are discussed below sequentially, but in actuality they are interdependent, and some iteration is likely to be required. It may be a good idea to work out two or more schemes in some detail and compare the practical problems, costs, and expected effectiveness before making a firm decision.

THE GENERAL APPROACH

The selection of the most suitable combination of pricing methods and other measures for any city warrants careful study in the context of local conditions. Assuming that the objectives call for limiting traffic in a specific area, there are various general approaches to choose from. J. Michael Thomson has compiled a comprehensive catalogue of physical, legal, and fiscal methods of preventing vehicles from going where they would otherwise go or influencing people's travel choices 1/. The present report is primarily concerned with area road pricing. However, it is worth noting that road pricing may be combined with other approaches in a policy package. For example, Singapore, in addition to the Area License Scheme, also made use of parking charges, reserved bus lanes, and prohibition of trucks with more than two axles from the area inside the ring road.

Physical and legal restrictions include barricades and signs forbid- ding entry except by buses and emergency vehicles. These are suitable for setting off relatively small pedestrian shopping areas, but generally not for reducing congestion over a large area without making access too difficult. Area permits may be issued on the basis of profession, for example, to doctors or civil servants, or on the basis of officials' judgements about needs. It is generally considered almost impossible to administer such schemes fairly and to avoid corruption.

A method that controls through traffic without restricting access to destinations within the city has been successfully used in Bremen, Germany, and Goteborg, Sweden. It involves dividing the city into sectors (four in Bremen, five in Goteborg) separated by barriers so that traffic can enter any sector from the ring road but cannot move directly from sector to sector.

Other methods to reduce traffic congestion rely on giving priority to high-occupancy vehicles in bottleneck situations. For example, access to some freeways in the United States has been limited by reducing green light time on the access ramps as flow on the freeway approaches capacity, but

1/ J.M. Thomson: Methods of Traffic Limitation in Urban Areas, a report prepared for the OECD Environment Directorate (Division of Urban Affairs), Paris, 1972. - 260 -

buses, and in some cases high-occupancy cars, are enabled to bypass the queues thus created. A similar approach involved deliberate creation of delays to motorists trying to enter the central area of Nottingham, England, while buses were allowed to pass freely. Such confusion, frustration, and indignation was created that the scheme had to be abandoned. It may be that the failure of the Nottingham scheme was partly due to inadequate public education before the scheme took effect and partly to the unpredict- ability of the waiting times involved in entering the city at different points. There may have been other important differences between the two cases. However, it seems fair to state that the use of delays to ration low-occupancy vehicle access to congested roads succeeded in one case and failed in the other. This should be a warning against casual generalization and an indication of the importance of specifically tailoring each scheme to the local situation.

The other general approach to reducing traffic congestion is "pricing", i.e., charging fees for some aspect of the use of the city streets. Inasmuch as the subject of our research was Singapore, where pricing was the primary measure, the rest of the chapter will focus on pricing approaches and methods of implementing them.

BASIS OF CHARGING

Vehicles may be charged for entry into a designated zone, for operating within it, for parking in it, or for exiting from it. There may be a single zone or multiple zones with different prices. Charges may also be assessed for passing control points corresponding to concepts other than zone boundaries. On any of these bases, charges may be varied with time of day and type of vehicle. All of them are location-specific. Thus they all conform to the idea of focusing the charges on the times and places where congestion needs to be reduced.

Differences among the charging bases will result in different effects on through traffic relative to traffic bound for destinations in the central area, on people spending short versus long periods in the area, and on residents of the restricted area as compared to non-resident commuters.

Ideally, perhaps, the charging basis chosen would be that which came closest to having the desired impacts on different sorts of traffic, for example, requiring the user of each vehicle to pay the extra costs that his vehicle imposes on other road users in congested conditions. In practice, the method of implementation -- i.e., of ascertaining and collecting the amount due -- must be considered at the same time and may weigh heavily in the choice of the charging basis.

METHOD OF IMPLEMENTATION

Table 11.1 lists six general methods of implementation and shows the bases of charging to which each method could be applied. Some practical Table 11.1: IMPLEMENTATION OF DIFFERENT BASES OF CHARGING

METHOD - - -_ ------…------_ - - - Pre-Dated User-Dated Automatic Parking Windshield or Self-Can- Toll On-Vehicle Vehicle Charge BASIS OF CHARGES Sticker celing Ticket Stations Meters Identification Methods

ZONE, entry or exit Yes Yes Yes Yes Yes No

ZONE, presence in Yes /a Yes /a No Yes /b Yes lb No

ZONE, parking in Inconvenient LC Yes No No In principle Ld Yes

CROSSING A LINE Yes Yes Yes Yes Yes No > (not zone boundary)

ARRAY OF POINTS Few ------Possiblye… Yes Yes No Many No No No Yes Yes No

Notes: /a Windshield stickers on cars, either moving or parked, spot-checked by roving wardens. Lb By registering entry and exit, meters or AVI could ascertain length of time spent in the zone, if desired. /c Possible, but other methods are probably more convenient. /d For off-street parking places, entry and exit could be registered electronically. To do it for on-street parking would require equipment in every parking space. /e For a system with few control points, method could be the same as for zone entry or cordon crossing. - 262 - aspects of each method are briefly discussed below (except for conventional parking charging, which needs no discussion at this level). 1/ The subsequent section compares the implications of different charging bases.

Dated Windshield Stickers

Dated windshield stickers were the method chosen in Singapore to show that payment had been made in advance (for the day or month). The Singapore experience demonstrated that it is practical to administer and enforce a system of charging for entry to a zone by using dated windshield stickers. Success of such a scheme is not automatic, of course; the conditions for its success are discussed later in this chapter.

User-dated or Self-canceling Tickets

One inconvenience of the method used in Singapore is that people who only need licenses for one day have to purchase them on the specific day and at locations that are not convenient for everybody. A proposed method that would solve this problem would use a windshield sticker or ticket designed to have portions cut off by the user in such a way as to show the date of use and render it invalid for any other date. The user could purchase these singly or in quantity without committing himself to using them on a specific day. Yet another means to the same end is the "self-canceling ticket" -- still to be perfected -- which would undergo a chemical process and change color, thereby becoming invalid, at a convenient time (e.g., 12 hours) after having its cover removed to display it.

Toll Stations

From observation of the traffic reduction in Singapore it can be inferred that a scheme that reduced traffic so drastically could have used a form of on-street toll collection without seriously impeding the remaining traffic. That, of course, is true partly because traffic was reduced so much more than originally intended. However, even with considerably more drivers buying licenses and entering the zone, it would probably be possible to col lect tolls at the entry points for one-trip entries while continuing the windshield-sticker method for monthly licenses. The feasibility of this arrangement in any city would have to be carefully studied in terms of the space available at entry points and the expected volumes of one-day entries and all other traffic.

Collection at entry would also make matters easier for the taxi passenger who, under the present system, is sometimes forced to ride to the license booth, spending time and taxi fare for the trip as well as paying for the license, before the driver will take him down town.

1/ Fuller description and discussion of various methods will be found in K. Bhatt, J. Eigen, and T. Higgins, "Implementation Procedures for Pricing Urban Roads", Working Paper 5032-3-3, The Urban Institute, Washington, D.C. 1976. - 263 -

Electronic Metering

Two basic electronic approaches have been under study and development for a number of years. One, which seems to have relatively few propo- nents at this time, would require that every vehicle carry a meter, which would be triggered by signals from control points in the streets. An on- vehicle meter could run like a taxi meter, but charging at different rates depending upon the signal received from the most recently passed control point. The signal and, therefore, the charging rate would depend on the location and on the time of day. A simpler version could simply register a charge for entry of the vehicle into a zone and/or its exit therefrom, without accounting for the duration of stay in each zone, or it could register charges for passing through selected intersections or other critical points unrelated to a zone system. In practice, a difficulty with on-vehicle meters is the design of a simple payment system that cannot be tampered with by the user.

Similar functions could be performed with automatic vehicle identification (AVI) equipment, which would identify each vehicle as it passed certain check points and record the information necessary for charging each car owner's account. AVI also requires installation of an electronic unit in every car, but the unit is much simpler than the on-vehicle meter. One difference, which would be especially important if a complicated charging system were used, is that the on-vehicle meter could be designed to show the driver the charges as they were incurred, while bills from the AVI system might come as a surprise to the motorist who did not pay close attention to his crossings of critical points.

Prototype AVI systems have been tested in toll-collection situations and shown to be reliable. 1/ The time will come when they will be used for zone-entry charging or for some more complex approach. Even then, however, there will have to be a parallel method such as toll booths or windshield stickers for non-local visitors or cars that enter the zone too infrequently to warrant installing the electronic devices.

IMPLICATIONS OF DIFFERENT BASES OF CHARGING

Zone Entry or Exit

Singapore's zone-entry charge affects traffic from outside, whether it is bound for within-the-zone destinations or is crossing the zone to go out again. Traffic originating within the zone is unaffected so long as it either stays inside or goes outside without returning during the hours of restriction. The fact that enforcement is by visual observation makes it possible to grant exemption to high-occupancy cars, which would be difficult, if not impossible, with an electronic system.

1/ See Robert S. Foote, "Collecting Revenue for Road Use with AVI", presented to the Annual Meeting of the International Bridge, Tunnel, and Turnpike Association, Seattle, U.S.A., September 1976. - 264 -

In Singapore, the absence of restrictions in the afternoon allows trans-zone travelers who by-passed the zone enroute to work in the morning to go freely through the zone on the way home, with the result that the afternoon peak was not reduced as much as expected. Requiring the license for entry in the afternoon, just as in the morning, would divert some of the trans-zone traffic, but it has not been done because it would probably also deter people from going downtown for non-work purposes. To avoid this unde- sired effect and still divert trans-zone traffic, another possibility would be to require the license for leaving the zone in the afternoon peak but not later in the evening. To date, however, the problem has not been judged serious enough to justify introducing afternoon restraints.

Presence in the zone

If the charges were applied over the whole day, a charge for presence within the zone would affect entering and transiting traffic in the same way as the entry charge, but would also apply to residents within the zone, unless they were issued free or reduced-price licenses. Charging on the basis of presence, however, cannot be as sharply focused on a specific time period, such as the morning peak, since cars that entered before that period would also be charged. The supplementary licensing plan proposed for London 1/ would have charged for presence rather than entry because policing the presence of both moving and stationary vehicles was judged more feasible than policing entry to a zone. Which approach would be more practical in a given city would depend upon the area, number of entry points, and other local conditions.

Parking

Since methods of charging for parking are well known and are already in effect in many cities, it is sometimes proposed that traffic be controlled simply by charging high prices for parking in the central part of a city. However, making such charges have the effect of congestion pricing would not be a simple matter of raising parking meter rates and assessing a tax on commercial parking spaces. In addition, it would be necessary to eliminate the practice by many businesses and government agencies of providing free or underpriced parking to their employees. It would also be necessary to change the usual schedule of diminishing hourly rates and the reduced-price monthly contracts, both of which favor commuter parking and thereby reduce the impact on peak traffic. A more appropriate system would have increasing hourly rates and would eliminate monthly discounts.

The increased parking charges in central Singapore were believed to have had some adverse effect on business there, while the area license, because of the timing of the restricted period, was thought not to have had such an effect. Perhaps it would have been desirable to impose a surcharge for entering parking spaces during the morning peak period instead of raising the rates for all times of day.

1/ A Study of Supplementary Licensing, Greater London Council, London, 1974. - 265 -

No scheme of parking charges, of course, would apply charges to through traffic. Unless complemented by other methods such as a physical partition of the area as in G8teborg or a charge for crossing a dividing line (see below), high parking charges might deter people from going downtown while the road space they relinquished was taken over by an increase in through traffic.

Another class of traffic that would be immune from parking charges and in most cases from measures designed to impinge on through traffic is chauffeur-driven cars, which are important in some cities in less-developed countries. Taxis also would be missed, but whether they should or should not be charged is a matter of some debate.

Crossing a dividing line

Either to compensate for the main deficiency of parking rate policy, or to reduce traffic with as little effect as possible on trips to central destinations, it may be desired to apply restraints specifically to through traffic. For this purpose, a variant of the Bremen and Gbteborg schemes could be used. Instead of physically separating the different sectors of the city, a charge could be assessed for crossing the dividing line between one sector and another. A single line would suffice in some cities; others might require two or more lines. The means that could be used for collection would be the same as for pricing entry to a zone.

Other pricing bases

Electronic devices, once they become operational, will not be limited to substituting for manual methods of implementing the approaches discussed above. They will open up possibilities that would otherwise be infeasible.

Instead of one or two pricing zones, there might be dozens of "pricing points" with different prices, all varying by time of day, to tailor the price in each place more closely to the degree of potential congestion or the social marginal cost. Charges could reflect not only entry into a congested area but also the length of time spent, or distance traveled, in it. There could be specific charges for passing through critical intersections or for turning in directions that delay other traffic. There has even been a proposal for measuring the frequency and sharpness of accelerations and decelerations in a vehicle as an indication of the degree of congestion in which it is operating and to which it is therefore contributing, and charging accordingly.

Even disregarding the last, extreme proposal, there is a danger in any of these relatively complicated approaches that the motorist will not have a sufficiently clear and accurate understanding of the prices confronting him to be able to respond as he is expected to. Moreover, a system that is hard to understand is likely to be unpopular. One of the principles considered important by the Smeed Committee in England in its 1964 study of road pricing was: - 266 -

"Charges should be ascertainable in advance. As the object of road pricing is to influence the decision of people before they use congested roads, it is desirable that intending drivers should know the charges payable before making a journey. Any system that imposed heavy charges without giving prior warning would fail in its main purpose and would give rise to much dis- satisfaction." 1/

Thus, care should be taken to keep the scheme relatively simple, with prices known in advance.

DELINEATION OF A PRICING ZONE

Delineation of the zone for an entry pricing scheme will depend heavily on the size and shape of the city, its traffic flow patterns, the local geography, and the pattern of roads. Nevertheless, some general principles can be suggested.

The size of the zone -- aside from considerations of local geography and existing roads -- should be sufficient to include the area subject to congestion, not just at the present time but also in the future. It should also take into account the congestion that may be created outside the zone by diverting traffic that would otherwise go through. If the zone is too small, it may merely act as a local obstruction, causing more congestion around its periphery than it eliminates inside.

The larger the zone, the more people there will be whose destinations lie within it and who (if they live outside) will be affected, having to choose whether to pay the price or travel by a different mode or at a different time. However, a larger zone might include so many residents that a large fraction of trips would be internal and charging for entry would be ineffective.

Thus, the zone should be big enough to contain the important potential congestion area and to be more than a local obstacle, yet small enough to be primarily a central business area, excluding as much residential terri- tory as possible.

There should be a reasonably convenient route for those who wish to bypass the zone. That may, in practice, be one of the most important determinants of the zone size and its boundaries, although the bypass route does not have to be the boundary. As Figure 2.2 shows, Singapore's zone boundary and ring route are separated except at a few points.

If possible, the boundary should not divide neighborhoods or shopping areas and should not favor one retail store over its near competitor by

1/ Roth, Gabriel, Paying for Roads, Penguin Books, London, 1967. - 267 -

making customers enter the zone to reach one but not the other. Where a physical division already exists -- such as a river or a railway -- it may be desirable to use it as a boundary.

Other considerations in delineating the boundary are: 1/

(a) It is important to avert congestion in places where it would impede public transport.

(b) There must be opportunity for drivers to alter their routes if they find themselves approaching the zone boundary and do not wish to enter.

(c) If car parks or suitable locations for car parks exist near the proposed zone, the boundary should be drawn so as to leave them on the outside.

(d) The number of entry points should be kept small to minimize the policing requirements.

(e) If minor roads are closed rather than policed as entry routes, it should be done at points that will cause minimal disruption of local traffic.

(f) Where possible, major through routes should be left uninterrupted, to avoid impeding traffic that is not bound for the restricted zone. Thus, if a freeway passes through the area, its exits, rather than a point on the freeway, might be treated as entry points into the zone.

THE HOURS OF RESTRICTION

A pricing scheme for the purpose of reducing excess demand for road space and consequent congestion obviously does not have to be in effect 24 hours a day. The question is whether it should be in effect only for the morning commuting hours, or for one period in the morning and another in the afternoon, or during the whole working day. This depends upon the intended impact of the scheme and on local conditions such as normal business hours, the kinds of activity located in the area, whether it is customary to drive home for lunch, and so on.

In Singapore, the concern to avoid any negative effect on business dictated that restrictions should be focused on the morning commuting period. The beginning of the period was purposely set late enough (7:30 a.m.) to en- able many drivers to avoid the charge by going to work about half an hour earlier than before. Others avoided it by going after the end of the period.

1/ The list is adapted from Lim Leong Geok, op. cit. - 268 -

While the peak flow was suppressed, two new peaks were created. Although these peaks were lower and of shorter duration than the original one, the second one (at 9:30 a.m.) was still higher than desired by the officials, and they eliminated it by extending the restricted period to 10:15 a.m. It was believed that this extension would not affect retail business, since most stores did not open until 10:00 a.m. or 10:30 a.m., but that any further extension would do so. Opinions expressed by businessmen generally agreed with this judgment.

Restricted hours in the afternoon were also considered. Had the zone included only non-consumer-oriented offices and industries, afternoon restraints might have been adopted. They were not instituted for fear of their possible effects on the business of stores that stayed open in the evening, restaurants, and entertainment centers, all of which were vulnerable to diversion of their potential customers to places outside the zone.

In a city where going home for lunch is prevalent practice, charging inbound morning traffic would be far from an adequate approach. People might use car pools and buses for both of their morning trips, and then drive separate cars after lunch. Thus it would be necessary at least to consider a second period at the time of the after-lunch inbound trip or to charge for outbound trips in the afternoon.

It should be noted that reducing peak flow levels is not the sole purpose of a pricing policy. Reducing inflows is also a method of reducing the number of commuters' cars moving around in the zone later in the day or occupying parking spaces that might otherwise be used by shoppers.

SETTING THE PRICE

When the Singapore officials set the price for the area license, there was no previous experience to guide them. Now the Singapore experience is on the record and could be used as a rough basis for choosing a price by relating it to the other costs of making trips, the mean income level, and the desired reduction in traffic. However, in other cities the demand relation may differ because of differences in the quality and availability of alternative transport, the distances involved, and a variety of less tangible factors.

Whatever price is set must be regarded as an experiment, and the authorities must be prepared either to change it after they observe its effect or to accept the resulting reduction in traffic congestion even if it is different from what they had intended.

If the effect is greater than intended, it may be that the price, while too high for the immediate conditions, is right for a future situation when there will be higher incomes and more cars. In that case it may be best to leave it alone rather than reduce it and have to raise it again. If the effect is too little, the price should probably be raised before people get used to it; it is important that the pricing scheme have some shock effect - 269 -

and make people re-evaluate their travel habits rather than merely adjust their budgets to absorb a small change in costs. (The gradual increases in price of gasoline in the United States have not had much impact on automobile use.) However, it might be effective to increase charges gradually if it was announced at the beginning that charges would be increased at regular intervals until a specified reduction in congestion had been achieved. MULTIPLE ZONES AND VARIABLE PRICES

The discussions above concerning zone boundaries, hours, and prices assumed that there would be a single zone with a fixed price for entry (or possibly for presence within the zone) during prescribed hours. In order to modify the "edge effect" of a single zone, or to intensify the effect in a critical area, it might be worth while in some cases to have two or even more zones with different prices or with different hours. While more complex arrangements would be more difficult to administer with any except electronic methods, windshield stickers could be used for a two-zone system in which the same license was valid for both zones but was required for different time periods -- for example, during the morning commuting period for an outer zone and throughout the day for an inner zone. Slightly more difficult would be a two-price scheme, which would require stickers of different shapes or colors for the two zones.

"Edge effects" occur not only at geographical boundaries but also at the beginning and end of the restricted hours. At zone entry points in Singapore there was a quite noticeable rush to get into the zone before the lights on the "Restricted Zone" signs were turned on at 7:30 a.m., and one could often see drivers, who had clearly intended to enter but arrived too late, turning onto the ring road instead. To avoid such effects, Columbia University's Professor William Vickrey -- one of the earliest advocates of road pricing and still its most ardent spokesman in the United States -- has suggested that the tolls on bridges leading into New York City be varied continuously, minute by minute, charging a low price in the very early morning, gradually increasing it to a maximum at what is now the time of peak flow and lowering prices gradually after that time in such a way as to flatten the peak and spread demand uniformly over several hours. While this proposal seems extreme and might entail problems about drivers' awareness of the charges to be encountered, it illustrates an approach to the ideal of achieving smoothness at the "edges" of a pricing scheme.

While it is conceivable that variable prices could be charged with a toll collection system, windshield sticker methods could not be used unless perhaps for only two prices in different periods, using different colors or shapes to show what price was paid in advance for the sticker. Even this might add just enough complication to give serious difficulties in administration and enforcement. It is recommended that such refinements as different prices for different entry times or even different prices for different zones be considered only after a simple one-price scheme has been successfully implemented. - 270 -

EXEMPTIONS AND DIFFERENT PRICES

Discrimination among types of vehicles is feasible, both in terms differ- of exempting certain types from the license requirement and charging ent prices. This has been demonstrated in Singapore, where buses, commercial and where vehicles, emergency vehicles, motorcycles, and scooters are exempt three different prices apply to business-owned passenger cars, privately- in owned passenger cars, and taxis. There are hazards and difficulties making such distinctions, however, and there should be compelling reasons for any departure from charging the same price to all without exception.

The exemptions in Singapore were based on the desire to avoid imposing any new expense on bus operators or on business and on the difficulty of reading license numbers of moving motorcycles or scooters for enforcement that purposes. If commercial vehicles are to be exempt, it is essential seems they be readily distinguishable from vehicles subject to charge. This hard to to have been no problem in Singapore, but in some cities in might be van used distinguish a delivery van (exempt as a commercial vehicle) from a no as a personal car. To avoid such problems, it might be better to have exemption for commercial vehicles.

The car-pool exemption in Singapore requires that the police, before recording a car's license number for a violation, check the number would be of occupants. This somewhat complicates the job of policing (and a car impractical with electronic methods). Since having four people in exemp- automatically reduces the cost per person, it has been argued that to tions for car pools are not necessary. However, the exemption serves to coun- reinforce the concept of encouraging high-occupany-vehicle use and to drive teract objections that an area license scheme only allows the rich in down town. While these arguments may not be relevant in some cities, others they may be important for public acceptance of the scheme.

Charging three different prices for taxis, privately-owned cars, and company cars required six easily distinguishable shapes of windshield order to stickers, three for monthly licenses and three for daily ones. In them, it verify that company-owned cars have paid the higher fee required of but also is necessary not only to check the shape of the windshield sticker to have a way of distinguishing them from privately-owned cars. For this were purpose, before the scheme was implemented, all company-owned cars assigned registration numbers prefaced with the letter "Q".

This illustrates some of the ramifications of what appear to be rather simple distinctions for pricing and exemption purposes. Granting (start- special prices or exemptions to particular categories of car owners re- ing, probably, with government officials) is a temptation to be firmly will sisted. Once that door has been opened, there is no end of groups that voice their need for special treatment, and in any exempt category there will be many individuals for whom the rationale for the exemption is not valid. Exemptions of this sort are also an invitation for corruption, as bet- someone has to decide on the qualifications of individuals. It is far should ter to have no exemptions -- especially for government officials, who set a proper example for others. - 271 -

COMPLEMENTARY MEASURES

Charging a price for the use of scarce road space is not a policy to be considered in isolation. It should be part of an overall transportation policy for the city. If it is to be adopted, it will probably need to be complemented by other measures in the realms of traffic management and public transport.

PROVISION FOR CHANGES IN TRAFFIC PATTERNS

For drivers wishing to by-pass the restricted zone, it may be necessary to upgrade or widen some roads along a circumferential route, modify intersections to handle increased traffic, and perhaps convert some parallel two-way roads into one-way pairs. Improvements of this sort were done along the ring route before inauguration of the license scheme in Singapore, but even so, a few spots of local congestion appeared when the scheme took effect, and some further work had to be done to remove the bottlenecks. Traffic lights around the periphery will probably have to be reset to reduce the time allocated to radial flow and give a greater share of "green time" to the circumferential flow.

If fringe car-parks are to be established, some changes in the adjacent roads may be needed to provide adequate approaches. Where circumferential automobile traffic may interfere with buses, there may be a need for reserved lanes, priority at intersections, or other means of avoiding delay to the buses.

ALTERNATIVE TRANSPORT FACILITIES

If people are to be induced to give up driving cars into the city, they must have other means of getting there. Car pooling will take care of some of them. Some may switch to motorcycles, bicycles, or walking. In most cities, however, the main alternative to driving is public transport, usually in the form of buses or buses and jitneys. If public transport capacity were insufficient to absorb the increased load, public reaction could force the pricing scheme to be abandoned.

Long before implementation of the Area License Scheme, Singapore officials undertook several measures to expand capacity of the standard bus service -- using school and private buses to carry part of the peak load, improving maintenance to keep more buses operating, and buying additional buses. In addition, believing that there should be higher grades of service (at correspondingly higher prices) for those who were giving up the comfort and convenience of automobiles, they provided the park-and-ride service, an express bus service, and an air-conditioned service using sightseeing coaches during commuting hours. Rather surprisingly, none of these services drew very much patronage. It would probably be wrong to conclude from this experience that such services would not be used in another city, where different trip lengths and different public attitudes might yield different results. Perhaps a market survey would help forecast demand. In any case, contingency planning and experimentation are advisable. - 272 -

ESSENTIAL FACTORS IN IMPLEMENTING THE SCHEME

Successful implementation of the Area License Scheme in Singapore has led some people to proclaim that it can now be done anywhere, while others say that Singapore is a unique case and proves nothing. In a sense, both are partly right, and totally wrong. It has been demonstrated that an area license scheme can be successfully implemented under some conditions. The important thing is to try to identify what the conditions were, which ones were essential to success, and which ones could be different without making success impossible.

The following factors, all of which were present in Singapore, are in our judgement essential to the successful implementation of any pricing scheme.

(a) Competent management, with an organizational structure that fosters comprehensive policy-making and planning for all aspects of transport in the metropolitan area, including traffic management, traffic policy, public transport services, and motor vehicle regulation.

(b) Carefully worked-out, detailed provisions for issuing licenses, erecting signs on the approaches to the restricted zone, enforcing the rules, and handling all administrative details.

(c) Good design of the scheme, including upgrading the circumferential route, expanding public transport facilities, and laying out the boundary.

(d) Advance education of the public as to the reasons for the scheme, its expected benefits, both short-run and long-run, exactly how it will work and what choices people can make. (In Singapore, this process started a year before implementation of the scheme and involved publication of an explanatory booklet 1/ and expression in speeches and newspaper articles of opposing views as well as those of the government.)

(e) Pragmatism, as exemplified in Singapore by an effective program of monitoring results and quickly making traffic management changes or modifying the rules of the scheme to overcome observed problems.

1/ Appendix 1. - 273 -

OTHER CONTRIBUTING FACTORS

Some other factors undoubtedly made successful implementation of the scheme in Singapore easier but, in our judgement, are not absolutely necessary for success elsewhere.

(a) The fact that automobiles were used by only a minority of downtown commuters, even before the Area License Scheme was instituted. This helped, not just because there were fewer potential opponents of the scheme, but also because it implied a greater capacity for public transport to absorb those who changed their travel mode.

(b) Cultural-social attitudes favorable to compliance with regulations. Singaporeans are disposed to believe the government acts in the general social interest, and to accept rules and costs imposed on them. This does not mean that no effort was necessary to enlist popular support. (See item (d) above.) However, in a society where protest is the typical response to government actions, an even greater effort to gain acceptance would be necessary.

(c) The relative ease of planning and decision making in a city-state, with one level of government dealing with one geographical area. In a multi-jurisdictional setting, it would be necessary to have a special agency with authority to establish and administer the necessary measures for the area as a whole. While it is often awkward and time-consuming to get agreement on essential decisions among the members representing different jurisdictions, it is not impossible. That is illustrated by the work of the Metropolitan Washington Council of Governments, which represents the District of Columbia, together with the six counties and eight cities that make up the suburbs in two states.

(d) Laws that made enforcement simple. While it was indeed convenient to be able to collect fines from car-owners by observing their license numbers and mailing citations to their homes, nevertheless other procedures would be feasible (and perhaps less open to possible error.) With the reduced flow on the streets just inside Singa- pore's Restricted Zone, there would have been plenty of space to stop all violators and cite them on the spot. This would have required a larger number of police at each entry, but would have been perfectly practical. - 274 -

(e) The relative isolation of the region from outside traffic. Having significant numbers of vehicles from other jurisdictions entering the restricted zone would require more convenient provisions for issuance of one-day licenses and a different approach to the treatment of violations. The use of toll booths for those without monthly stickers would be a convenient substitute for one-day licenses and would eliminate the problem of violations by strangers who did not know about the system.

Considering both the essential factors and those that were helpful but not absolutely necessary, the possibility of successfully implementing area pricing schemes in other cities has to be judged on a case-by-case basis. Given a conviction that such a scheme would be worthwhile, appropriate organizational and administrative arrangements, and competent people in charge, it should be possible to design and implement a workable scheme within the constraints of local conditions in any of a wide variety of cities. APPENDIX ONE

A PLAN FOR THE RELIEF _, OF

TRAFFIC CONGESTION

IN THE CITY - 276 -

F1- __ i -__~~~~~~~~~~~~~~~~-

i __

b ~~~~~~~~~IEEII _~1F r, . INTRODUCTION

The motor car has long been recognised as a comfortable and convenient means of transport and this has led to its widespread use Cities have come to realise, however, that this widespread use has itself frustrated and defeated the mobility and convenience of the motor car - the very reason for its popularity Each additional car coming on the road adds to the congestion for all other motor vehicles already on it This is particularly so during peak hours when large numbers of motorists travel along the same roads to the same places at the same time Daily traffic congestion results in delay and frustration to motorists, bus commuters, goods and emergency vehicles, and poses danger to pedestrians and other road users It also causes deterioration of the "Long wait for buses ...... environment through noise, air pollution and visual blight "To cope with increasing traffic, new roads were built, existing roads A" .S.C,- * ...... "widened and various traffic management schemes were implemented _____ All these efforts, involving considerable sums of money and large scale often property acquisition and demolition, barely kept pace with the rapid increase in traffic In Singapore the rapid development of our economy over past the decade has brought more and more families into the car-owning income range In 1968 the rate of increase in cars was 8% over the previous year This had accelerated to 1 5% per year at the MI 116 end of 1973 The average annual growth rate over the last five years was about 10% If this average annual growth rate of 10% is maintained, there will be more than half a million private cars in Singapore in the next 14 years, a four-fold increase over the 1973 car population of 135,000 If the 1973 growth rate of 15% is maintained, the 500,000 car population will be reached much earlier This tremendous would result in a increase in road traffic and its attendant problems cor______It is estimated that to keep this amount of traffic moving, an extensive road development programme involving the construction of an elaborate system of expressways, grade-separated interchanges and "Delays and frustration ...... "other major roads over the whole island together with large scale " a,A f XA , I ...... widening of roads in the city area would be needed Singapore ' -~- t R - z . . cannot afford to continue to allocate scarce and valuable land to build unlimited miles of roads to keep pace with this uncontrolled increase in traffic, neither can we afford to allow the prolonged and frequent snarling of traffic which is surely and steadily choking up our roads and our economic life. While the many advantages and benefits offered by the motor car should not be unnecessarily withheld from those who can afford to own one, the uncontrolled use of the car in the city must 2,U K I%_' - -\ fi_ be discouraged Specifically, the daily convergence of car commuters into the central area during peak hours, which severely strains the available road capacities and which is largely responsible for all the undesirable side effects of motor traffic, must be controlled

_; -- o ! The majority of the motorists who drive during the morning and

*1 , - \k _evening peak hours are people commuting to and from work Many of these drive into the central area in the morning, park all day, "These cars carry 240 persons ...... then drive home again in the evening It is fair to state that, given

" ME. fi .t , i 240 f' a comfortable alternative mode of travel, it will not be essential for many of such motorists to commute into the central area by car

in_ r -j _ S ,Theaverage car occupancy rate of motorists commuting to work in Singapore is 1 3 persons per car A large bus which is licensed to carry 70 to 80 passengers, uses the equivalent road space of only -,__->' - - - athree passenger cars The benefits to be gained by inducing a substantial number of car commuters to travel into the central area by bus are obvious p . - 7 , 1

Xr. ;* r 'PLAN FOR RELIEF OF TRAFFIC CONGESTION

A scheme is being proposed to relief traffic congestion on our roads during peak hours It aims to encourage commuters to travel by bus into the city during peak hours or by group travel through car pooling To achieve the desired changes In travel habits, some restraint measures are necessary

...... 5so do these three buses" 11...... it -- e.F + XL se, - i RATIONALE OF THE PLAN As the capacity of existing bus services cannot meet the additional demands of diverted motorists, commuters who presently drive to In conceiving a scheme of restraints on private car travel into the work will be encouraged to park their vehicles at the periphery parks city, the primary aim must be to limit the non-essential categories of the defined restricted zone where strategically located car of motorists, viz those who do not need their cars during the course will be provided They will then transfer onto a new, comfortable to their of their work and those who can change their times of travel to and reliable shuttle bus service which will transport them off-peak hours so as to provide relief to peak hour traffic congestion. various destinations within the restricted zone will be Careful attention must be given to minimising disruption to the Within the restricted zone in the city, car parking charges economic activity of the city Thus, activities which contribute to raised to discourage long-term parking of paying a supplementary the vitality of the central area, such as shopping, the transaction of Thus, motorists will be given the choice business, the provision of services etc, must be allowed to continue licence fee plus higher parking charges for the privilege of driving cecfeplshgepakgcagsfrtenveeofdvg with minimal hindrance into the restricted zone during peak hours or opt for the alternative shuttle An attractive, comfortable and reliable bus service must be provided of driving to the fringe area car parks, and then taking the for those who opt not to drive into the city bus services into the city at considerably less cost The scheme also provides an element of choice Those who choose to continue driving into the city during peak hours may do so, THE RESTRICTED ZONE but at a price The restricted zone covers areas of severe traffic congestion and areas of concentrated employment which attracts the greatest SUMMARY OF THE SCHEME number of work commuters

A careful study has been made of various possible methods of It excludes, where possible, major by-pass routes to minimise restraining private car travel into the central area Some of these disruption to traffic not destined for the congested parts of the city measures have been implemented in other cities, while others have The size of the restricted zone is kept to the minimum to minimise time is been subject of considerable study but have yet to be tried out The undue disruption to the motoring public while at the same conclusion of the study is that a supplementary licence scheme large enough to achieve the objective of reducing traffic congestion coupled with increased parking charges would be the most suitable The defmed restricted zone Is shown on Figure A measures for Singapore The scheme recommends that a special licence will be required for specified categories of vehicles to be driven into a defined area FRINGE AREA CAR PARKS & SHUTTLE BUS of the city during a specified time period As an incentive to car SERVICES pooling, vehicles with four or more persons would be exempted from this licence requirement Other categories of vehicles such as Fringe area car parks will be strategically located to provide con- public transport vehicles, emergency vehicles, public service venient parking spaces for motorists approaching the restricted vehicles, taxis etc will also be exempted zone from all directions The selected sites are shown on Figure B

5 4 The parking charges would be a nominal $10/- per month, except VEHICLES AFFECTED BY THE SCHEME for a few multi-storey car parks located immediately adjacent to the restricted zone boundary where the parking charge will be The scheme is designed to discourage commuters who drive alone $30/- per month or with one or two passengers in their cars during peak hours Hence all passenger vehicles, including station wagons and mini-buses The car parks will be landscaped and trees will be planted to provide with seating capacities of less than 7 persons, will be subjected to shade to the parked cars Initially, about 10,000 parking spaces will the supplementary licence, irrespective of the type of licence (eg be provided private car licence or goods vehicle licence) To encourage car pooling, vehicles carrying 4 or more persons will A new shuttle bus service will be operated for those who opt for be exempted the "park-and-ride" scheme The buses will only carry seated passengers to ensure that a reasonably comfortable service, which The following categories of vehicles will be specifically exempted is free from over-crowding, is provided. Direct bus connections will and be allowed free access into the restricted zone at all times - be provided from the fringe area car parks to the major destination a Public transport vehicles, including regular scheduled buses areas within the Restricted Zone (See Figure B), viz (a) Area around school buses, private buses with seating capacity of 7 or more Shenton Way, (b) The South Bridge Rd/New Bridge Rd area, persons and taxis (c) Orchard Rd The buses will run non-stop until they enter the b Emergency vehicles, eg ambulances, fire engines, police and restricted zone so as to minimise travel times On the return journeys military vehicles during the evening peak period, some buses will start from inter- 00 mediate locations along the bus routes so that passengers who do HOURS OF RESTRI not start their journeys from the city terminals will still be assured STRICTION of getting seats in the buses and not face the frustration of having To minimise inconvenience to the public, the hours of restriction to wait for long periods at the bus stops will, initially, be for 2 hours only - 7 30 am to 9 30 am except for Sundays and public holidays During off-peak hours, frequent shuttle bus services between major activity centres within the restricted zone will be provided by It Is hoped that the motorng pubic will respond favourably to the this new fleet of comfortable buses as supplementary services to scheme and take full advantage of the park-and-ride service provided taxis If such is the case, it will be possible to keep the hours of restriction to the minimum The response will be closely monitored and the hours of restriction will be lengthened if it is found to be necessary A flat fare of 50 cents per trip will be charged This fare is con- t os siderably lower than taxi fares and only slightly higher than fares to do so on the regular scheduled bus services It is felt that the slightly higher fare is commensurate with the greater level of comfort and FEE FOR SUPPLEMENTARY LICENCE reliability of the service provided Season tickets at $20/- per month will be available for the convenience of regular commuters Those who choose to continue to drive into the restricted zone will have to pay a licence fee of $60/- per month Licences will be issued for periods of six months by the Registrar of Vehicles

6 7 For the convenience of those who occasionally need to drive into X Qi the restricted zone during hours of restriction, daily licences can' be purchased at $3/- at the fringe area car parks Monthly licence holders will be issued with special identification plates which have to be fixed on the front and rear of their vehicles These plates will be in addition to a licence disc which will have to be fixed on the bottom left hand corner of the front windscreen of the car

Daily licences will be in the form of licence discs only which will have to be fixed on the left hand corner of the windscreen i D alana ,

PARKING WITHIN THE RESTRICTED ZONE

All car parks within the restricted zone will be required to charge a minimum of $60/- per month for long-term parking spaces Hourly parking charges will vary according to location as follows- a Core Area of Restricted Zone- 50 cents for the first hour, OD,/. $1- for the second hour, and

$1- for each subsequent half Sc. hour b Rest of Restricted Zone - 50 cents for the first hour, and 50 cents for each subsequent half hour TEIOK AYER The main aim is to discourage long-term parking while maintaining reasonably low rates for short-term parkers

OPERATION & ENFORCEMENT LEGEND R1OI The entry points along the boundary of the restricted zone will RESTRICTED ZONE be prominently identified by signs Additional warning signs will onVi la also be placed along the approaches to the zone boundaries The Ll CORE OF RESTRICTED ZONE check points have been carefully selected to allow unwary motorists 5 / M" PM 9" 4' PA the opportunity to turn around with little difficulty / A A A

8 RESTRICTED ZONE Mi 1 p-

0 000 2000 3000 400FOT fJ i =2g / ~~~~~~~~LEGEND O -='

SHUTTLE- BUS ROUTE ~~~~~~~~~~~XrESyA\ Igt b- 1S@ M v R\\ - k 42 cr + * 6ia-

AREA %0\ ,," v 2 | | ~~~~~~~SHENTON WAY

BRIDGE ROAD AREA t>,'.> / SS //v7 | | ~~~~NEW BRIDGE/SOUTH

Je 4* NJ1- 0- i Yn/ wtQl iZA t <2// ~~~~~Ai/Uth ^.tR A r t/ Xstor rsw > \ \1ORCHARD ROAD AREA

l jt t q ta a ~~~~~FRINGECAR PARKS

':s $t1t /u_ ;te / ~FIGURE B 11B

( = ~~~~~~~~FRINGE CAR PARKS AND SHUTTLE BUS SERVICEI

2000 300 ...... I I~~~~~~~~~~~~~~~~~~o ^ b:' t ffi t z e $ 4 *~~~SC.LfEEi f It would be an offence for non-licensed or non-exempt vehicles to be dropped at the fringe area car parks to catch the connecting to enter the restricted zone during the hours of restriction Vehicles, shuttle bus services into the city and be picked up again on their within the zone will be free to move within or out of the zone at return They could thus, probably dispense with the expense of all times running a second or third car in the family The luckier ones who live within walking distances of the fringe area car parks would Residents within the restricted zone will thus be free to move within likewise benefit from the scheme In all cases, there will be cost the restricted zone or to leave the zone without the supplementary savings from not having to drive into the central area and paying licence at all hours They will, however, be subjected to the same high fees for parkmg and for the area hcence restrictions if they should drive Into the zone during the hours of restriction Motorists who choose to drive into the restricted zone during peak hours will have to pay for the supplementary licence and higher parking charges However, they too will benefit from the scheme EFFECTS, BENEFITS AND CONCLUSIONS through driving in less congested streets

At the end of 1973 approximately 28,000 private cars entered the Not all of the additional revenue through higher parking fees within proposed restricted zone during the peak 2 hours each morning the restricted zone will accrue to the owners of garages Car park Approximately 14,000 of them carry commuters on their way to licence fees will be raised so that some of this additional revenue work The proposed scheme aims to reduce the number of such can be channelled to subsidise the fringe area car parks In this way, work commuters by about half In addition, it is expected that the those who prefer to drive into the city during peak hours will subsidise number of non-work car commuter trips during the morning peak those who decide to use the park-and-ride service hours will also be reduced By inducing some motorists to park at the periphery of the city in Residents within the restricted zone will only be affected if they the morning, there should be a corresponding decrease in the should leave the zone and wish to re-enter the zone during the evening peak hour traffic Unless it is necessary, there will not be restriction hours an evening peak restriction There is also some flexibility to extend the morning hours of restrictions should the need arise School children who attend schools located within the restricted zone will not be adversely affected if they should travel by car as Business and other activities within the restricted zone should they have to reach school before 7 30 am - the start of the restriction not be adversely affected In fact, businessmen, shoppers and hours. visitors to the area will have less difficulty in finding short-term parking spaces The environment would also improve with the Large commercial vehicles have been left out of the restraints reduction of traffic measures at this stage to minimise disruption of their operations Motorists, who are the main targets of the scheme, will find that Consideration will be given to restrict this category of vehicles if the scheme will also bring some benefits to them Those who opt their numbers on the streets pose serious traffic problems for the park-and-ride scheme, will find the fringe area car parks convenient and the shuttle bus service comfortable and reliable It is hoped that the motoring public will respond to the scheme They will be freed of the burden and frustration of crawling through which is designed to bring an end to their daily frustration and to the traffic jams in the city Some motorists may be able to arrange improve our living, shopping and working environment

12 13 Detailed planning and other arrangements to implement the scheme ff * Pi 3z A N 1 A - are in progress Comments and suggestions are invited from the public who should send all correspondence to - The Secretary -W Road Transport Action Committee 6th Floor, National Development Building o J 4 Bg£ Maxwell Road Singapore 2 i ±JlII[b

El 3 M A T-9i ±' l 135XDf9t - 'tNJA f O 3ftiib

" ) on o

14 15 RECENT PAPERS IN THIS SERIES:

No. TITLE OF PAPER AUTHOR

254 Valorization Charges as a Method for Financing W. Doebele (consultant) Urban Public Works: The Example of Bogota, 0. Grimes Colombia

255 The Employment Impact of Industrial Investment: J. Stern (consultant) A Preliminary Report

256 A "Stages" Approach to Comparative Advantage B. Balassa

257 Measuring the Performance of Family Planning P.S. Mohapatra Programs

258 Multi-Level Programming & Development Policy W. Candler, R. Norton

259 Alternative Concepts of Marginal Cost for J. Saunders, J. Warford Public Utility Pricing: Problems of P. Mann (consultant) Application in the Water Supply Sector

260 Meeting Basic Needs in Malaysia: A Summary J. Meerman of Findings

261 A Model for Estimating the Effects of Credit T. Husain, R. Inman Pricing on Farm Level Employment and Income Distribution

262 Commodity Price Stabilization and the E. Brook, E. Grilli Developing Countries: The Problem Choice J. Waelbroeck

263 Industrial Policy and Development In Korea L. Westphal, Kwang Suk Lim

264 The Incidence of Urban Property Taxation in J.F. Linn Developing Countries: A Theoretical and Empirical Analysis Applied to Colombia

265 India's Population Policy: History and Future R. Gulhati

266 Radio for Education and Development: P. Spain, D. Jamison Case Studies (Vols. I & II) E. McAnany

267 Food Insecurity: Magnitude and Remedies S. Reutlinger

268 Basic Education and Income Inequality in J. Jallade Brazil: The Long-Term View

269 A Planning Study of the Fertilizer Sector in A. Choksi, A. Meeraus Egypt A. Stoutjesdijk No. TITLE OF PAPER AUTHOR

270 Economic Fluctuations and Speed of Urbanization: B. Renaud A Case Study of Korea 1955-1975

271 The Nutritional and Economic Implications of L. Latham, M. Latham Ascaris Infection in Kenya S. Basta

272 A System of Monitoring and Evaluation of M. Cernea, B. Tepping Agricultural Projects

273 The Measurement of Poverty Across Space: V. Thomas The Case of Peru

274 Economic Growth, Foreign Loans and Debt G. Feder Servicing Capacity of Developing Countries

275 Land Reform in Latin America: Bolivia, Chile S. Eckstein, G. Donald Mexico, Peru and Venezuela D. Horton, T. Carrol (consultants)

276 A Model of Agricultural Production and Trade R. Norton in Central America C. Cappi, L. Fletcher, C. Pomareda, M. Wainer, (consultants)

277 The Impact of Agricultural Price Policies on M. Osterrieth, Demand and Supply, Incomes, and Imports: An E. Verrydt, J. Waelbroe( Experimental Model for South Asia (consultants)

278 Labor Market Segmentation and the Determination D. Mazumdar of Earnings: A Case Study M. Ahmed

279 India - Occasional Papers M. Ahluwalia, J. Wall S. Reutlinger, M. Wolf R. Cassen (consultant)

280 Educational Effects of Class Size W.D. Haddad