Transport and Road Research Laboratory 0 0 A COMPARISON OF BUS OPERATIONS IN CITIES OF DEVELOPED AND DEVELOPING COUNTRIES by Dr G D Jacobs, D A C Maunder and P R Fouracre Overseas Unit Transport and Road Research Laboratory Crowthorne Berkshire 1 9A1 Reprinted from TRAFFIC ENGINEERING & CONTROL June 1979 A comparison of bus operations in cities of developed and developing countries by G. D. Jacobs, PhD, MCIT, MIHE, D. A. C. Maunder, BA, MCIT, and P. R. Fouraicre, BSc, MIHE Overseas Unit, Transport and Road Research Laboratory Introduction. Over the last 20 years something like 600m. people have been born in or have moved to urban areas with populations of more than 100000 (now generally signified internationally by the word 'city'). Over the next 25 years it has been estimated' that they will be joined by a further I 500 in-more than ,,&If the world's entire population as it stood in early 1960's. Unfortunately this expected cTrease is. unlikely to spread evenly through the world. Some three-quarters of these newcomers will join the teeming masses of those Third World cities that are at present experiencing great difficulty in feeding, hous- ing and transporting the millions who already live there.In many cities in developing countries, the rapid rise in population, coupled with limited financial resources available for investment in urban infrastructure, has produced severe transport problems. In the near future, an improved public transport system is likely to be the only solution for reducing traffic congestion and chaos in most Third World cities 'since it is unlikely that the communities concerned can (or even should) afford to build the road network needed to accommodate unrestrained travel by private car'". Indeed, a number of cities in South-East Asia have taken a lead in the possible introduction of area restraint systems aimed at reducing private car usage within the city centres. Although one or two cities in developing countries are building rail mass transit systems, the vast majority will still be relying for many years to come on road-based systems. Public transport in cities of the developing world comprises more than just the traditional bus and train of developed cities. Thus shared taxis, motorcycle, scooter, cycle and hand rickshaws, and even horse-drawn vehicles, all play an important role in personal mobility. Previous articles'.' explained the role and function of 'paratransit' systems in cities in develo ping countries; this article makes a broad review of the demand for public transport in cities in the Third World and some of the factors that might be affecting demand, comparing and contrasting them with con- ditions in the developed world. It also outlines some of the special problems faced by public transport operators in developing countries Table I. Average annual percentage changes over the period 1 964-75 in the number of buses and passengers in some cities in developing countries and the U. K. Passengertrips per Average Trips per daily person City tion on road a n num per b-us by bus Ahmedabad (India) + 5.5 + 2.1 + 7.3 + 0.4 + 1.1 Barbados (island) + 2.0 + 1-6 + 5.0 + 2.7 + 2.4 Blantyre + 9.7 +12.3 +10.6 -0.7 + 0.5 Bombay + 5.5 + 4.1 + 7.0 + 2.0 + 0.9 Calcutta +18.4 + 1.5 + 5.0 + 3.0 -4.4 Delhi + 4.8 +13.1 +22.7 + 4.0 +11.6 Istanbul + 6.1 NA + 8.0 NA + 1.3 Mauritius' (island) -3.4 + 9.8 +14.8 + 3.4 +21.0 Mombasa + 7.3 . + 9.6 +15.4 + 2.8 + 4.5 Nairobi + 8.1 + 9.0 +16.6 + 3.8 + 4.5 Nicosia + 6.4 -0.4 + 9.3 +10.4 + 1.7 Surabaya- + 3.0 + 6.7 16.7 + 7.7 +11.2 U. K. London Transport Executive -1.8 -2.9 -1.3 Passenger Transport Executive -0.7 -2.7 -2.2 Municipal Operators + 0.4 -2.5 -2.7 '1970-75--1970-76 NA: not available 0 and the economic climate under which they operate. Urban public transport usage in developed and developing countries In the U.K. the use of stage bus services hat' been declining since about 1950. A major cause of this reversal in patronage trends has been identified 4 as the rapid increase in car ownership which has taken place over the last three decades. Although car ownership has also increased in developing countries over the last 30 years, it is still at a relatively low level and there is a great reliance on public transport. Percentage changes in the number of buses and passengers over this period 1964-75 for some 12 cities, most of which would be regarded as being in the 'developing world', are given in Table I. In. almost all the Third World cities there were considerable increases in the number of buses in use and passengers carried. (In Nicosia, the only city to show a decrease in the number of buses, the public transport sector was subject to a comprehensive modernisation programme leading to the scrapping of a number of old vehicles.) The number of passenger trips made on public transport increased at a faster rate than the population growth rates in all cities except Calcutta where unusual changes took place during the period 1964~7 5-in particular, the large influx of refugees from Bangladesh and alto rural migrants with very low incomes who are un- likely to be able to afford any form of public transport. In comparison, in the U.K. there was an approximate 3 percent reduction in the number of passenger trips per annumi. Apart fomn Blantyre, Malawi, buses were heing used more intensively in 1975 than in 1964. In Surabaya, Indonesia, for example, there was an increase of almost 8 per cent per annum in the average number of passengers per bus. In other cities, increases of 4 per cent were common. In the U.K. there were again consistent decreases for all three transport organisations for which data were available. The number or trips made per hc;,d of' population (i.e. the trip rate) increased in all Third World cities except Calcutta. where there was an average reduction of 4.4 per cent per annum. In this city, in 1968, shared tax( services were sanctioned bY the Regional Transport Authority and there might well have been a movement from conventional public transport towards unconventional forms such as shared taxis, scooter rickshaws, etc. I t might well be that in a number' of developing countries a modest increase in real incomes 1 Averageflumber Popula- of buses q coupled (in some cities) with improved services has led to .increased public transport usage. whereas it has been shown that in the U.K. a more substantial increase in real incomes has led to increased car ownership which, together with a decrease in service levels, produced a decrease in public transport usage. In order to compare and contrast some of the factors affecting public transport oper- ations in cities in both developed and developing countries, basic data for 139 bus companies operating in cities throughout the world were obtained from the UITP Handbook for Urban Transport'. For the year 1974, data were obtained on such socio- economic characteristics of each city as population, area and car registration, together with details of each bus company operating in the city, e.g. fleet size, staff employed, routes operated, system length, passengers carried and kilometres operated. The sample of 139 cities were grouped into 97 'developed' and 42 'developing', based on a gross national product per capita for the countries in which the cities were located. A value of U.S. $750 was taken as the criterion; above this value, countries and hence cities were classified as 'developed'. belowv this value, as 'developing'. Following the method developed by Legris' and Lesley' a number of relationships between public transport usage and supply and the. physicat and socio-economic parameters of the cities described above were derived using regression analysis. An attempt was made to use multiple regression analysis to determine how the various socio-economic parameters, acting together, were related to the number o f passenger trips per annum. Unfortunately, )ecause the data sets were inadequate, the analysis was statistically invalid. Simple regression analysis was, therefore, used to describe relationships between public trans- port usage and socio-economic characteristics of the cities and bus companies. Two dependent variables were used: (i) the total number of passenger trips per annum; and (ii) the average number of trips per person per annum. It should perhaps he emphasised that the object of the analysis was not to derive a milodel' for public transport trip-making in Third World cities: it was simply to obtain a clearer understanding of some of the factors that might affect bus usage in these cities. In both developed and developing cities there was a close correlation between the population of the cities and the number of bus passenger trips per annumn. From Fig I it can he seen that, for a city of givcn population, the number of passengers carried in both de- veloped and developing cities is. on average, very, similar indeed. This is a surprising result, since with low car ownership in Third World cities tisage of pu hIic t ranissport might he thsiughi io he cosrrcspoindingly gi caicr ilhan lin cities of Furispe and Nort I Amecrica with simlilar popuilation. Five points can p.rsisbe mlade in eIplainling this resuilt: (11 MIan pi eopic isl des eloping cou istries Irisel hs svipblic iraiispori niodes otlser than Alie Coliventiiioiial bIsi'. e.g. shaired taxis. iiooiecyle atid hiaind rickshawsk.. 121 l~eeautic. in mnany Third World countries., wages are low and large numbers are unemployed, people cannot afford to travel by bus even though the fares are low (see later). (3) There is probably sila lack of supply in developing countries which would be reflected inl the number of' passengers 'carried. 1400 -1200 c 1000 CL 6. 8 00 C0 ' 400 C. 200 0 0 0 1000 2000 3000 4000 5000 6000 7000 8000 Popuiations (thousands) lx) Fig 1. Bus passenger trips per auinuos bar cities oj dijjerinig population in developed and dev'elopinig countri~es. (4) Population may not be a very satisfactory way of comparing the sizes of' cities in developed and developing countries because patterns of social behaviour are different. Thus in the Third World, only one member of a large family may be found to use public transport, usually to travel to and from work, other members of the family staying at home. (5) The data are cross-sectional and do not reflect the trends described earlier, i.e. that public transport supply and use is increasing in developing countries. In comparison with the above result, the regression equations relating the number of passenger trips per annum with area of city were m-arkedly different for cities in developed and developing countries (although corre- lations are, in this case, weaker). From Fig.2 it can be seen that the number of passengers per annum in cities in developing countries was much greater than in cities of similar area in developed countries. This could reflect dif- ferences in land utilisation and greater population densities that exist in most African and Asian cities. The lower densities of cities in the developed world adversely affect the economics of bus services and at the same time provide a powerful inducement to use private transport. This could explain in part the reason for the different results shown in Figs I and 2. Meanwhile, detailed research is being con ___ ducted in Delhi by the TRRI-'s Overseas UnitW on the socio-economic factors affecting bus usage in that city. As might be expected, in both groups of cities there was a close correlation between passenger trips per annum and the in- dependent variables that describe bus oper- ations. The number of passengers carried per bus fle'et size and per bus kilometre was fairly similar for both groups, although there was a tendency for the larger fleets in developed cities to carry correspondingly more passengers, indicating perhaps the greater efficiency of bus operations in Europe and North America. There were marked differences in tlse relation- ships derived for passengers carried and stall' employed, number of routes operated and system length operated for the two groups of cities. The greater number of passengers per route and per system length operated in developing cities might be a consequence of greater population density, an inadequate number of routes (or total network), or perhaps an. inadequate supply of vehicles to routes operated. There were considerably Fig,2. Bus pasusenger tri p~spertaipipiti for c.itiesof dil.tiring area its developedan~ddlevelopiuig coias fries. 1400 1200 1-- 1000 l- 800 l- -gc 0, 600 I' 400 k 200 n0 100 200 300 400 500 600 700 800 900 1000 1100 1200 Area (km I) (x) 2 O Developing cities * Developed cities 0 - 0 ~~~~~~~~~0 0 Developing 0 -_ -7.+13 lDeveloped c 0 r , 0 .6 000 * **~~~~~~ ~ = 69.0 + 0.26(x) .**~~~~~~~~he I .1 more passengers per bus company employee in developed cities than in those of the Third Worl d. This illustrates the very high staffivebicle ratios that exist in developing countries (see later). The trip rate, i.e. the average numher of' a~zssenlger trips per person per annum111 was found, in both groups of cities. to be related to gross national product (GNP) per capita of the various countries. Since the GNP values relate to the countries as a whole and not to the specific cities this must be regarded as an approximation of the real relationships that may exist. Nevertheless the relationships derived between trip rates in the cities and the GNP/capita for the countries as a whole were statistically significant at the 5 per cent level. The slope of the regression line in Third World cities was found to be positive, the trip rate increasing rapidly with the increasing GNP, whereas in the European anid North American cities the slope was negative, the trip rate decreasing slightly with increasing GNP. This implies (although this may perhaps be an over- simplification) that as people become more affluent in the Third World they may make more bus journeys each year, whereas in the developed world they buy more cars and consequently make fewer bus trips. In comparing the trip rates with the ,ideedent variables that decribe bus comn- #ny, characteristics, the trip rates in both groups of cities were found to be significantly related to the number of stall' employed, the annual bus kilometrage travelled and number of routes operated. In these relationships, the trip rate was greater in developed countries for the same number of staff employed or bus kilometres travelled. In both groups, the trip rate increased with increasing stall', number of routes or bus kilometres, i.e. the larger the 'service', the greater the number of trips made. per person per annum-due possibly to a. better service. Current research in Delhi is studying in much greater detail factors affecting both supply of and demand for public transport. Level of service As seen earlier, the numbers of passengers carried per annum in a number of cities in developing countries increased considerably over the period 1964-75. As shown above the number of bus journeys made per person per annum was found to increase with increasing vel of service. It may well be that an ladequate fleet size and route network is constraining the total number of passenger trips and there may be a considerable suppressed demand for public transport in these Third World cities. In order to obtain an estimate of the level of service provided in the two groups of cities, the average values of parameters that describe, in part, the level of public transport service were calculated. These are given in Table II. As can be seen from Table 11 there were, on average, about 40 per cent fewer buses per Table Ill. Service output levels Servicesoperated Kms Fleet to services operated to utilisation Company/city, scheduled kms scheduled level year (percentage) (percentage) (percentage) Delhi Transport Corporation (DTC), 1976 86.2 87.7 67.4 Bombay Electric Supply & Transport Undertaking (BEST), 1976 98.3 - 92.1 Pallavan (Madras) Transport Corporation Ltd (PTCL), 1976 97.5 94.9 88.6 Suva, 1973 66.1 88.4 Mombasa, 1977 100 96.1 94 Nairobi, 1 977 100 97.9 90 Kingston. Jamaica, 1977 88.4 80.0 88 Table IV. Costs and revenues in various cities Co st per bus Revenue per bus 'Profit' per bus City km (pence) km (pence) km (pence) Bombay. 1976 21.8 20.0 -1 8 Blantyre, 1 977 24.3 23.4 -0.9 Chieng Mai, 1976 9.1 6.0 -3.1 Delhi, 1976 18.8 10.5 -8.3 Kingston, Jamaica, 1977 48.9 47.7 -1.2 Kuala Lumpur. 1977 9.4 10.2 +0.8 Madras, 1 976 14.8 14.9 +0.1 Mombasa, 1977 20.1 21.0 +0.9 Nairobi, 1 977 21.0 25.2 +4.2 Nicosia, 1977 5.1 6.3 +1.2 Poona, 1976 16.1 1 54 -0.7 Surabaya, 1976 8.5 11.7 +3.2 Average values of 5 municipal operators in 31.9 28.1 -3.8 U.K. 1974-75 head of population in cities of the Third World eviM though the level'of car owrnership is very much lower than in the industrialised countries. Similarly, the number of route kilometres per head of population is about half that in Europe and North America. It should, of course, be remembered that, in many Third World cities, paratransit systems provide additional transport services. This is discussed later.The number of buses per route kilometre is the same in both groups of cities. Closer examination of these values shows that bus routes in cities in developed countries are, on average, only half as long as those in developing countries. In order to try to provide a reasonable level of service on these long routes, operators in developing countries have to run more buses per route than is the case in developed countries. One of the most characteristic features of bus operations in Third World cities is the very long routes that meander across large cities with terminal-to- terminal times of up to several hours. Delhi was one city where a major route rationalis- Table II. Level of service in developed and developing cities Average value of Average value of Service level 97 developed cities 42 developing cities Busesper 100 000 population 93 63 Route kilometres per 1 00 000 population 100 54 Busesper route kilometre 1.2 1.2 ation scheme was adopted and passengers now travel directly between nodal interchange points. The effectiveness of this system is presently being examined by a joint team from TRRL and the Indian Association of State Road Transport Undertakings (ASRkTU) in Delhi.The level of service of a bus company is dependent not only' on the number of buses owned, but also on the reliability and effectiveness with which they are operated. The internal operating efficicency of a bus company can be assessed by an examination of factors such as ratios of services operated to services scheduled; kilometres operated to kilometres scheduled; and the proportion of vehicles fit for use (fleet utilisation levels). Table III shows values obtained for a small number of bus companies operating in India, Kenya. Fiji and Jamaica. It can be seen that call bus companies reported suprisingly 'high se rvice levels (com- parable with London Transport). Unfortunately such figures can be misleading, since thay provide no indication of the condition (and hence the comfort and safety) of the buses, the frequency of service and passenger waiting times. Such factors are also being examined by the TRRL/ASRTU team based in Delhi. Costs and revenue As seen earlier, the demand for public trans- port has declined steadily in the U.K. over the past decade. This in turn has led to general agreement that public transport services should be regarded, in part, as providing an 3 Table V. Operating costs for public transport operators (percentages) > ~~~~~~~~~~~~~~~~0 CL 0 0 0 E Cost factor c R i -0 0 c M,~~~ 0 c .. 0 0 0s a. 5 . . Staff 59 72 44 39 34 40 39 39 34 31 23 37 39 44 60 37 35 38 Fuel 3 6 1 5 1 5 21 20 1 2 1 7 1 9 1 5 20 1 8 1 6 6 4 22 1 3 1 2 Tyres/Spares 1 ~ 10 1 2 1 4 1 4 1 1 1 2 5 1 1 1 1 6 5 3 24 6 1 8 6 Maintenance 20 j 1 8 1 6 13 1 9 27 1 6 21 20 Depreciation/Interest 4 3 1 4 24 9 1 2 27 1 7 '11 1 5 36 7 6 12 5 10 9 9 Taxes/licences 1 - 7 4 1 3 8 2 -7 6 2 3 - 4 - 3 - - Other 12 - 10 6 9 6 9 8 7 10 7 1 9-- 1 5-- 4 7 6 4 15-- Total 100 1001100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 *Bus companies allocate maintenance costs mainly as cost of tyres and spare parts Depot and traffic operating costs and administration essential service and, as such, should receive a nmeasure of financial support. In contrast, it mlight be expected that with the large and increasing demand for public transport in Third World cities, bus companies should have no difficulty in being economically viable enterprises. The small sample shown in Table i V suggests that this is not the case. In over half the cities for which data were available, operating costs per bus kilometre exceeded revenues. In many of these countries, India being a particularly good example, there has been a policy of maintaining low fare levels irrespective of the cost of providing the service. ~equests for fare increases are subject to ilengthy enquiries and small increases (if .ianted at all) are often out of date by the time .hey aie implemented. An examination of the structure of ---perating costs in cities in developed and Dcveloping countries reveals some interesting differences. It can be seen from Table V that in the U.K. the major cost component, about two-thirds the total, is that of staff, mainten- ance and spares making up most of the remainder. Operating costs in the Third World cities are clearly different, with stafrcosts being about one-third the total, but with fuel and depreciation/i'nterest costs being much greater than in the U.K. The number of staff employed, per bus in the 42 cities of developing countries was, on average, over twice as high as in cities in developed countries (7.4 compared with 3.6). Thus, even with one-man operation on many buses in the U.K., staff costs are high; in comparison, in many Third World cities there are often three persons operating each bus, but nevertheless staff costs would seem to be relatively low. The distribution of costs in Nicosia, Cyprus and Kingston, Jamaica, appear to lie some way between the costs in the U.K. and other Third World cities. In these countries of relatively high levels of GNP, staff costs are obviously more than in Africa and Asia. Fuel costs are also similar in Nicosia and Kingston to the U.K. In the U.K. fuel is exempt from duty for bus companies; in many developing countries this is not the case and fuel costs are proportionately much greater. Perhaps the mnost interesting difference in operating costs between cities in the U.K. and in the selecied cities in the developingeountrics, is the proportion of' total costs allocated to depreciation and interest. For example, in Chieng Mai. Thailand. over one-third ol the total operating costs of public transport is spent on these two components. In many Third World countries the operational life of a bus is often extremely short, sometimes as low as six years, consequently vehicles have (in theory) to be replaced at very short intervals of time and depreciation costs are correspondingly high. In addition, many public transport operators in developing countries receive virtually no government support at all. In order to continue supplying public transport services, loans have to be acquired (sometimes fromn Central Government itself, as in Delhi, or fromi State Governments, as in Calcutta). The interest paid on these loans is often a considerable proportion of total operating costs for these bus companies. Financial costs of operation are further complicated in India by the imposition of 8 .7 1 00 'c .2 00 4 8 12 1 6 20 Distance (km) direct and indirect taxes. Such taxes consist ol passenger taxes (levied on the number of~ passengers carried), vehicle taxes and income taxes. This means that the more successful and efficient the transport organisation, the greater* the penalties imposed by government. To a visitor from Western Europe or North America, bus fares night appear to be low in many cities in developing countries. Where a fiat-fare system operates, passengers are able to travel up to 16 km for the equivalent of a 2p fare. In the U.K., a similar journey would cost over 50p.Figure 3 shows how fares per kilometre travelled vary in parts of England and in a number of developing countries (where a fiat- fare system is not in operation). In both England and the developing countries for which data were available, the cost per kilo- mietre of a bus journey decreases with increas- 24 28 32 36 40 4 Fig 3. Bus fares per distance tra ililed otid distonce Jbr dijcreit countries. intg distance. A number of' bits operators in developing countries show particularly sharp dececases in cost for journeys of over six kilometres iii length. Studies of public transport operations inl Chieng Mail and Surabava indicated that lw comec travellers usebhuses. as opposed. to minibuses or other, forms of paratransit. In order to relate the cost of bus travel to some measure of the wealth of passengers. thecoist of a five-kilometre journey in different cities was regressed against the GNP value of the various countries. As was pointed out in the earlier comparisons of trip rates and GNP levels, relationships derived are approximations since GNP values for city dwellers are likely to differ considerably from those of the country as a whole. However, the relationship derived (see Fig 4) was statistically significant at the 1 per cent level. Thus, even in those cities where particularly low fares mnay appear to exist, the fares would seem to be commensurate with the general wage levels of the countries concerned. DiscussionWith the very low levels of car ownership that exist in most cities in developing countries, it is essential for the social and economic well- being of the community that an adequate public transport system be provided. In a . number of cities paratransit systems play an important role, but in most urban areas in the Third World there is, and will continue to be. a heavy reliance on the conventional bus system. The analysis of trends in bus usage in cities for which data were available has indicated the growing demand for public transport that has taken place in recent years. Whilst this growing demand for public transport has been taking place, government support in many countries has been minimal and operators (be they private, municipally- owned or controlled by central government) are often treated as commercial enterprises and consequently the problem of remaining econ- omically viable lies with the bus company. Indeed, in India successful bus operators are often penalised by means of taxes for providing an efficient service to the general public. On the one hand bus organisations are treated as commercial enterprises, whilst on Fig 5.Boadig abusi Dei 35 30 25 h Z E m CLA 20 [ 15 1~ 10 1~ 5 00 400 800 1200 1600 2000 2400 2800 3200 3600 4000 GNP per capita in U.S. $ Fig 4. The cost oI a 5-kni bus journey aurd GPN per. capita (1974) ini c.itie in di vet.oped anid dev'elopinig coirutries. the other hand they are asked to provide a ,social service, operating uneconomic routes. providing particularly cheap fares to students and being subject to rigid fare controls. With the large demand that exists for public .transport in developing countries it might be thought that bus companies are thriving, with relatively few financial problems; from data available this appears not to be the case. Although the operating costs per kilometre are lower than in Great Britain, revenues are also much lower. In addition, the large demand appears to be associated with a tower level of service: the total numbers of buses and bus kilometres available per head of population are much lower in Third World cities than in those of Europe and North America. In many cities in developing countries, paratransit sys- tems have develope d to satisfy this demand for public transport. Although this paper has outlined differences in levels of service in the broadest sense, it is important to identify deficiencies in greater detail. The Overseas Unit, TRRL, and the Indian Association of State Road Transport Undertakings (ASRkTU) 'are studying the transportation system of a Third World city (Delhi) to see how it functions, how it is used. who uses it and what satisfaction users receive. All modes of transport (including walking. cycle, auto-rickshaw, etc.) are being examined. The aims of the two-year research programme are: (I) To provide information on the effective- ness of urban public transport in meeting the needs of the traveller, particularly from the poorer sections of the community; and (2) To provide information on the effecctive- ness. of mcasu res to improve the efficiency of' .public transport operations. The research will identify important traveller characteristics and transport supply chatrac- teristics and how these vary in response to changes in each other and in the operating environment. This information clan be used to provide a sound basis for policy decisions on maitiners such ats airest ructurte. route rationalIis- atisin. control of' parilrtiaiisit and lieeds of, tile urbanl poor. Meanwhi ile, throughout the dexeloping world many passengers travel on unsafe buses in extremely overcrowded conditions (see Fig 5) having waited (as preliminary results from Delhi indicate) up to two hours for a bus to arrive. If levels of service are to improve in the future, but with fares related to the low income levels in developing countries, public transport services may well need some kind of government support in meeting the needs of the community. ACKNOWLEDGMENTS The work described in pths paper Joiris part of/ the research programme of thle Overseas Ujiji, Trans port and Road Research Laboraiory (Un ii Head: J. N. Bulnan) and the paper is published by permission of ihe Director. The authors would like to ackniowledge the assistance of the transport consultants who proivided essential data and also the help of Miss W A. tHirds. Newcastle Polytechnic, and N. Becketi, Utniversity ol Reading, with the subsequent analysis. Crown Copyright 1979. Any views expressed iii this pa per are not necessarily those of the Department oif the Environment, the Department of Tratisport or tie Ministry of Overseas Development. Extracts from thy text may be reproduced except for comniercial purposes provided the source is acknowledged. REFERENCES' 'WH-SHERs. J. and R. RIGHTER. The Explodinig Citiei. Andre Deutsch Ltd. London, 1974. 'J.s-otis, G. D. and P. R. FOURACRE. Intermediate forms of urban public transport in developing countries. Traff Engmig Conttrol. 17 (3). March 1976. 98-100. 'FoURACRE. P. R. and D. A. C. MAuNDERi. Pubtic transport in Chieng Mai, Thailand. Trq1ff Engng Control. 181(51, May 1977, 260--26t. 'Bvs, P. H. and R. H. OLC)FIELio. The effect of car ownership and. income on bus travel. Trall. Emigmg and~ Control. 19 (8/9). August/Sepivem er t978. 392-396. 407. 'international Union of Public Transport. lmitii- miatioiiol Statistical Ha~ndh,,ek ont Urhani Pulitic Trun~spimrl. Brussels. 1975. 'Li GRIS, R. Measuring and iimproiving pr ...iiciivii) iii urbain passenger st rl'ae iransporti. PPii 3 9Vii h i,. ( o ng rviss ut th e m t r o ii s m u io Piuhlii Transpiort. Rotor. 1971. 'L r , L. L. S. A study of public transpo..rt iii relation to urban furia. PhD thesis, Universits otf Strathclyde. 1971. 'I-UtRlACRE. P. R. and 1). A. C. MAsUNDis p1ubliC tratnsport in Surabaya. Indonesia. TR RI Suip. plemen a ry R epuirt SR3711. Transport anid Ru nal Research La boraioury. Croiwi huorn. 197N. and Rood Rese tohi Labioiratoriii. Cioithu lBerkshirie R(;HI 6.4 :- 0 5 *Kent*Midtand Red Mahd * Mauritias * ogKong * .Kalia Lumpur e Kuching * Beirut *Caracas 9 Calcutta9 Manila MI 1