High Volume Transport

Vital transport research to ensure accessible, affordable and climate friendly transport for all.

A comparison of freight transport operations in Tanzania and Indonesia


Publications with the same themes

View all

PDF content (text-only)

A comparison of freight transport operations in . Tanzania and Indonesia by J L Hine (TRL) J H Ebden and P Swan (Mott Macdonald) TRL Report 267 The Transport Research Laboratory is the largest and most comprehensive centre for the study of road transport in the United Kingdom. For more than 60 years it has provided information that has helped frame transport policy, set standards and save lives. TRL provides research-based technical help which enables its Government Customers to set standards for highway and vehicle design, formulate policies on road safety, transport and the environment, and encourage good traffic engineering practice. As a national research laboratory TRL has developed close working links with many other international transport centres. It also sells its services to other customers in the UK and overseas, providing fundamental and applied research, working as a contractor, consultant or providing facilities and staff. TRUS customers include local and regional authorities, major civil engineering contractors, transport consultants, industry, foreign governments and international aid agencies. TRL employs around 300 technical specialists - among them mathematicians, physicists, psychologists, engineers, geologists, computer experts, statisticians - most of whom are based at Crowthorne, Berkshire. Facilities include a state of the art driving simulator, a new indoor impact test facility, a 3.8km test track, a separate self-contained road network, a structures ha] 1, an indoor facility that can dynamically test roads and advanced computer programs which are used to develop sophisticated traff]c control systems. TRL also has a facility in Scotland, based in Edinburgh, that looks after the special needs of road transport in Scotland. The laboratory’s primary objective is to carry out commissioned research, investigations, studies and tests to the highest levels of quality, reliability and impartiality. TRL carries out its work in such a way as to ensure that customers receive results that not only meet the project specification or requirement but are also geared to rapid and effective implementation. In doing this, TRL recognises the need of the customer to be able to generate maximum value from the investment it has placed with the laboratory. TRL covers all major aspects of road transport, and is able to offer a wide range of expertise ranging from detailed specialist analysis to complex multi-disciplinary programmed and from basic research to advanced consultancy. TRL with its breadth of expertise and facilities can provide customers with a research and consultancy capability matched to the complex problems arising across the whole transport field. Areas such as safety, congestion, environment and the infrastructure require a multi-disciplinary approach and TRL is ideally structured to deliver effective solutions. TRL prides itself on its record for delivering projects that meet customers’ quality, delivery and cost targets. The laboratory has, however, instigated a program me of continuous improvement and continually reviews customers satisfaction to ensure that its performance stays in line with the increasing expectations of its customers. TRL operates a quality management system which is certificated as complying with BS EN 9001. & @@ *O ,5 ‘*RQ“~~?h Wansport Research Foundation Group of Companies b * ~ G Transpon Research Foundation (a company limited by guarantee) trading as Transport Research Laboratory. D z 70 ssnm Registered in England, Number 3011746. TRL Limited. Registered in England, Number 3142272. ~> ~$’ lmmAmn f.7”9”Q\ umlnOm Registered Office: Old Wokingham Road, Crowthorne, Berkshire. RG45 6AU Mmm Transport Research Laborato~ Old Wokingham Road Crowthorne, Berkshire,RG456AU DFID Department for International Development 94 Victoria Street London, SWIE 5JL TRL REPORT 267 A COMPARISON OF FREIGHT TRANSPORT OPERATIONS IN TANZANIA AND INDONESIA by J L Wne (TW), J H Ebden and P Swan (Mott Macdonald) Subsector Transport Theme: Increase the efficiency of national and regional transport systems Reject title: Reducing the cost of freight in Africa Reject reference: 6240 Copytight Transport Research Laboratory 19W. All rights reserved. This documentis m outputfrom a projectawtid to MottMacDonaldand~ by the ~ Department for htemadond Development ~~) forthe benefitof developingmuntries. The viewsexpressedm not n-ssdy thow of the D~. Transpoti Researeh Foundation Group of Companies Transport Research Foundation (a company limited by guarantm) trading as Transport Research hboratory. Registered in England, Number 3011746, ~ hmited. Registered in England, Number 3142272. Registered Offices: Old Wotin@m Road, Crotiome, Berbhi~, RG45 6AU. First fiblished 1997 ISSN 0968-4107 CONTENTS Page Executive Summary 1 Abstract 3 1. Introduction 3 2. Road freight transport in the two countries 3 2.1 Taztia 3 2.2 Indonesia 4 3. The surveys 5 4. Survey results 6 4.1 Trip characteristics and vehicle type 6 4.2 Truck ownership, crew characteristics and methods of finding loads 9 4.3 Vehicle age, utilisation and empty running 10 4.4 Journey speeds and fuel consumption 12 4.5 Freight tariffs 13 5. An analysis of vehicle operating costs 17 5.1 A comparison of component prices 17 5.2 Vehicle maintenance and tyre costs 19 5.3 A comptison of overall operating costs 20 6. Possible solutions to encourage the development of low cost transport in Africa 7. Acknowledgements 8. References Appendix 1 1. Roadside truck survey questionnaire 2. Transport operator’s survey questionnaire 20 22 22 23 23 25 EXECUTIVE SUMMARY Previous research has shown that during the mid 1980’s road freight transport costs in Francophone Africa were in the region of four to five times the costs of freight transport in Pakistan. Data collected from other countries confirmed that in general African transport costs were much higher than in Asia. In order to investigate further the differences in freight transport costs and efficiency between Africa and Asia, and to help identify measures which might reduce costs, a comparative study of road freight transport operations was carried out in Tanzania and Indonesia. The main data collection of the study wasthrough roadside interview surveys of truck drivers which wassupplemented with a series of structured interviews of transport operators at their offices. In Tanzania 270 roadside interviews were carried out on the main road just outside of Dar es Salaam. In Indonesia 250 drivers were interviewed attwo commercial rest stops atPemalang, about 250 km east of Jakarta on the main northern corridor connecting Jakarta with Semerang and Surabaya. In both surveys data were pnncipdly collected from drivers undertaking long distance movements. Average loaded trip distances were found to be 803 km in Tanzania and 898 km in Indonesia. Thirty per cent of trucks were found to be empty in Tanzania while only five per cent were empty in Indonesia. However it is befieved that in the latter case,the low percentage of empty running may have occurred because of a bias in the sampfing technique (only vehicles stopping at the rest stops were interviewed). Two axle trucks made up 56 per cent of the survey in Tanzania but only 43 per cent in Indonesia. Articulated trucks accounted for 25 per cent in Tanzania and 44 percent in Indonesia. The ownership stmcture of the vehicle fleet were similar in both countries with private transport companies accounting for about hdf of the fleet and private individuals a further quarter. Although there was little difference in mean vehicle age for two axle trucks the larger vehicles in Tanzania were found, on average, to be two years older than in Indonesia. The annual utifisation of two and three axle trucks was found to be about 60,000 km and 80,000 km respectively for both countries. However, Indonesian vehicles were found to achieve 82,000 km per annum compared with 60,000 km for those of Tanzania. These figures are higher than the utitisation recorded in Francophone Africa but less than that found in Pakistan where dl types of vehicles travelled more than 110,000 km per year. In terms of methods of finding loads the drivers reported that freight agents were used in 17 percent of cases in Tanzania and 18 percent of cases in Indonesia. In contrast in Pakistan over 60 per cent of all drivers found their loads through freight agents. Two axle vehicles were found to have a mean load of 5.6 tonnes in Indonesia and 7.1 tonnes in Tanzania. Three axle trucks were much heavier loaded in Indonesia, at 15.6 tonnes, compared with 10.5 tonnes for Tanzania. Articulated vehicles had very similar mean loads (21 tonnes) in both countries. Average vehicle journey speeds were found to be very much higher in Tanzania than in Indonesia. For example, for three axle trucks mean journey speeds of 50 kph were reported for Tanzania compared with only 15kph for Indonesia and 24 kph in Pakistan. The high journey speeds and differences in load factors helped to account for the much higher fuel consumption @ertonne km) found for larger vehicles in Tanzania compared with Indonesia. For three axle trucks a mean fuel consumption of 0.06 ltrs per tkm was recorded for Tanzania compared with 0.03 ltrs per tkm for Indonesia and 0.02 for Pakistan. Tariff rates per tonne km for trucks carrying 3 to 13tonnes were estimated to be 8.6 US cents (at mid 1995 prices) for Tanzania, 3.8 US cents for Indonesia and 2.8 US cents for Pakistan. For trucks carrying over 13 tonnes the tariff rates were estimated to be 9.6 US cents for Tanzania, 2.1 US cents for Indonesia and 2 US cents for Pakistan. A large proportion of the differences in tariffs between Tanzania and Indonesia appears to relate to the differences in vehicles and fuel prices. For example a common two axle truck in Indonesia costs about 22,000 US $ and about 65,000 US $inTanzania. Atractorandserni-trailercosts about 135,000 US $ in Tanzania while a comparable vehicle costs 73,000 US $ in Indonesia. Fuel is particularly cheap in Indonesia at 0.166 US $ per litre compared with 0.435 US $ per litre in Tanzania. A number of measures were identified which could contribute to lowering transport costs in Tanzania and the rest of Africa. These include: a) Competition is the main mechanism by which costs are kept under control or reduced within the transport sector. Every encouragement should be given to ensure that there is competition in both the supply of vehicles and parts and that transport operations remain competitive. Efforts should be made to control the power of the forrnd and inforrnd trucking associations that are common in Africa. The associations often work to restrict supply and share out available demand through vehicies queuing at lorry parks. b) Efforts should be made to restrict the use of exclusive dealerships in the import of vehicles and spare parts. High cost, high specification vehicles are frequently used in Africa whereas in Asia deliberate attempts are made to import very cheap basic vehicles which are then, if necessary, modified after purchase. 1 c) In Pakistan delays are minimised, empty running reduced and tariffs kept low by the extensive use of freight forwarding agents which act in competition with each other. d) A high utilisation can be achieved by using two drivers that drive night and day. Night driving need not be dangerous if a vehicle’s lights are in working order and the vehicle drives slowly. e) Fuel consumption was found to be high in Tanzania partly as a result of high driving speeds. Slow tiving speeds not only reduce fuel consumption but can keep parts consumption low and reduce the risk of accidents. ~ One of the main factors keeping costs low in Pakistan relates to the large measure of responsibihty given to the driver. He is aware of costs and revenues of the truck he runs because he finds loads, arranges and pays for maintenance pays for fuel and keeps accounts. He is very aware of the benefits of driving slowly and of paying close attention to routine maintenance. 2 A COMPARISON OF FREIGHT TRANSPORT OPERATIONS IN TANZANIAANDINDONESIA ABSTRACT Previous research has revealed that road freight transport costs in Francophone Africa were in the region of four to five times the costs of freight transport in Pakistan. To investigate further the differences in freight transport efficiency and costs between Africa and Asia new surveys were carried out in Tanzania and Indonesia. This report presents findings from these surveys together with comparative data from Pakistan. The results suggest that in Tanzania long distance freight transport tariff rates and overall tariff revenues pertonne-km are between two to five times those of Indonesia and Pakistan for different vehicle types and load weight categories. The report identifies a range of factors contributing to these differences including disparities in input prices, utilisation, load factors, fuel efficiency and maintenmce practices. A range ofsuggested measures are proposed to help improve efficiency and reduce transport costs. 1. INTRODUCTION Previous research has been carried out which showed that long distance road freight transport costs in Francophone Africa (ie C6te D’Ivoire, Cameroun, and M~) were in the region of four to five times the freight transport costs of Pakistan. Limited data collected from Central Southern Africa (Zimbabwe, Zambia and Malawi) indicated that high transport costs dso appeared to be common in this region of Africa. In contrast, data collected from India and Vietnam showed that other Asian counties dso had low transport costs (fine and Rizet, 1991, Rizet and Mne, 1993). To investigate the differences in transport costsfurther, and to identify any measures which might help to reduce costs, it was decided to carry out similar surveys in Tanzania and Indonesia. In the following the results of these surveys are presented together with data from the earlier Pakistan research. A full description of the Pakistan work is given in TRRL Research ReportNo314 (fine and Chilver, 1991). In Tanzania a structured interview survey of transport companies was carried out in Dares Salaam and a roadside survey of truck drivers was undertaken on the main road just outside of Dares Salaam. The surveys were carried out by staff from the Ministry of Works, Communications and Transport in addition to privately hired enumerators. In Indonesia a structured interview survey of transport companies was undertaken in Jakarta and a survey of truck drivers was carried out at commercial vehicle rest stops at Pemalmg, about 250 km east of Jakarta, on the main northern corridor route of Java. The truck drivers interviews were carried out by staff from Institute of Road Engineering (IRE) located at Bandung. 2. ROAD FREIGHT TRANSPORT IN THE TWO COUNTRIES 2.1 TANZANIA Tanzania has an estimated population of about 27 milfion. The mainland land area is 881,000 sq km; hence the population density at about31 people per sqkm isrelatively low. In 1993 it had an estimated Gross National Product (GNP) percapita of US$ 90, one of the lowest in the world. From 1980 to 1993 the GNP growth rate was 3.6 per cent per annum which was just above the rate of population growth (World Bank, 1995). Over half of the Gross Domestic Product (GDP) is derived from a@culture and the population is predominantly rural. Industry accounts for about 14 per cent of GDP. Urban economic activity is heavily concentrated at Dar es Salaam; it has about one third of all manufacturing employment. Other important centres include the port of Tanga, Arusha and Moshi in the north, Mwanza on Lake Victoria and along the main TANZAM corridor connecting Dar es Salaam with the Zarnbian border. Dar es Salaam accounts for over 90 per cent of the total mainland port trtilc ofthe country. Besides handling most of the domestic trtilc Dar es Sdaarn is dso a key transit port taking traffic for Zambia, Malawi, Rwanda, Bumndi and eastern Zaire. In total there are reported to be 82,000 kilometres of natiomd, regional and district roads. Most of these roads are rural earth roads. The main trunk road network is 10,300 km in length of this about 3,800 km has a bitumen surface and 3,300 km has a gravel surface. In 1986 the Economic Recovery Programme @RP) was introduced and various measures were subsequently taken to promote economic recovery and development through liberalisation and deregulation. As a result of these measures the private sector was able to undertake activities previously restricted to the pubtic sector. Within the transport sector private companies could market and supply agricultural crops and inputs. Freight tariffs were decontrolled and there was a hberalisation of foreign exchange allocation procedures. 3 In an attempt to restore the essential road network to acceptable conditions the Integrated Roads project ~RP) was introduced in 1987 with considerable donor support. As a result, there has been a large increase in the funding allocation for roads. For example in 1987/88 US$ 21m of the government development budget was spent on roads, whereas by 1992/93 this had risen to US$ 76.5m, of which 76 per cent was from foreign sources. In June 1993 it was estimated that by the end of 199564 per cent of paved (trunk) roads will be in good condition (Ministry of Works, 1993). Within Tanztia the total transport sector accounts for about six per cent of GDP and, in view of the importance of transit traffic, about 15 percent of foreign exchange earnings. Road transport accounts for over 70 percent of freight movement in the country. It has been estimated that in 1993/94 total road freight transport demand was composed as follows: international road freight: 750 m tonne-km inter-regional road freight: 2850 m tonne-km intra-regional road freight: 150 m tonne-km total road freight: 3750 m tonne-km International road freight mainly consists of transit trtilc to and from Dares Salaam. There are believed to be about 60 private Tanzanian companies operating in this sector. They tend to use high capacity trucks and have typical fleet sizes of between 10 to 30 vehicles. During 1994 and 1995 demand is believed to have fallen as traffic destined for Rwanda and Burundi has switched to using the port of Mombasa while Malawi trtilc has switched to using Mozambique ports. Inter-regional road freight accounts for about 75 percent of the total freight transport demand. A high proportion ofthis movement is between regional centres and Dares Salaam. Most of the companies operating in this sector are private with a high proportion of transpofiers only owning one truck. Most of the trucks used have two axles with load capacities often tonnes or less. Theintra-regiond market only accountsforasmdl proportion of freight movement. This component of trtilc mostly consists of short distance movement between district and regional centres and villages. Most transport is by smaller capacity two-axle trucks. The public sector has a higher proportion of the market than for the longer distance movements. Public sector transporters comprise cooperatives, parastds and the Regional Transport Companies (MTCOS). The exact number of freight vehicles in Tanzania are not known but, for vehicles of three tonnes and above, it may now be the region of 15,000 compared with an estimate of 10,300 in 1986 and 13,200 in 1991. Over the last 20 years there have been a number of reports which have argued that there was a shortage of transport capacity in the country. @xamples include the 1975 Fourth Highway Project, the 1977 World Bank’s Trucking Industry Rehabilitation and Improvement Project and more recently the 1993 Road Transport Study for the National Transport Corporation). The arguments used to support this case have included the severe difficulty of finding trucks in rural areas, the numbers of immobile vehicles without spare parts and an analysis based on estimated demand compared with estimated capacity assuming certain levels utilisation. To help tackle this problem there have been various donor assisted prograrnmes to improve the procurement of vehicles and spare parts. Besides the World Bank’s 1977 project there have been import support schemes finaced by UK, Italy, Sweden and India as well as prograrnmes supported by the Nordic countries, Intemationd Finance for Agricultural Development (FAD) and Japan. A recent report has argued that supply now exceeds demand, particularly for intra-regiond transport. The evidence for this is that freight transport rates have fallen in red terms and that the import of new trucks has declined from the peak levels in 1992 Wational Transport Corporation, 1995). 2.2 momsm Indonesia has an area of 1.9 m sq km and, in 1993 a population of 187 miltion. However more than 60 per cent ofthe population hve on the island ofJava which has an area of 132,000 sq km. Java’s population density is very high at about 877 people per sq km. In 1993 it had an estimated GNPper capita of US$ 740. Between 1980 and 1993 it had a GDP growth rate of 5.8 per cent per annum and a population growth rate of 1.7 per cent per annum. In contrastto Tanzania in 1993 agriculture only accounted for 19 per cent of GDP while industry accounted for 39 per cent. The Jabotabek region, an area around the capital, Jakarta (pop. 6.5 m) is the most important economic region of the country. On Java other important towns are Surabaya (pop. 2m),Bandung (pop. 1.5m) and Semerang (pop. 1m). Within Indonesia coastal, shipping, road and rail dl play an important role. For example in 1991 the following interurban domestic freight modd spht was estimated for Java bilhon tonne-km Road 27.6 Rail 1.1 Coastal shipping 42.1 Total 70.8 Long distance road freight transport tends to be chosen over rail and coastal shipping for non bulk commodities because of speed and reliability. During the 1980’s, the annual vehicle fleet grew at an annual rate of 9 per cent. To accommodate the growing 4 oshi ■ Tabora 9 Tanga o Zani Dodoma ■ A zibar ‘:ge:rEs Salaam \~JItinga . v ■ Mbeya Fig. 1 Map of T-ania traffic volumes there has been a considerable expansion in the classified road network from 160,000 km in 1979 to 260,000 km in 1992 (of which 17,700km are national roads and 34,100 km are provincial roads). In addition it has been government policy to significantly increase expenditure on the road sector over the last ten years. The five year highway sector development plan, Repetita V (1989-94) placed a strong emphasis on the rehabihtation and maintenance of the national and provincial road network. In real terms the 1995/96 expenditure is expected to be around three times the level of 1984/85. During the period 1991- 1994 expenditure on maintenance and betterment for the national and provincial network has averaged around US$ 1biltion. This represents around US$ 19,000perkm, a very high figure compared with other developing countries. On Java dl of the national and provincial roads are paved and most are in good or reasonable condition.For example, in 1993 it was estimated that 53 per cent oftheir length was below 4 IN (Intemationd roughness index) while a further 30 per cent of their length was between 4 and 6 IRI. Congestion is now a major problem in Indonesia, for Java the average annual daily traffic level (AADT) on national and provincial roads was estimated to be 3900. By 1994, 374 km of urban and inter-urban high capacity toll roads were built principally to serve the Jakarta area (Ministry of Works, 1993). In 1990 it was estimated that there were one million trucks in Indonesia. Since 1985 government control of road and sea freight tariffs has been relaxed and nowprices are set by the market. Although route ficences are required for freight vehicles to cross provincial boundaries in Java the regulations have been relaxed and it is notthought to cause major economic distortions. Own account trucking operators are believed to account for more than hdf of all road transport. The majority of for hire operators own only a small number of vehicles although there are some operators who own relatively large fleets. The market is generally beheved to be competitive. 3. THE SURVEYS In Tanzania 270 roadside interviews were carried out with truck drivers on the main road just outside of Dar es Salaam on four separate daysin late July and August 1995. In total 28 percent of trucks passing the survey point were stopped. Amap of Tanzania showing the location of the major towns is given in Figure 1. A range of questions were asked about the origin and destination of the journey, the load, tariff, vehicle utilisation, operating costs and ownership of the truck. This survey was supplemented with information from about a dozen structured interviews with transport operators in Dar es Salaam where more detailed questions were asked about operating costs and management practises. Example of the questionnaires are given in Appendix A. Although driver cooperation was good, their knowledge on 5 Fig. 2 Map of Java operating costs was rather limited and some of the data was difficult to interpret. While some transport operators were willing to cooperate with the structured interview others were not. There appeared to be suspicions that the interviewers were acting on behalf of the tax authorities! In Indonesia 250 interviews of truck drivers were carried out attwo commercial vehicle rest stops atPemalang, about 250 km east of Jakarta on the main northern corridor connecting Jakarta to Semerang andSurabaya. This road is one of the busiest in Indonesia with an estimated traffic flow of nearly 15,000 vehicles per day. Over the three day period approximately 1.6 per cent of the passing trucks were surveyed. A map of Indonesia showing the location of the survey station and major towns is given in Figure 2. The interviews were undertaken during the first three days of December, 1995. The interview form (translated into Bahasa Indonesia) was almost identicd to that used in Tmzania (see Appendix A). The data was supplemented with ten structured interviews with transporters in the Jakarta area. Cooperation with both drivers and transporters was good, although again it was not possible to collect much detailed information on operating costs. 4. SURVEY RESULTS For ease of comparison, the results of the two surveys are presented together as far as possible. Where appropriate data from the Pakistan freight transport study are also given. 4.1 TRIP CHARACTERISTICS Am VEHICLE TYPE As can be seen from Table 1 there was a close match in the total number of trucks surveyed traveling to and from Dar es Salaam. Loaded distances from Dar es Salaam were higher than those traveling in the opposite direction. With a mean loaded trip distance of over 800 km the survey is representative of long distance trip movements. Because of the location of the survey all of the trucks were going to or from Dares Salaam. About 13 per cent of the loaded trucks surveyed were travelhng out of the country; Zambia, Kenya and Rwanda were the most important origins and destinations. A wide range of intemd origins and destinations were recorded. Out of the 186 loaded TABLE 1 Tanzanian truck survey Loaded trucks Empty trucks Direction No. mean trip No. mean trip distance km distance km To Dares Salaam 85 598 48 629 From Dar es Sdaarn 101 979 31 285 Total 186 803 79 478 6 TABLE 2 Indonesian truck survey Loaded trucks Empty trucks Direction No. mean trip No. mean trip distance km distance km To Jakarta 114 924 8 651 From Jakarta 103 859 4 1125* Total 217 898 12 809 *distances based on very small sample surveyed trucks the most important origins and destinations were: Mwanza (19), Mbeya (19), Tanga (18), Morogoro (16), tinga (12), Dodoma (11), and Moshi (10). Basic information on the Indonesian survey is given in Table 2. Here it can be seen that shghtly more trucks were surveyed going towards Jakarta than in the opposite direction. Only a very small proportion (5 per cent) of the trucks surveyed were empty. This is well below what might be expected from other freight surveys. Because the survey was not a pure random roadside survey (in which trucks are stopped at random) but was dependent on interviewing trucks aheady stopped atthe two commercial truck parks it is possible thatthe survey was biased towards long distance loaded trucks. In addition, because the survey was conducted well away from major urban areas, it is to be expected that fewer empty trucks would be surveyed. It is usualto record much higher levels of empty running in and close to urban areas as empty trucks travel between customers and to and from their base depot. Out of the 217 loaded trucks 112had ori@ns or destinations in Jakarta while81 were traveling to or from the Surabaya area. The two next most important origins and destinations were Semerang (18) and Bafi (10). Trip distance distributions for the two surveys are given in Figures 3 and 4. Although there is httle difference between Tanzania and Indonesia in the average loaded trip distance (803 and 898 km) it can be seen that there was a much greater variation in trip lengths in the Tanzanian survey with many very long distance international trips as well as many shorter trips. A high proportion of the Indonesian survey recorded trips between Jakarta and Surabaya which isreflected inFigure 4. Again an element of bias may have occurred in the Indonesim survey as drivers on shorter distance trips may not have wanted to stop for a break. Information on truck configurations is given in Figures 5 and 6. Here it can be seen that two-axle trucks appear to be more important for long distance transport in Tanzania. In contrast truck and trailer combinations are proportionately more numerous in Indonesia while there is little difference in the proportions of the larger three-axle trucks and the tractor and semi-trailer combinations. A wide range of vehicle makes were recorded in both countries. In Tanzania the most populm makes were Scania (25 percent) and Isuzu (18 percent). In Indonesia by far the most popular make was Mitsubishi with 36 percent of the total fleet. Japanese makes were far more popular in Indonesia, accounting for 60 percent of the total, compared with only 32 per cent in Tanzania. An analysis of data showed.that,in the Indonesian survey, there was no clear relationship between vehicle capacity and trip distance. Mean loaded trip distances for two-axle trucks and articulated vehicles were very similar at about 920 km and 900 km respectively, while three-axle trucks appeared to have an average trip distance of about 800 km. In contrast vehicle size appemed to have a marked effect on trip distances in the Tanzanian survey. Two-axle trucks travelled the shortest distances while articulated vehicles traveling from Dar es Salaam, (with a high propotion on international trips) had mean journey distances of more than double those of the two and three-axle trucks. The results are shown in Table 3. Data on truck body types are given inFigures 7 and 8. Here it can be seen that high sided trucks are more important in Indonesia. In the Tanzanian survey tankers were an important component of the trtilc while very few were recorded in the Indonesian survey. The latter probably reflects a peculiarity of the northern corridor route astankers appear to be more commonly observed on other routes in Java. 7 -- 50 40 30 20 10 0 W loaded ❑ empty 201-500 501-800 801-1000 >1000 Trip distance (kms) Fig. 3 Number of trips Ioadedempty by trip distance - Tanzania 2 20 0 C200 201-500 501-800 801-1000 >1000 Trip distance (kms) Fig. 4 Number of trips IoadeWempty by trip distance - Indonesia TABLE 3 Tanzania Mean loaded trip distance by vehicle type I Direction 2 ale 3 mle articulated To Dares Salaam 496 623 728 From Dar es Salaam 616 632 1520 Mean 561 627 1190 a 3 and 4 9 2 ule truck 55.87. +ighsidad 46,2Y. Tractor and semi trailer 14.6% Fig. 5 Truck confi~rations - Tanzania 2 wle truck 43.3% 3 and 4 ale truck 13,0% Tractor and semi 13.0°A Fig. 6 Truck configurations - Indonesia Low eided 18.l% Flat 1o.40A Other Tpper 0.8% 2.?% Fig. 7 Truck body type - Tanzania “ Low eided 21.870 Other 2.O”A Tanker 2.97. trailer Fig. 8 Truck body &pe - hdonesia 4.2 TRUCK O~RS~, CREW CHARACTERISTICS Am ~THODS OF F~~G LOADS Information on the patterns of truck ownership are given in Figures 9 and 10. Important similarities were observed in the ownership structure with both surveys recording that half of the trucks were owned by private transport companies. Very few owner-tivers were recorded in either survey, in contrast in the Pakistan survey 20 per cent of drivers had a full or part share in the vehicle. The biggest difference observed between the two surveys was that in Indonesia private non-transport companies (operating mainly on own-account business) owned 20 per cent of the trucks whilst in Tanzania this was just six per cent. This difference may well reflect the more diverse industrial structure of Indonesia. Another difference recorded was that in Tanzania state transport companies owned over six per cent of the vehicles while in Indonesia this was less than one per cent. In both Tanzania and hdonesia there was an average of 1.06 drivers to each truck. In addition there were dso averages of 1.06 assistants in Tanzania and 1.16 assistmts in Indonesia to help with looking after the load and various other routine jobs required by the drivers. In comparison in Pakistan there was an average of 1.57 drivers and one assistant to each truck. In Tanzania 85 per cent of drivers were regular employees while 12 per cent were casual employees. Very similar results were recorded for the drivers in Indonesia where the corresponding figures were 83 per cent and 10 per cent. Information on finding loads (for non-tankers) is given in Figures 11 and 12. Directions from the transport company’s dispatch clerk were much more important in hdonesia (54 per cent) than in Tanzania (37 per cent). However in Tanzania 13 percent of drivers found their loads from truck parks compared with two per cent of drivers in Indonesia. In both surveys, similar proportions of drivers found their own loads (20 per cent) or used freight forwarding agents (17 and 18 per cent). 9 Pfivate Trans~fl 50.2% Fig. 9 Truck Ownership - Tanzania Patinerstip 2,5% nia and Indonesia in three-axle (60,000 and 56,000 km) and articulated trucks (80,000 and 78,000 km) but for two-axle trucks the Indonesian trucks appeared to achieve a higher annual utilisation (82,000 km compared with 60,000 km). In part these differences are a reflection of the mean trip distances of the different types of trucks. Much higher utihsation rates were recorded in the Pakistan surveys where all truck types travelled, on average, more than 110,000 km. Despite this the survey data for Tanzania and Indonesia compares well with the utilisation rates found in Francophone Africa where articulated vehicles achieved only 50,000 km and three-axle trucks 35,000 km per year (Nzet and Hine, 1993). Information on mean vehicle loads (for non-tankers) is given in Figure 14. Here it can be seen that Pakistan achieved consistently higher loads, for each vehicle type, than the other two countries. Two-axle Indonesian trucks Pflvate TransWfl ~. 50.4% =lrdr.er2.0% ~pa” Consignor found elsewhere 4.4% ~ S@te mn-Tramw~% ~reight ~ge”t = Smte TransWti CO. {//////A 0.4% 16.6% I//////r ti/// 201% Pdvate non-Trans~ti b. \ Y Fig. 10 Truck Ownership - hdonesia Oiipatch Clefi 37.2% In contrast in Pakistan over 60 per cent of all drivers found their loadsthrough freight agents, of the remainder the vast majority found their own loads. Very little use was made of company dispatch clerks or of truck parks to find loads. 4.3 ~HICLE AGE, UTILISATION, LOAD~G AND EMPTY RUNN~G Estimates of mean vehicle age, for different vehicle types are given in Table 4. This shows that for the larger vehicles Tanzania has an older vehicle fleet than Indonesia. In comparison at the time of the Pakistan surveys mean vehicle age was 9 yrs. Estimates of annual vehicle utilisation are given in Figure 13. The figures were calculated from drivers’ estimates of their own utilisation. In order to reduce the effect of unrepresentative extreme values the estimates shown for Tanzania and Indonesia in Figure 13 are based on median values. The figures show tittle difference between TanzaFig. 11 Methods of finding loads (for non-bnkers) - Tanzania Freight Agent tinsignor found at Vuck park 4 Q% srfouti bd 20.2% Other 5.3% 54.4% Fig. 12 Methods of finding loads (for non-~kers) - hdonesia 10 TABLE 4 Estimated mean vehicle age (yems) Truck type 2 ale 3 wle articulated Tanzania 7.2 9.0 8.6 Indonesia 7.3 6.5 6.5 160 I 40 20 I 100 80 60 40 20 0 1 2 Nle trucks 3 hle tmcks Articulated trucks I ❑ Tanzania ❑ Indonesia ❑ Pakistan Fig. 13 Annual truck utilisation (1000’s kms) 30 25 ~ 20 -[ 2 hle trucks 3 hle trucks Articulated trucks ❑ Tanzania ❑ Indonesia ❑ Pakistan I Fig. 14 Mean load (tonnes) 11 TABLE 5 Extent of vehicle overloading mean maximum mean mean amount permitted load load per cent of overloading tonnes tonnes overloaded tonnes Tanzania 2 axle 8.5 7.1 0 0 3 axle 16.5 10.5 0 0 artics 33.1 20.6 7 9.8 Indonesia 2 axle 6.4 5.6 13 2.1 3 axle 14.9 15.6 19 14.2 tics 24.2 21.7 19 8.9 -\ l— Tanzania I \ \ \ -~ .- . . . . .% <. ””.””.”” . .. . . -.. ,.. . . --. . . . . . . . . . . ..O \- . . . . --- . ..- -. .+ ~ . . . . . . ..” I I .~v4”z”_ _ ( 100 350 650 900 >1000 Ttip distance (kms) Fig. 15 Recorded emp& running (per cent) were found to have much lower maximum permitted loads appears to have a higher level of empty running than than those in Tanzania or Pakistan and this accounts for the particularly low load factors for these vehicles (5.6 tonnes compared with 7.1 and 8.2 for Tanzania and Pakistan). The mean maximum permitted load recorded for Tanzanian three-axle trucks was found to be relatively high at 16.5 tonnes however the mean recorded load (10.5 tonnes) was much lower than for three-role trucks in the other two countries. An analysis of vehicle overloading was carried out for Tanzania and Indonesia and the results presented in Table 5. Here it can be seen that Indonesian trucks were much more hkely to be overloaded than Tanzanian trucks. The extent of empty running is shown in Figure 15.Here it can be seen that apart from the shortest trips Tanzania Pakistan or Indonesia. However because the Indonesian survey was not based on a pure random sample (seeSection 4.1 above) there are grounds to suspect that the level of empty running has been underestimated. In comparison with other surveys the level of empty running in Tanzania does appear to be well below that found in Ctina (Hine et al 1995) and also below that found in Francophone Africa (Rizet and Hine, 1993). 4.4 JOU~Y SPEEDS AND FUEL CONS~PTION During the surveys drivers were asked to give their overall journey times, including rest stops, and from this data overalljourney speeds were estimated. This data is given in 12 Figure 16. Here it can be seen that Tanzania has much higherjoumey speeds than the other two countries.Journey speeds appear to be two to three times those in Indonesia and up to twice the speeds of Pakistan. Overalljourney speeds are afunction of anumberof factors including rest stops, desired speed under ideal conditions, road geometry and traffic congestion. The relatively high speeds in Tanzania probably reflect a high desired speed in combination with low traffic congestion. Research carried out by the Transport Research Laboratory in Kenya ~ide et al, 1975) found that spottruck speedsfor trucks on good condition wide, flat, straight paved roads without congestion were about 68 kph. In contrast research carried out in Pakistan (Majeed 1980) found that, for a range of sites, the maximum mean spot speeds for trucks was 59 kph while the average mean speed was 52 kph. While research carried out in Indonesia (B&g, 1994) suggested that desired speedsfor medium and heavy trucks were 69 kph and 66 kph respectively. The very low journey speeds found in Indonesia probably relates to long rest times and high congestion. Although much of central Java is mountainous the northern corridor route is fairly flat and in good condition. Most main roads in Pakistan are flat and straight however they do not have the same high trtilc levels asin Indonesia and congestion is not a severe problem. The low journey speeds in Pakistan are much more a reflection of the low desired speeds than any other factor. Resttimes inPakistan are probably not as long as in Indonesia or Tanzania because amajority of trucks havemore than two drivers and a high utilisation is achieved as a result. Fuel consumption is a function of driving speed where minimum consumption for a truck is achieved at around 30 to 40 kph (Hide et d, 1975). Average fuel consumption, per vehicle kilometre, is given inFigure 17. Here it can be seen that Pakistti vehicles have lower fuel consumption rates than vehicles from the other countries. However once the higher load factors are taken into account the differences are much greater. For similar vehicle types Pakistani trucks have fuel consumption rates, per tonne-km (tkrn), that are approximately half those of Tanztia and well below those of Indonesia (see Figure 18). 4.5 FmIGHT TARIFFS Survey data collected on freight tariffs (for non-tankers) in Tanzania and Indonesia are shown in the graphs of Figures 19 to 24. Only a minority of drivers had knowledge of the tariff charged and very little rehable data could be collected for the larger loads in Tanzania. The graphs show a wide variation in tariff rates, per tkrn; similar wide variations have been found in Pakistan and in other surveys. For two and three-axle trucks in Tanzania the average rates found were about 57 Tanzanian shilhngs (Tsh) (9.3 US cents) per tkm. After adjusting for inflation very similar rates have been reported for inter-district movements in both 1993 and 1994. For example, in 1993 for 7-8 tonnes trucks rates were 63.5 Tsh per tkm; for 10-12 tonnes: 53.1 Tsh per tkm; md for 15 tonnes: 50.5 Tsh per tkm. In 1994 a mean rate of 52.4 Tsh per tkm was recorded (National Transport Corporation,1993 and 1995). For the heavier vehicles in Tanzania a mean rate of around 50 Tsh (8.1 US cents) per tkm was recorded but, as can be seen from the graph, this cannot be taken as a rehable 60 t 50 t 40 10 0 2 hle trucks 3 hle trucks Articulated trucks I ❑ Tanzania ❑ Indonesia ❑ Pahstan I Fig. 16 Average journey speeds 13 I I 1,11111, I 1 200 150 100 50 0 700 600 500 400 300 A A A A 200 100 0 0 0 500 1,000 1,500 2,000 2,500 Distance (kms) Fig. 20 Tamanian tariffs for loads greater than 13 tonnes I A . . A . . A 200 400 600 800 1,000 1,200 1,400 Distance (kms) Fig. 21 Indonesian tariffs for loads between 3 and 13 tonnes 140 I 120 100 80 60 40 20 A .“ A A A A A 2 o~ I I I 400 500 600 700 800 900 1,000 1,100 1,20C Distance (kms) Fig. 22 Indonesian briffs for loads greater than 13 tonnes 15 35 I — Tanzania --------- Indonesia ——— Pakistan 0 E a v 10 I . . .-”-....... w . . . . -p. . ““.----- 35 --- “-. . . . . ...=-- ---- -— --- -- —-- ----- . . . . . . . .. . . ----- -. 01’ I I I I I I 1 100 150 350 600 650 900 1300 1400 w c c o Distance (kms) Fig. 23 Tariffs for loads of 3 to 13 tonnes (US cents per tkm) m -------— __ ------- _, n I I I I I I I 1 “ 100 350 500 650 700 900 1100 1600 Distance (kms) Fig. 24 Tariffa for loads of more than 13 tonnes (US cents per tkm) estimate. Discussion, held in August 1995, with transporters and representatives of the World Food Programme quoted rates in the range of 33 to 55 Tsh per tkrn for the movement of heavy loads on international routes, with an average of about 44 Tsh (7.1 US cents) per tkm. Rates as high as 70 Tsh per tkm were quoted for late 1994 and early 1995 for traffic to Burundi. Many transporters complained that there had been a substantial reduction in international traffic from Dar es Salaam and that rates had fallen as a result. An analysis of the survey data showed that, for trips of less than 500 km tariff rates for two and three-axle trucks traveling to Dares Salaam were in the region of 50 per cent higher than for trucks travelhng in the opposite direction. In contrast, for trips that were greater than 1000 km and traveling from Dar es Sdaarn tariff rates were more than twice the rates of those traveling towards Dares Salaam For articulated vehicles there were insufficient data to identify any direction differences in tariffs. In Indonesia the average rates found in the survey were, per vehicle km, 619 Rupiah (Rs) (27 US cents) for two-axle trucks, 919 Rs (40 US cents) for three-axle trucks and 1178 Rs (52 US cents) for articulated vehicles. Expressed per tonne km the corresponding figures are: 115 Rs (5 US cents), 70 Rs (3 US cents) and 65 Rs (2.9 US cents). In comparison higher rates were quoted by transporters moving containers (2000 Rs perkm, and 2300 to 2500 Rs perkm for 20ft and 40 ft containers respectively). However higher rates for containers are to be expected because it is often difficult to arrange backbauls for containers. For loads of 3to 13tonnes, Figure 23 shows how tariff rates for Tanzania, Indonesia andPakistan compare for different distances. The tariff rates are expressed in US cents. The following rates of exchange used were: Tanzania Sh 616 to US$ 1,Indonesim Rs 2285 to US$ 1 andPtistan Rs 31.32 to US$ 1 with an additional adjustment factor of 2.328 for inflation in Pakistan to convert 1986 prices to 1995 prices. The graph shows that tariff rates in Tanzania are much higher than for Ptistan with Indonesia taking an intermediate position; the differences are most marked for short distance for loads in the 3 to 13 tonne range. 16 Table 6 provides an overall comparison of freight tariff revenues. Here the tariff revenues quoted are calculated for all trucks, excluding tankers, carrying loadsin the different load categories using the following formulae: tariff revenue per km= toti tariff revenue/ total loaded distance tariff revenue per tonne km= total tariff revenue/ sum ( loaded distance* load weight) The Table indicates that, on an itiation adjusted basis, overall tariff revenues, per tkrn, in Tanzania are in region of three to five times those of Pakistan and two to four and a hdf times those of Indonesia. The differences between Tanzania and Pakistan are comparable with the differences found in earlier studies between Francophone Africa and Pakistan (Rizet tid Hine, 1993). For heavier loads (above 13 tonnes) overall tariff revenues in Indonesia are very similar to those found in Pakistan. In this category, the average load factors for Pakistan are lessthan for the other countries. This is because, at the time of the survey in Pakistan, there was a higher ratio of three-axle trucks compared with articulated trucks than were found in the Tanzanian and Indonesian surveys. For the smaller loads overall tariff revenues, per tkm, are around one third higher in Indonesia than in Pakistan. The main reason for this relates to the much lower loads found in Indonesia. 5. AN ANALYSIS OF VEHICLE OPE~TING COSTS In order to explain the differences in tariffs found in the previous section it is useful to carry out an analysis of vehicle operating costs (VOCS). Inter-country variations in vehicle operating costs are mainly derived from differences in: a) component input prices b) vehicle productivity c) rate of consumption of fuel, tyres and spare parts In the results of the drivers’ interview surveys given in Section 4 differences in vehicle productivity and fuel consumption were presented. In the following sections component prices and tyres and spare parts consumption will be considered together with estimates of how total vehicle operating costs are built up for representative vehicles of the different countries. 5.1 A COMPARISON OF COMPONENT PRICES ‘ In a comparison of Pakistani and Francophone vehicle operating costs it was found that more than hdf of the difference in operating costs related to the initial difference in component input costs (See Rizet, 1990 and Rizet and Hine, 1993). In Table 7 typical component price data are provided for Tanzania, Indonesia and Pakistan. The exchange rates and inflation adjustments to bring to 1995 US dollar prices are given in Section 4.5. Total taxation (including import duties) on vehicles generally varies according to vehicle size or capacity, its origin and where it was built. In addition inPakistan the source of foreign exchange finance dso influenced the total tax rates. Estimates of tax rates are given in Table 8. Table 7 shows that, with the exception of crew costs, Tanzania is by far the most expensive for dl items. Fuel prices are over two and a hdf times the price in Indonesia and a third more than in Pakistan (adjusted for inflation). TABLE 6 A comparison of overall tariff revenues per vehicle and per tonne km, 1995 prices Mean Mean Tariff Revenues Observations Distance Load US cents US cents No. km tonnes per km per tonne-km 3-13 toMes Tanzania 44 616 7.8 71 8.6 Indonesia 66 850 5.9 23 3.8 Pakistan 1463 586 8.6 25 2.8 13.1-50 tonnes Tanzania 14 907 23.6 222 9.6 Indonesia 47 859 23.6 51 2.1 Pakistan 259 961 21.1 42 2.0 17 TABLE 7 Comparative component prices (US$ 1995) Item Tanzania Indonesia Pakistan Two-axle truck Three-axle truck Tractor & semi Trailer Truck tyre for 2 axle truck for tractor unit Diesel per litre Lubricants per ltr. Maintenance Labour Per hr. Crew costs per month for 2 axle truck for artic. 64,900 97,400 135,000 292 357 0.438 2.27 2.11 136 203 22,300 n.a. 73,100 142 201 0.166 1.66 2.19 219 263 24,200 37,800 46,500 169 261 0.32 1.0 0.7 320 400 TABLE 8 Estimated tax rates (per cent) Item Tanzania Indonesia Pakistan Two-axle truck 50 30 35 Tractor unit 5 34 20 Truck tyre 80 23 32 Diesel 3 11 11 Indonesia is amajoroil producer and has adopted a particularly cheap fuel policy. The higher crew costs in Pakistan are, in part, a reflection of the higher crew manning levels (see Section 4.2 above). The comparison of vehicle prices is somewhat difficult because in all countries there is a range of makes and models and in the surveys it was not possible to collect precise specifications. A selection of typical mid range vehicle prices were chosen to be as representative for Tanzania and Indonesia. For Pakistan the Bedford truck (7 tonne carrying capacity) was used as a representative vehicle; at the time of the surveys Bedfords accounted for 77 percent of the total fleet. In spite of the difficulties in the comparison the differences in vehicle prices are sufficiently large to suggest that in Tanzania truck prices could well be over twice the price of comparable vehicles found in Pakistan and Indonesia. The taxation data presented in Table 8 suggestthatthe high costsfound inTanzmiaare not primarily brought about by higher vehicle taxation, Part of the reason for the very low prices for tractor units in Pakistan relates to the practice of locally converting very simple two-axle trucks into tractor units. In Tanzania (and Indonesia) far more conventional factory made tractor units are in use. In both Pakistan and Indonesia a range of modifications are made to vehicles, after they have left the factory in order to take heavier loads. Modifications include the fitting of stronger springs and the addition of extra steel to strengthen the chassis. Tyre prices appear particularly low in Indonesia and maintenance labour costs are particularly low in Pakistan. The reason for the latter is that most repairs in Pakistan are carried out at small, informal workshops which have very htited overheads. In contrast more conventional garages and owners’ workshops are used in the other countries. 18 5.2 VEHICLE MA~TENANCE Am TYRE COSTS In the main roadside surveys carried out in Tanzania and Indonesia drivers were asked to estimate their maintenance costs, but not tyre wear. However the data was felt to be too unrehable to use on its own and so other sources of information were also used to provide estimates of vehicle maintenance and tyre costs. The following four sources of information were used:- i) drivers’ estimates ii) transporters’ estimates iii) estimates from accounts and written report iv) estimates derived from the road investment planning models, HDM and RTIM Tables 9 and 10 give different estimates of maintenance and tyre consumption costs. Survey interviewers in Indonesia reported that they had bttle faith in the driver’s estimates of maintenance costs and an examination of the data confirms that fittle credence can be given to this data. For example articulated trucks were reported to have lower maintenance costs than two-axle costs, despite the fact that they were over three times as expensive. Drivers’ estimates of the maintenance costs of articulated vehiclesin Tanzania also appear to be suspect in view of the reported close correspondence with the maintenance costs of two-axle trucks. In contrmt the tivers estimates of maintenance costs inPakistan were found to be both intemdly consistent (ie variations with vehicle age and between vehicle types) and consistent with written records. A com~tison of the more reliable estimates shows that TABLE 9 Estimates of vehicle maintenance costs ~S cents per km, 1995 prices) Item Tanzania Indonesia Pakistan Two-axle trucks Drivers 6.4 2.8 2.7 Transporters 5.3 5.8 (12 yrs) Accounts 6.7 4.3 1.8 (4 yrs) HDM (Brazil) 21.5 7.6 10.2 R~M 33.6 19.5 Articulated trucks Drivers 6.7 1.6 4.2 (4 yrs) Transporters 31.7 3.0 (new) Accounts HDM (Brazil) 42.4 29.2 26.7 R~M 74.7 TABLE 10 Estimates of tyre consumption costs ~S cents per km, 1995 prices) Item Tanzania Indonesia Pakistan Two-axle trucks Transporters 7.3 1.9 Accounts 8.2 1.2 1.1 HDM (Brazil) 3.6 1.7 2.5 R~M 3.5 2.9 Ardculated trucks Transporters 17.5 8.3 HDM (Brazil) 2.1 R~M 9.6 4.0 4.9 14.8 19 Pakistan had particularly low maintenance costs. In Pakistan nearly dl vehicles are maintained by numerous informal workshops and the driver is predominantly responsible for organizing and paying for repair and maintenance. In contrast in both Tanzania and Indonesia the transporters’ surveys reported that most trucks were maintained by the truck owner’s own maintenance yard. In most cases @particularlyfor the larger fleets) if a truck broke down with a serious problem, some way away from the home base, the driver was expected to contact the base and await for the fleet repair vehicle and a mechanic to assist. This rarely happened in Pakistan. Application of the road planning models, RTIM (T~’s Road Transport Investment Model, based on research in Kenya and the Caribbean) and HDM (the World Bank’s Highway Design and Maintenance Standard’s Model based on research in Brazil) tended to give the highest estimates of vehicle maintenance costs which were often difficult to reconcile with the other data and the overall estimates of tariff revenue and operating costs. In estimating the ~M and RTIM maintenance costs shown in Table 9 it was assumed that the main road roughness levels were 4.5 IN in Tanzania, 3.5 IW in Indonesia and 6 IRI in Pakistan. While these road roughness values may reflect conditions on the main paved network connecting most of the surveys’ origins and destinations it is recognised that much higher values will apply to the unpaved main road sections in Tanzania andPtistan. Ehigherroughness values had been used then, obviously, the models would have predicted even higher maintenance and tyre costs. 5.3 A COMPARISON OF O~RALL OPERAT~G COSTS Overall approximate estimates of vehicle operating costs are given in Table 11 for two-axle trucks. Because of the lack of reliable data it was not possible to provide the same cost breakdown for the larger vehicles. The maintenance and tyre costs are derived from samples of accounting data. To enable a comparison to be made with tariff revenues, which are assumed to be constant in real terms over the life of the vehicle, the interest rate specified in the calculation of capital costs was assumed to be net of itiation. A three per cent real interest rate was assumed in dl cases. The revenue data (expressed per vehicle-km) reflects an assumed level of empty running (25 percent for Tanzania and 15 per cent for Indonesia and Pakistan). The data clearly show that, apart from the crew, all of the cost components are higher in Tanzania, when expressed on a per km basis. One of the main factors keeping Indonesian tariffs @er vehicle-km) low relates to the particularly low fuel prices in that country. The Indonesia two-axle vehicles have much lower loading factors than in Pakistan and as a result their cost per tonne-km are higher than in Pakistan (See Section 4.5 for a comparison of tariffs). If, in the analysis, uncalibrated data from the road planning models HDM or RTIM had been used to calculate maintenance costs then two-axle vehicles would have been shown to be unprofitable in all cases, and by particularly wide margins in Tanzania and Pakistan. In the case of Pakistan this issue has already been examinedin detail (See Hine and Chilver, 1994). However there are some indications thatfor articulated vehicles it would be somewhat easier to accommodate HDM or RTIM derived maintenance costs within the revenue estimates. 6. POSSIBLE SOLUTIONS TO ENCOURAGE THE DEVELOPMENT OF LOW COST TRANSPORT IN The results confirm earlier research which indicated that African freight transport tariffs and costs were much higher than in Asia. In this case long distance freight transport tariff rates and overrdl tariff revenues per tonne-km in Tanzania are between two to five times those of Indonesia and Pakistan for different vehicle types and load weight categories. The reasons for the higher costs and tariffs relate to the following combination of factors: higher input prices, higher maintenance costs, lower utilisation, higher levels of empty running and higher fuel consumption. High levels of taxation and higher operating costs due to road condition do not appear to be major reasons for the marked differences in operating costs. Possible solutions to help reduce transport costs and tariffs in Tanzania might be along the following lines. a) Encourage all concerned, including transporters, consignors and officials to realize that transport tariffs could be much lower. Competition is the main mechanism by which costs are kept under control however there are reasons which sometimesprevent competition fromworking. Such factors include poor market information; formal or informal cartels; state controls; a low or fluctuating level of demand; and high costs, or other impediments, of entry into the industry (including high costs of finance). b) It is not known exactly why initial vehicle costs are so much higher in Tanzania (and in other African countries) than in Indonesia orPakistan. However there is evidence to suggest that high import prices to Africa are not unique to the transport sector (Yeats, 1989). Overall, vehicle taxation only appears to be a minor factor in the differences in the domestic vehicle prices. It seems that the structure of commercial vehicle supply is far more important. High prices may be the result of exclusive 20 TABLE 11 Estimated composition of operating costs for two axle trucks (US cents per km, 1995 prices) Tanzania Indonesia Pakistan Capital costs 10.6 2.7 1.8 Fuel 15.4 5.8 9.3 Crew 2.7 3.2 3.2 Oil 1.0 0.7 1.0 Maintenance 6.1 4.3 2.2 Tyres 7.8 1.2 1.1 Overheads 6.5 1.8 2.4 Total 50.1 19.7 21.0 Revenue 56.8 20.0 21.8 dealerships in vehicle importing combined with the small market size. Also, some vehicles have been brought into Tanzania under aid terms and hence price competition for these vehicles may have been firnited. Often it seems that in Africa expensive vehicles are purchased with very high specifications whereas in Asia deliberate attempts are made to buy very cheap vehicle types and then, if necessary, to locally modify them after purchase. Encouragement should be given to dl involved in vehicle purchase to look very carefully at initial vehicle costs. and Tanzania only a relatively small proportion of loads are arranged via freight agents. Tariff levels can be kept to a minimum when freight forwarding agents are in active competition with each another. Unfortunately freight forwarding agents are sometimes viewed as parasites within the freight transport industry. In fact recognition and encouragement ought to be given to agents for the part that they can play in increasing the efficiency of the industry. c) Spare parts costs will tend to be very expensive when ~ In many African countries utihsation rates are kept low exclusive dealerships are involved in keeping a wide stock of supplies for very small vehicle numbers. In Indonesia and Pakistan the large numbers of similar vehicle types help to keep the costs of spares low. In addition it is known that in Asia many parts will be manufactured locally (either in batches in small factories or by a mechanic on the spot) for the local market. It is recognised that locrdly made parts may not be up to the same standards as imported parts from Japan or Europe. However, in Pakistan operators have found that by running their trucks very slowly, and changing the oil very often, they can run their vehicles both cheaply and safely with using Iocdly made parts. Another benefit of slow running speeds, frequent oil changes and close attention to routine maintenance appears to be a much lower incidence of major engine overhauls than in other countries. d) Vehicle utilisation can be increased by using two drivers. Pakistti vehicles achieve a high utilisation with two drivers by driving both night and day. A sleeping area allows one to sleep while the other drives. Night driving need not be dangerous if a vehicle’s lights are in working order and the vehicle travels slowly. e) Delays incurred by looking for loads are minitised and empty running reduced by an extensive use of freight forwarding agents in Pakistan whereas in both Indonesia and freight rates high by the action of formal or informal associations of truck operators. Although this may not be a major problem in Tanzania Government authorities should be made aware of how such cartels reduce efficiency and increase costs. The power of trucking associations in other countries is most often manifest in up-country locations at lorry parks. They operate to restrict supply, and share out the available demand, through vehicles queuing at lorry parks. To hire a truck a customer has to take the truck atthe head of the queue. The customer is prevented from negotiating a lower rate with other vehicles in the queue. Overall operating costs are kept high because the more efficient operators, with low variable costs, are prevented from increasing their utifisation by undercutting other higher cost operators. In this way older vehicles with higher maintenance costs are kept in business instead of being driven out by competition. g) The survey showed that fuel consumption per tonne km was very high in Tanzania. Part of the reasons probably relate to the higher driving speeds in Tanzania. Slow driving speeds are also likely to have other benefits in both lower maintenance costs and lower accident rates. The Government, transporters and drivers should be made much more aware of the benefits of slow and careful driving. 21 h) One of the main factors keeping operating costs low in Pakistan relates to the large measure of responsibility given to the driver. He is aware of the costs and revenues of the truck he runs because he finds loads, arranges and pays for maintenmce, pays for fuel and keeps accounts. He is very aware of the benefits of both driving slowly and of paying close attention to routine maintenance. 7. ACKNOWLEDGEMENTS The work described in this report was carried out under the Technology Development Research programme of the Engineering Division, Overseas Development Administration. The work was undertaken by staff of Mott MacDonald and the Transport Research Laborato~ together with assistance from Ministries of Works, Communications and Highways, in Tanzania and the Institute of Road Engineering, at Bandung in Indonesia. 8. REFERENCES B~G KL (1994). Indonesian highway capacity manual project phase 2:Interurban roads, Ministry of Public Works and Sweroad, Bandung. ~NE JL and AS CH~~R (1991). Pakistan road freight industry: An overview. TRRL Research Report No. 314. Transport and Road Research Laboratory, Crowthorne. ~NE JL and AS CH~~R (1994). Pakistan road freight industry: An analysis of tariffs, revenues and costs. TRL Project Report No. 94. Transport Research Laboratory, Crowthome. ~NE J, BARTON A, CHEN GOUJ~G and WANG WENLONG (1995). The scope for improving the efficiency ofroad freight transport inChina. 7th World Conference on Transport Research, Sydney. Mm J and C REET (1991). Halving Africa’s freight transport costs: could itbe done ? International Symposium Transport and Communications in Africa, Brussels. mDEH, ABAYNAYAKA SW,SAYER I andRJ WYATT (1975). The Kenya road transport cost study: research on vehicle operating costs.TRRL Laborato~ Report No. 672. Transport and Road Research Laboratory, Crowthome. MAJEED A (1980). Highway SpeedSurvey. NTRCReport No. 5f. National Transport Research Centre, Islarnabad. MINSTRY OF PUBLIC WORKS (1993). Indonesian Highway Statistics. Directorate of General Highways, Ministry of Wblic Works, Jakarta. MI~STRY OF WORKS (1993). Presentation of the Second Integrated RoadsProject. Ministry of Works, Ministry of Communications and Transport, Dares Salaam. WET C (1990). Coftts et surcoOts du camrnionnage en Afrique - Application d’une m6thode de comparison des cotits entre pays. Les Cahiers Scientifiques du Transport, 21,85-102. RUET C and JL ~NE (1993). A comparison of the costs and productivity of road freight transport in Africa and Pakistan. Transport Reviews, 13(2), 151-165. NATIONAL TRANSPORT CORPORATION (1995). Restructuring road transport services in Tanzania. ~is Transport Management Services, Dar es Salaam. NATIONAL TRANSPORT CORPORATION (1993). Road transport study. Economic Research Bureau, University of Dares Sdaarn. WORLD BANK (1995). World Development Report. World Bank, Washington DC. YEATS AJ (1989). Do African countries pay more for imports? Yes. Policy, Planning and Research Working Paper No. WPS 265. The World Bank, Washington DC. 22 APPENDIX 1 1. Questionn& No. 2. Survey Station No. 3. Date 4. ~tion: Dar es Srdaam 5. Registration No: ~o or From) 6. Truck base Country 7. Truck Make 8. Cotilguration: 9. * body type: a) Truck a) bw sided b) Tmck & Ttier b) High sided c) Tractor & Semi Ttier c) Hat d) Tmctor unit otiy d) Tanker e) Mom than two tiers e) Tipper Q Other ~ Other 10. No of des 11. Mtium load tons or fitres Don’t know_ 12. Tmck owne~ 13. How many crew m with the truck now? a) Driver (sole or joint owner) Drivers b) Private individti Assistants c) Partnership d) Private Transport Company. e) GovernmentiState Transport Co. ~ Private non-transport Co. (opemting on own account) g) Government non-~sport Co. (opemting on own account) h) Coopemtive i) Other j) Don’t know 14. Driver’s ~htionship to truck: 15. In toti how many drivers share driving this a) Driver is sole owner truck m~larly? Don’t know _ b) Driver has part share c) Reguk employee 16. Over the last month what propotiion of toti d) Casual employ= distance was by the driver(s) now with the truck? e) Rented tick per cent Don’t know O Truck bomwd g) Don’t know Pl~se estiate the following: 17. Age of truck yrs Don’t know 18. Average distance driven per week kms Don’t know— 19. ficluding tym what are the avemge monttiy service and repair costs currency 20. Average fuel used titres per kms or fitres per days 21. Average monttiy tariff income currency 22. Btimated current value of truck currency Don’t know Don’t know— Don’t know — Don’t know — Don’t how — 23 23. badd/Empty ~f empty go to question 34) U haded plme amwer the fo~owing qu~ions: 24. 25. 26. 28. Commodity or type of load Don’t know had weighdVol: tons or titres Don’t know had origin 27. bad destination Journey distance kms Don’t know 29. ToM otigk to destination journey tfie (with stoPs) _ ~s _ days Don’t know _ 30. Between dropping the kst load and picking up this load how long was truck empty? hrs _ days Don’t know _ 31. 32. 33. Betwmn dropptig the last load and picking up this load how much empty mnning was there? kms Don’t know ~timate tariff for this load currency Don’t know What method was used to fmd this load? Don’t know a) Driver found load b) Consignor found truck at lorry park c) Consignor found truck elsewhere d) tiugh a freight agent e) Instructions from Transporter’s dispatch clerk ~ Other methods g) Don’t know 34. Journey origin 35. Destination 36. Distance kms 37. fitirnated empty journey tfie (with stops) _ ~ _ days Don’t know _ 38. Empty journey pupse: a) boking for load b) Return to base c) Setie accounts d) Repairs e) Personal reasons ~ Other Thank you for your cooperation. 24 1. 3. 5. 6. @estionnaire No._ 2. Date Company Name 4. hcation Type of company a) Private individud b) Partnership d) tivate Transport Company. e) GovernrnentiState Transport Co. ~ Private non-transport Co. (opemting on own account) g) Government non-tmsport Co. (operating on own aaunt) h) Coopemtive i) Other What is the main type of transport business carried out by the company? Pl~se describe: 7. What are the principal wmmodities moved? 8. What are the main destinations? 9. What proportion of business is obtained by the foflowing ways: a) Through freight agents b) By telephone/fu etc from customers c) By trucks waiting at lorry parks d) By drivers finding theh own loads e) By wing goods for the non-transport side of the company ~ Other channels 10. Plmse fist the numbers of different types of vehicles currentiy in 2 tie rigids _ 3 tie rigids _ Tractors and semi truck and tiers vehicles with more than one ttier _% % % ‘% ‘% _% ope~tion with the company. tiers 25 11. Plase give examples of the current estimated complete purchase price in T~ania for new and second hand vehicles (if used): i) 2 tie truck make max load tons body (new/ _ yrs old) ii) 3 tie truck make max load tons body (new/ ~rs old) M) @ctor unitl) make hp* gcw* tons (new/ ~rs old) iv) tractor unit2) make hp* gcw* tons (new/ ~rs old) v) semi ttier ties _ body (new/_ yrs old) vi) drawbar tier body (new/ _ y~ old) hp = horse power, gcw = gross combination weight 12. Please give examples of the cumnt freight rates to the most fquent origins and destinations servd. i) Fmm to Distance kms Tariff per tons load ii) From to Distance tis Tariff per tons load iii) From to Distance kms Tariff per tons load iv) From to Distance kms Tariff per tons load v) From to Distice hS Tariff per tons load 13. How are repairs and servicing carrid out? a) Own ‘workshop facfities b) Use of commercial gmge c) titure 14. mat happens if a vehicle b- down away from base? a) Driver approaches lod gmge b) Driver ctis base md own mobtie repti truck is sent out c) Drivers CWSbase and base ~ges for l@ garage to coflect and repair d) Other, pl=se expti 15. In manning the trucks what poticy does the fm have? a) Wocate one main driver to a vehicle b) tiocate two or th drivers to a vehicle for shift work c) Wow many drivers to drive each vehicle d) Other, please explain 26 16, Does the fm have any restriction on night-time driving? (yes/no) If there is a restriction what is it and why ? Plase explain. 17. Over the last five y- have any trucks been stolen from the fm? 18, mat annual deage do you expwt from your vehicles? Type a) kmslyr Type b) kmslyr Type c) kmslyr d) kmslyr 19. For different vehicle types opemtd by the fm a you give some estimates of current vehicle opemting costs. i) Truck type Tyre price truck tier A driver’s wages and Wowances per An assistant’s wages and Wowances per Avemge budget for maintenance. & repair per kms Average budget for fuel per krns Avenge budget for tyres per krns ii) Truck type ~re price truck tier A driver’s wages and ~owances per An assistant’s wages and Wowanms per Average budget for maintenance & repair per krns Avemge budget for fuel per kms Avemge budget for tyres per kms 20. What are the main problems facing tisport operators in T-a and what are the main constits preventing more efficient transport opemtions? Pl~se explain: Thank you for cooperating with this survey. 27