~-d Transport& RoadResearchLaborato~ L~5r~~] I Il:i:il Research Report 333 Pakistan road freight industry: The productivity and time use of commercial vehicles ( by J L Hine TRANSPORTAND ROAD RESEARCH LABORATORY Depadment of Transpofi RESEARCH REPORT 333 PAKISTAN ROAD FREIGHT INDUSTRY: THE PRODUCTIVITY AND TIME USE OF COMMERCIAL VEHICLES by J L Hine Crown Copyright 1991. The work described in this report forms part of the programmecarried out for the Overseas Development Administration, but the views expressed are not necessarily those of the Administration. Extracts from the text may be reproduced, except for commercial purposes, provided the source is acknowledged. Overseas Unit Transport and Road Research Laboratory Crowthorne, Berkshire, RG11 6AU 1991 ISSN 0266-5247 Digest of Research Report 333 1991 PAKISTAN ROAD FREIGHT INDUSTRY THE PRODUCTIVITY AND TIME USE OF COMMERCIAL VEHICLES by J L Hine Introduction A study of Pakistan’s freight transport industry was carried out by the Overseas Unit of the Transport and Road Research Laboratory in cooperation with the National Transport Research Centre, Islarnabad. As part of this study a ‘Vehicle Activity Survey’ was undertaken in which the different vehicle activities (i.e. loading, unloading, moving, resting) were recorded and timed as they occurred. The data were collected by both truck drivers and survey staff as they travelled with each vehicle on a continuous basis for periods lasting from five days to over four weeks. hformation on distances travelled, costs incurred and freight tariffs charged was also recorded. In total data on 45 survey periods were collected covering 405 loaded and 327 empty trips. The main purpose of this survey and the subsequent analysis was to investigate the following three aspects of time related costs: a) The use of time savings following road investment. b) The amount of time a vehicle spends in different activities (e.g. in traveling, loading, unloading, or at rest). c) The relative importance of time and distance in the explanation of costs and tariffs. Findings The use of time savings From a regression analysis relating trips made per day to mean trip working time elasticities between -0.84 to-1 were found. These imply that in Pakistan’s conditions, journey time savings following road investment are likely to be fully translated into extra trips. Time distribution of vehicle activities Freight vehicles were in movement for 40 per cent of the time and including loading and unloading vehicles were in active use for over 12 hours per day. The relative importance of time and distance Vehicle operating costs depend on both time and distance but there is uncertainty as to the relative importance of each. Pakistan has a very competitive freight transport industry so tariffs probably reflect underlying costs. A dA Department of Transport < A statistical analysis of the relationship between tariff, time and distance was undertaken but the high degree of correlation between time and distance made it difficult to identify the separate explanatory power of either variable. The work described in this Digest forms part of a programme of joint research between the Overseas Unit (Head J S Yerrell) of the Transport and Road Research Laboratory, UK and the National Transport Research Centre (Head M S Swati), Pakistan. If this information is insufficient for your needs a copy of the full Research Report RR333, may be obtained, free of charge (prepaid by the Overseas Development Administration) on written request to the Technical Information and Library Services, Transport and Road Research Laboratory, Old Wokingham Road, Crowthorne, Berkshire. CROWN COPYRIGHT. The views expressed in this Digest are not necessarily those of the Department of Transport. Extracts from the text may be reproduced except for commercial purposes, provided the source is acknowledged. CONTENTS Abstract 1. Introduction 2. The valuation of commercial vehicle time savings 3. The road freight transport industry in Pakistan 4. The vehicle activity survey 4.1 Survey procedure 4.2 The pattern of vehicle use 5. The relationship between trip frequency and trip duration 6. The relationship between revenue and journey time and distance 7. Conclusions 8. Acknowledgements 9. References Appendix A Page 1 1 1 3 3 3 4 6 11 14 14 14 15 PAKISTAN ROAD FREIGHT INDUSTRY: THE PRODUCTIVITY AND TIME USE OF COMMERCIAL VEHICLES ABSTRACT As part of a wider study of freight transport undertaken in Pakistan during 1985-87 a survey of vehicle utilisation was carried out. Different vehicle activities were timed and recorded on a continuous basis for periods lasting from five days to four weeks. Using these data it was possible to investigate three aspects of time related costs:- i) ii) iii) 1 A series of elasticities relating trips to travel time showed that, in Pakistan’s conditions, time savings following road investment are likely to be fully used. An analysis of vehicle time budgets found that freight vehicles were in active use (traveling, loading, and unloading) for more than 12 hours per day. Both time and distance could each independently provide a good explanation of tariffs. Although difficult to identify their separate influences, the best estimate suggests that time and distance account for 45 per cent and 55 per cent of tariffs respectively. INTRODUCTION The rapid growth in traffic volumes experienced by many developing countries has, in recent years, led to an increased interest in providing high capacity roads that are primarily designed to reduce congestion and provide journey time savings. Despite this, the analysis of journey time savings in developing countries has attracted comparatively little research. The two main elements of time-related costs are passenger time, and vehicle and crew time. To improve our understanding of the latter for commercial vehicles this Repoti presents and analyses data collected from a survey carried out in Pakistan during 1985-86. The Repoti examines three aspects of time-related costs. These are:- a) the use of time savings following road investment. Common-sense suggests that if road investment brings about journey time savings then, on average, commercial vehicles should be able to make use of the time saved by making extra trips. However there is disagreement among various investigations (Fleischer 1963, Dawson and Vass 1974, Thomas 1983) about the extent to which this can be achieved. b) c) the amount of time a vehicle spends in different activities (e.g. in traveling, loading, unloading, or at rest). This data is needed to help estimate the benefits from changes in travel time for those cost components that are time dependent. the relative importance of time and distance in the explanation of costs and tariffs. Most components of vehicle operating costs can be separated into those that are clearly time dependent (e.g. driver’s wages) and those that are clearly dependent on distance travelled (e.g. tyre wear). However for some components, (such as depreciation or maintenance), it is known that there is a dependency on both time and distance although there is some uncertainty as to the relative importance of each. An analysis of the extent to which tariffs can be explained by time worked and distance travelled would be useful in checking the validity of current models of vehicle operating costs used in road investment appraisal. To collect information on the time utilisation of freight vehicles a “Vehicle Activity Survey” was undertaken. Data were recorded by both truck drivers and survey staff who travelled with each vehicle for periods lasting between five days and four weeks throughout Pakistan. The purpose of the survey was to record the number of trips made and the time spent by the vehicle in different activities (namely traveling, loading, unloading, resting, and under repair) during each period. The data were recorded 24 hours-per-day. In addition data were also collected on the distance travelled, the revenues earned and the expenses incurred. This Report forms part of a study of Pakistan’s freight transpoti industry that was carried out under a programme of cooperative research between the Overseas Unit of the Transport and Road Research Laboratory and the National Transport Research Centre (Islamabad). A repoti providing a general description of the industry and presenting more information on vehicle operating costs and freight tariffs has been published separately (Hine and Chilver, 1991). 2 THE VALUATION OF COMMERCIAL VEHICLE TIME SAVINGS In order to value the benefits of journey time savings for commercial vehicles arising from road investment assumptions are necessary about the extent to which the potential time savings may be translated into productive use. Some authors, such as Dawson and Vass (1974), 1 assume constant working hours per year in their formulation of vehicle operating costs; this implies that all time savings following a road investment will be fully used. In vehicle operating cost formulations, which do not assume constant working hours, annual time-related costs are divided by distance travelled per year. The latter figure may either be estimated by a formula derived from vehicle speed or estimated by the user. Neither procedure is very satisfactory. For example both Winfrey (1969) and DeWeille(1966) have proposed a formula from which it can be calculated that a one per cent decline in travel time would bring about a one per cent increase in distance travelled. It is assumed that the time spent traveling per day will remain constant. This implies that, with a fall in journey times, the time spent working per day (including traveling, loading and unloading) must rise as more trips are made because of the increased time devoted to loading and unloading. The assumption of constant traveling time is clearly unrealistic for commercial vehicles where loading and unloading time is an important component of the working day. This approach will nearly always tend to overestimate time savings benefits. Other vehicle operating cost model formulations (such as TRRL’s “Road Transport Investment Model” or the World Bank’s “Highway Design Model”) provide the user with the option to predict utilisation following road investment or to choose one of a variety of models to achieve the same result. Again no empirical evidence is presented to show how utilisation is likely to change following road investment. The assumption that travel time savings can be readily translated into extra trips has been questioned by Fleischer (1962) and Thomas (1983). Their empirical investigations suggest that there is an inflexibility in vehicle operations relating to the constraints of drivers’ hours and the scheduling of work which will prevent travel time savings from being fully used, particularly in the short-term. It maybe thought that the more inflexibility there is in vehicle operations then the greater the probability that time savings will not be fully used following road investment. To test this the author built a computer based simulation model in which the operations of a freight vehicle were constrained by a set of drivers’ hours regulations and permitted loading and unloading times. The vehicle was assigned to undertake a random sequence of trips of different lengths. The constraints were shown to impinge on vehicle operations where it was predicted that certain activities (i.e. loading, unloading and short trips) could not be completed within the working day. When this happened the vehicle was assumed to remain idle for the rest of the day and the activity was postponed to the next day. Using different input assumptions the results of the model suggested that while additional constraints on vehicle operations could be expected to reduce efficiency there were no a priori grounds to suggest that they would reduce the probability of using time savings. The model provided as many cases where total idle time fell (i.e. time savings were more than fully used) as cases where total idle time rose (i.e. time savings were not fully used) following a reduction in trip times. The exact outcome was shown to be dependent upon the particular trip length distribution and the set of constraints assumed in the model. There are several ways of trying to estimate how time savings may be used following road investment. One approach is to ask operators to estimate what their response would be, given the predicted reduction in journey times. Even if it be assumed that each operator can forecast correctly his response the results are very difficult to interpret because unless a lot of detail is known there is no simple way of comparing those who say they can make productivity gains with those who say they cannot. It is possible, for example, that time savings will, on average, be fully used if only one operator in twenty can benefit. Another approach is to look at changes in utilisation before and after an investment has been made. This was done by both Fleischer (1963) and by Thomas (1983). Fleischer looked at one company operating on the route between Grants Pass, Portland and Seattle on the West Coast of the USA. There was a series of road improvements which gradually reduced journey times over many years. Fleischer found that an extra trip between Portland and Seattle could only be made after journey times had been reduced to a given amount. On the route between Grants Pass and Potiland he found that it was only when the firm could relocate its depot could the reductions in journey times be translated into an increase in the distance travelled which occurred many years after the first road improvements, though it is not known how representative or complete the case was. Without a total survey of all operators using the route it is not known whether other operators were able to make use of the time savings. Likewise it is possible that in other situations with only one major route impressive productivity gains could occur with only very small time savings. Thomas carried out an historical “before and after” analysis as well as a cross-sectional analysis to determine how vehicle productivity might change following road investment. For the historical analysis, data was collected on vehicle productivity before and after the new Kuala Lumpur - Karak highway was opened in West Malaysia which reduced average vehicle trip times from 4 to 5 hours by about 45 minutes. By collecting data on the operations of various types of commercial vehicles Thomas could not find any dramatic improvements in vehicle productivity following the opening of the new road section. In fact for certain vehicle types the level of trip making per day fell in the Kuala Lumpur area; also during the period which was covered by the data collection large changes were recorded in both the Malaysian commercial vehicle fleet numbers and in the levels of economic activity. Both of these factors could have swamped any beneficial effect of the road on vehicle productivity. In the Malaysian cross-sectional analysis Thomas derived a number of elasticities for different categories of commercial vehicles between trips made per day and average travel time per trip. Most of the Malaysian data was collected from a roadside interview survey in which 2 drivers were asked what trips had been made in the previous 24 hours. The elasticities found ranged from -0.2 to -0.6; i.e. if travel times were reduced by one per cent then trip making could have been expected to rise by between 0.2 and 0.6 per cent. This analysis suggests that the models which predict a one per cent increase in distance travelled for one per cent decline in travel time will clearly overestimate the effects of journey time savings. However to estimate the extent to which vehicle working time savings are used it is necessary to include loading and unloading time within trip times in the calculation of the elasticities. Loading and unloading times were not collected in the Malaysian study and as a result the derived elasticities underestimate the extend to which time savings may be used following journey time savings. It was against this background that it was decided to develop further the cross-sectional analysis using vehicle activity data from Pakistan. 3 THE ROAD FREIGHT TRANSPORT INDUSTRY IN PAKISTAN Currently in Pakistan there are about 45,000 trucks in operation, of which about 95 per cent are privately owned. Road freight transpoti is organised on a free market basis and freight tariffs are determined competitively by supply and demand with little government intervention. The industry is dominated by a large number of individual entrepreneurs operating a ‘hire and reward’ service. Entry into the industry is cheap and easy; there is a relatively lax licensing system and little enforcement of axle load limits or construction and use regulations. During the study several different surveys were undertaken. General information on private road freight transport was collected principally from the Roadside Interview Survey of 3,500 truck drivers. Three quatiers of the trucks surveyed were two-axle Bedford trucks, 14 per cent were two-axle Japanese trucks and the remainder was divided between three-axle rigid vehicles and tractor-trailer combinations. Less than one per cent were owned by a company for its own account operations. Most trucks were found to operate on the basis of picking up business where they could and going from job to job as demand required. It was very common for drivers to work away from base for up to three weeks. The Roadside Interview Survey found that on average drivers of Bedford trucks returned to base after 7 days and returned to their families after 17 days. The driver was found to be responsible for finding the load, for collecting revenue, and for repairing the truck. When he returned to base he was expected to account for the revenue earned and the expenditures incurred. About 80 per cent of drivers were employees, 17 per cent owned their own vehicle and the remaining three per cent owned a part share of the vehicle. There is an extensive network of freight agents who assist drivers to find a load. Virtually all agents were found to be connected by telephone which appeared to play an important pati in their business. Over 60 per cent of loads were found using agents. In the vast majority of cases vehicles could be found for a consignor within one hour. Little specialisation in body types was recorded. About 85 per cent of the Bedfords and 60 per cent of other trucks were high sided, while 8 per cent of Bedfords and 20 per cent of other trucks were tankers. Virtually all freight vehicles in Pakistan were found to have sufficient space within the cab to seat four people. In addition there was usually a purpose built space on top of the cab where a driver or his assistant could rest or sleep whilst the vehicle was in motion. Just over half the trucks were found to have two drivers and one assistant, the remainder had just one driver and one assistant. 4 4.1 THE VEHICLE ACTIVITY SURVEY SURVEY PROCEDURE In Pakistan a high proportion of journeys last for more than one day. In order to get an accurate picture of timing and duration of the different vehicle activities a “Vehicle Activity Survey” was undertaken in which activities were recorded and timed as they occurred. Data were collected from a sample of trucks on a continuous basis for periods lasting from five days to over four weeks. The data were collected by both survey staff and by drivers. The survey staff travelled with their allotted trucks continuously throughout the period, if necessay sleeping on board the truck as it travelled. Where drivers were used to collect data they were paid an additional sum to record their activities. In total over 600 days of useful data were collected, about one fifth of this being recorded by drivers, the rest by survey staff. This comprised 24 periods of data related to conventional two-axle Bedford trucks, seven periods to Bedford tanker trucks and 14 periods to conventional two and three axle Mercedes trucks. The latter trucks were owned and driven by Afghan refugees. Basic data relating to the trucks participating in the Vehicle Activity Survey are given in Table Al of the Appendix. In view of the difficulty encountered in placing the survey staff with vehicles many staff made use of their contacts to find vehicles with which they could work. Where possible, recorders sensitive to vehicle vibration were fitted to provide an independent check on the manual records of the survey staff. Detailed records of the timing of all movements, rest periods, loadings, unloading, waiting periods, emergencies and repairs were collected on a continuous basis. Vehicle stops of less than 15 minutes were ignored. Cooperation was sought from the other drivers and 3 assistants traveling with the vehicle to help record those details which the survey assistant or driver missed whilst he was asleep. Additional data on the distance travelled, costs incurred and revenue earned were also recorded. 4.2 THE PA~ERN OF VEHICLE USE The Vehicle Activity Survey collected data on trucks engaged on a variety of patterns of operation and covered a wide spectrum of trip lengths. The 45 survey periods covered 405 loaded trips and 327 empty trips. Distances were recorded for 92 per cent of the trips and for these trips the mean trip distances, loaded and empty, were 347 kms and 150 kms respectively. Fifteen per cent of loaded trips and 42 per cent of empty trips were under 50 kms while eight per cent of the loaded trips and two per cent of the empty trips were over 1000 kms. To analyse the pattern of vehicle productivity and time use it was necessary, as far as possible, to allocate time spent in different activities into empty and loaded trip periods. For each trip the times spent moving, loading, unloading, resting, or under repair were separately totalled. Because of the long journey distances one trip was usually broken up into a series of movements and rest periods. On average each loaded trip and each empty trip was composed of approximately 5 and 3 separate movements respectively. Because the vehicle was at least partially loaded during loading and unloading these activities were defined to be part of a loaded trip. In the analysis no distinction was made between resting time and waiting time. Waiting to unload was counted as part of a loaded trip while waiting to load after completing an empty movement was counted as part of that empty trip. For a loaded trip to finish the truck would have to be completely unloaded; so a sequence of multiple partial loadings or unloading would all be counted as part of one loaded trip. A loading is necessary to finish an empty trip. So a series of empty movements (made in any direction) was counted as part of one empty trip. Besides the loaded and empty trips there were also periods between two loaded trips when the vehicle did not move empty. In total 63 periods of this type were recorded. Detailed activity data is shown in Table A2 of the Appendix. The mean periods spent in each activity for the whole survey are shown in Table 1. On average loaded trip periods lasted 21.8 hrs and empty trip periods lasted 16.5 hrs. In total trucks were found to be loaded 56 per cent of the time. Rest and waiting time accounted for 30 per cent of loaded trip time, but overall 63 per cent of rest and waiting occurred while the truck was empty. The average duration between the end of loading and arrival at the destination was found to be 14.9 hrs which gives an average loaded journey speed (including intermediate rest periods) of 23 kph. This is the same overall journey speed found for loaded Bedford trucks in the Roadside Interview Survey (Hine and Chilver, 1991) . Using data from the Vehicle Activity Survey an estimate of 109,000 kms was calculated for annual vehicle travel. This is very close to the mean estimate of 112,000 kms annual vehicle travel found for Bedford trucks in the Roadside Interview Survey. The distribution of total loaded and empty trip times found by the survey is shown in Figure 1. Although 70 per cent of loaded trips took less than 24 hrs, 13 per cent lasted longer than 48 hrs. A breakdown of time by activity is shown in Figure 2. This demonstrates the high degree of time utilisation of the SUNeyed vehicles. Most vehicles worked round the clock with activity interrupted only by short rest and waiting periods. Vehicles were found to be moving 40 per cent of the time and loading or unloading a further 11 per cent of the time. Rest periods accounted for 46 per cent of the time. In total trucks were in active use 51 per cent of the time (i.e. over 12 hours per day). The high level of working activity was maintained throughout the whole week. Apart from a slight reduction in loading and unloading, the level of activity was not substantially reduced on a Friday (the normal rest day) compared with other days of the week. Figure 3 shows how vehicle use changed throughout the day. The most active movement times were between 16.00 hrs and 02.00 hrs. Trucks were most likely to be at rest between 02.00 hrs and 08.00 hrs. Even at the quietest time of day, 06.00 hrs, 37 per cent of vehicles were found to be working. The most active loading and unloading times were in the middle of the working day between 08.00 hrs and 18.00 hrs. Figure 4 gives the distribution of the stati times of loading and unloading. Although most loading and unloading took TABLE 1 Mean Times Spent on Each Activity for Loaded and Empty Trip Periods ‘Moving Loading Unloading Resfing Repair Total Hrs Hrs Hrs Hrs Hrs Hrs Loaded Trips 10.8 2.2 1.9 6.5 0.3 21.8 Empty Trips 4.4 11.3 0.9 16.5 Empty Periods Between Loaded Trips 7.1 0.6 7.7 Source: Vehicle Activity Survey 4 oA-.3m Per cent of all vehicles o ~ m & 03 0 0 0 0 I n o -. 0 ~w oN Per cent G g G E 0-4.0 .I I 4-8 8–12 - 12-16 16–20 20-24 24–28 $ -. :, 28-32 3m~ 32-36 38A0 4044 44+6 48-52 52-56 56-60 >60 — Start times Fig. 4 Loading and unloading stati times place during the “normal working day” 13 per cent of loading and 29 per cent of unloading started between 20.00 hrs and 6.00 hrs. Unloading (which lasts less time) was more evenly distributed throughout the day than loading. Because it is usually necessary to hire Iabour for loading and unloading the industry appeared to be quite flexible in allowing for it to take place throughout the night. An analysis of the data showed that a rest or waiting period occurred in 57 per cent of the time periods immediately prior to a loading. On average this period lasted 8.9 hrs. In other words, as expected, drivers tended to take a break after completing a delivery and before collecting their next load. But many also took the opportunity for a break before unloading. In 40 per cent of the periods immediately prior to an unloading a rest or waiting period occurred which lasted on average 6.1 hrs. 5 THE RELATIONSHIP BETWEEN TRIP FREQUENCY AND TRIP DURATION In order to estimate the extent to which journey time savings could be translated into extra trip making a series of elasticities was derived from the Vehicle Activity Survey data relating trip frequency to trip duration. The survey data were collected over periods lasting up to four weeks. During the survey some of the truck drivers 6 took their normal “day of~ recreational rest while other drivers did not. In order to estimate the extent to which this lack of uniformity may have biased the result an additional analysis was undertaken using data adjusted for long rest periods. A relationship had been found between rest days per month and loaded journey time from data collected from the Roadside Interview Survey and this was used to smooth the total amount of time devoted to long rest periods in the “adjusted data se~ referred to below. A further analysis was cartied out which excluded data from the tanker trucks which have their own unique pattern of operation. The following three data sets were prepared: i) The basic data set (45 cases) ii) The adjusted data set (45 cases) iii) The adjusted data set excluding tankers (38 cases) Each data set was derived from the 45 survey periods of the Vehicle Activity Survey. For each case the number of trips made per day, the moving and working time per trip and the mean working time per day was calculated. One approach to estimating the utilisation of time savings is to determine the relationship between total working time per day and average movement time per trip. If time savings cannot be fully used then working time per day (i.e. all moving, loading and unloading time) would rise with mean movement time per trip. This analysis is presented in Table 2. . TABLE 2 Regressions Relating Total Working Time Per Day To Movement Time Per Trip Basic data set. N =45 i) Total working time per day = 11.0 + 0.11 (movement time per trip) (se = 0.04) R2=0.13 Adjusted data set. N=45 ii) Total working time per day = 10.8 + 0.074 (movement time per trip) (se = 0.035) R’= 0.09 A@usted data set excluding tankers. N=38 iii) Total working time per day = 11.5- 5.7x 10-5(movement time per trip) (se = 0.07) R2=0 Although both of the first two regressions shown in Table 2 have positive slope coefficients and are significant at the 5 per cent level of probability they have very low explanatory power. The last regression has no significance at all. The analysis shows no consistent relationship between total working time per day and movement time per trip - hence there is little to suggest that journey time savings cannot be fully used. Elasticities derived between the number of trips made per day and movement time per trip (excluding loading and unloading time) cannot be used directly to estimate the extent to which time savings can be used following journey time savings. However they are useful for indicating how the total distance travelled per day might change with a change in travel time. Total distance travelled may be calculated from the average journey distance and the predicted number of trips made per day. An elasticity of -1 between trips made and movement time per trip is unlikely where loading and unloading times are significant. An elasticity of this magnitude would imply that total working time per day (including moving, loading, and unloading time) will increase with travel time savings. The elasticity is the coef~cient “b” in the following regression:- Log (trips) = a + b x Log (journey time) Table 3 gives the results of regressions relating trips per day to mean trip movement times, separately for “loaded trips” and for “all trips”. “All trips per day” refers to the mean number of loaded and empty trips made per day. Precise definitions of loaded and empty trips are given in Section 4.2 above. The regressions have high R’ values and are very significant. The elasticities lie within the range -0.70 to -0.77, which are much higher than those found in the Malaysian study. To estimate the extent to which time savings can be used following road investment elasticities using total working time were calculated. Regressions showing these elasticities are given in Table 4. For each data set three regressions are presented which relate the following variables:- i) “all trips per day” to “mean total working time per trip” ii) “loaded trips per day” to “mean loaded working time per loaded trip” ‘ iii) “loaded trips per day” to “mean total working time per loaded trip” The last provides an estimate (as far as is practical in Pakistan conditions) of the response of “round trip making to total working time per “round trip”. In the analysis total working time refers to all empty and loaded moving time as well as loading and unloading time while loaded working time is similar but excludes empty movement time. ,. Elasticities in Table 4 range from -0.84 to -1. All of the regressions are highly significant and have high R2 values. Figure 5 gives a typical plot of the data used in Regression (xv). Overall, by introducing loading and unloading into the analysis, the elasticity values have risen by 25 per cent. The mean value of the elasticities is -0.92; i.e. we could expect commercial vehicles to make use of over 90 per cent of the time savings following road investment. 7 TABLE 3 Regressions Relating Trips Per Day to Moving Time Per Trip Basic data set. N = 45 iv) Log (all trips per day) = 0.682-0.698 Log (movement time per trip) (se 0.036) R,= Ogo v) Log (loaded trips per day) = 0.518-0.709 Log (loaded movement time per trip) (se = 0.046) R,= 085 Adjusted’data set. N =45 vi) Log (all trips per day) = 0.715-0.755 Log (movement time per trip) (se = 0.032) R2 = 0.93 vii) Log (loaded trips per day) = 0.56-0.768 Log (loaded movement time per trip) (se = 0.042) R2 = 0,88 Adjusted data set excluding tankers. N=38 viii) Log (all trips per day) = 0.724-0.765 Log (movement time per trip) (se = 0.043) R’= 0.90 ix) Log (loaded trips per day) = 0.506-0.702 Log (loaded movement time per trip) (se = 0.048) R2 = 086 a TABLE 4 Regressions Relating Trips Per Day To Working Time Per Trip Basic Data Set. N= 45 x) Log (all trips per day) = 0.909-0.838 Log (mean total worting time per trip) (se = 0.047) R’= 0.88 xi) Log (loaded trips per day) = 0.873-0.913 Log (mean loaded working time per loaded trip) (se = 0.061) R2 = 084 xii) Log (loaded trips per day) = 0.90-0.864 Log (mean total workng time per loaded trip) (se = 0.045) R2 = Ogo Adjusted Data Set. N =45 xiii) Log (all trips per day) = 0.968-0.913 Log (mean total working time per trip) (se = 0.041) R2 = 0.92 xiv) Log(loaded trips per day) =0.951 -0.995 Log(mean loaded working time per loaded trip) (se = 0.054) R2 = 0.89 xv) Log (loaded trips per day) = 0.969-0.932 Log (mean total worting time per loaded trip) (se=O.038) R2 = 093 Adjusted data set excluding tankers. N=38 xvi) Log (all trips per day) = 1.00-0.95 Log (mean total working time per trip) (se = 0.054) R2 = 089 xvii) Log(loaded ttips per day) = 0.88-0.925 Log(mean loaded working time per loaded trip) (se = 0.066) R2 = 085 xviii) Log(loaded trips per day) = 1.03-0.98 Log( mean total working time per loaded trip) (se = 0,052) R2 = O.gl 9 0.: ( –0.4 –0.6 –0.8 \ . . ● ✎ ✎ ✎ ✎ ✎ ● ✼ ✎ ✎ ✎☛ ✎ ✎ ✎ ✎ ✎ 1 1 I 1 I 1 I I \ I 1 0.6 0.9 1.2 1.5 1.8 2.1 Log working time per loaded trip Fig. 5 Loaded trips per day against working time per trip 6. THE RELATIONSHIP BETWEEN REVENUE AND JOURNEY TIME AND DISTANCE Road freight transport in Pakistan is very competitive and in this environment there is every reason to believe that revenues and tariffs reflect operating costs. To examine whether it was possible to identify the proportions of operating costs that are dependent on time and distance an analysis was undertaken to determine the extent to which tariffs could be explained by these variables. This information could be useful in checking the relative importance of time related costs within the more complex cost models used in road investment appraisal. The analysis was carried out using simple and multiple regression techniques with data collected from the Vehicle Activity Survey. Four sets of data from the Vehicle Activity Survey were analysed. Because the tariff rates for carrying petroleum products are set by oil companies directly on a distance basis they were excluded from the analysis. The sets of data were as follows :- i) Bedfords - grouped data, 26 observations, data relating to periods lasting from five days to four weeks. ii) iii) iv) Bedfords - empty and loaded trip data, 176 observations, data relating to trips including empty movements. Bedfords - loaded trip data, 201 observations, data relating to loaded trips only. Mercedes - loaded trip data, 77 observations, data relating to loaded trips only (non tankers). Simple regressions relating tariffs (Rs) to distance (kms) and tariffs to time (hrs) are shown in Tables 5 and 6. They show that both time and distance are good explanatory variables of tariffs. The relationships are very significant and the R2 values are high. Distance is a better explanatory variable for the grouped Bedford data set and for the Mercedes data set whereas time is a better explanatory variable for the two Bedford trip data sets. Moving time appears to be a better explanatory variable than working time. Multiple regressions relating tariffs to working time and distance are shown in Table 7. These provide a slightly better explanation than the simple regressions. The R* value is raised in three of the four data sets by between two and three per cent compared with best alternative simple regression. However in the last two regressions the second term has very little significance. As expected, there is clearly a high degree of correlation present between time and distance. TABLE 5 Regressions Relating Tariffs to Distance Bedfords - grouped data set. N =26 xix) Tariff Revenue (for period)= 1499 + 2.50 (Empty & Loaded Distance) (se = 0.17) R’= 0.90 Bedfords - empty and /oaded trip data set. N = 178 xx) Trip Tariff = 286 + 2.28 (Empty & Loaded Trip Distance) (se = 0.12) R,= 0.67 Bedfords - loaded trip data set. N = 201 xxij Trip Tariff = 363 + 2.76 (Loaded Trip Distance) (se = 0.15) R’= 0.62 Mercedes - loaded trip data set. N = 77 xxii) Trip Tariff = 165+ 6.14 (Loaded Trip Distance) (se = 0.39) R2 = 0.76 11 TABLE 6 Regressions Relating Tariffs To Time Bedfords - grouped data set. N = 26 xxiii) Tariff Revenue(for period) = -1634 + 86.2 (Empty & Loaded Work Time) (se = 7.13) R’= 0.86 Bedfords - empty and /oaded trip data set. N = 178 xxiv) Trip Tariff = 246 + 73.7 (Empty & Loaded Moving Time) (se = 3.77) R2 = 0.69 xxv) Trip Tariff = 93.6 + 66.9 (Empty & Loaded Work Time) (se = 3.47) R2 = 0.68 Bedfords - loaded trip data set. N = 201 xxvi) Trip Tariff = 250 + 97.7 (Loaded Moving Time) (se = 3.9) R2 = 0.76 xxvii) Trip Tariff = 34.6 + 87.6 (Loaded Work Time) (se = 3.6) R2 = 075 Mercedes - loaded trip data set. N = 77 xxviii) Trip Tariff = 452 + 147 (Loaded Moving Time) (se = 12.2) R2 = 066 xxix) Trip Tariff = 381 + 115 (Loaded Work Time) (se = 12.9) R2 = 0.52 One way of testing the usefulness of the multiple regression model when multicollinearity is present between the independent variables is to compare the R2 value of the regression with the squared correlation coefficient between the independent variables. This is done in Table 8 where it can be seen that for the two Bedford trip data sets the multiple regression R2 value is lower than the squared correlation coefficient between the independent variables. This suggests that multiple regression results for these two data sets are not satisfactory. It appears that the multiple regression for the grouped Bedford data set provides the best explanation for Bedford truck operating costs. The analysis is designed to reflect the relative importance of time and distance for the freight transport market, however different vehicle types do operate in different markets with different tariff levels. The age and size of the Mercedes trucks may account for the relatively low proportion of tariffs explained by working time. These vehicles (which are imported from Afghanistan) are very old and consequently have relatively low capital values, in addition, because they carry much more than the Bedfords, Iabour costs account for a lower proportion of their total revenues and costs. Table 8 provides an estimate of the percentage of the mean tariff of each data set that is explained by time and distance in the different multiple regressions. The Bedford grouped data set and the Mercedes data set have a total percentage “explained” by time and distance that is greater than 100 per cent. This is because the constant term in the regression equation is negative. Table 8 suggests that multiple regression analysis cannot easily be used to explain tariffs. There is considerable variation in the relative importance of time and distance for the different data sets. Assuming that the grouped Bedford data set provides the best explanation and adjusting for the constant term then working time and distance account for 45 and 55 per cent of tariffs respectively. The grouped data regression suggests that 12 TABLE 7 Regressions Relating Tariffs to Time And Distance Bedfords - grouped data set. N =26 xxx) Tariff Revenue (for period) = -350 + 1.61 (Empty & Loaded Distance) (se = 0.324) + 35.3 (Empty & Loaded Work Time) (se =11 .4) R’= 0.93 Bedfords - empty and loaded trip data set. N = 178 xxxi) Trip Tariff= 141 + 37.8 (Empty & Loaded Work Time) (se = 8.06) + 1.1 (Empty & Loaded Distance) (se = 0.28) R2 = 0.70 Bedfords - loaded trip data set N = 201 xxxii) Trip Tariff= 47.5 + 80.1 (Loaded Work Time) (se = 7.87) + 0.292 (Loaded Distance) (se = 0.27) R2 = 075 Mercedes - loaded trip data set N =77 xxxiii) Trip Tariff= -3.08 + 5.43 (Loaded Distance) (se = 0.586) + 21.7 (Loaded Work Time) (se = 13.4) R,= 078 TABLE 8 Proportion of Tariffs explained by Time and Distance Per Cent Of Tariffs Multiple Squared corr. Explained By:- Regression coefficient Constant Work R2 bemeen Data Set Term Time Distance value Time & Distance Bedfords - grouped data -2.8 46.3 56.5 0.93 0.86 Bedfords - empty & loaded trip 13.4 51.5 35.1 0.70 0.83 data Bedfords - loaded trip data 4.2 88.6 7.1* 0.75 0.79 Mercedes - loaded trip data -0.1 16.5* 83.6 0.78 0.56 ● Term not significant at 5 Y. probability. 13 revenue will increase by Rs 1.6 for each extra kilometre travelled and increase by Rs 35.3 for each extra hour worked. An analysis of vehicle operating costs for Bedford trucks indicates that the key time dependent costs, namely crew, capital costs, taxation and profit amount to about one third of total revenues. Although there is a margin of uncertainty about how operating costs may be split up into time and distance dependent elements, nevertheless the tariff analysis probably overstates the importance of time dependent costs. 7. CONCLUSIONS A study of freight vehicle utilisation was carried out in Pakistan. From the data elasticities relating trips made per day to mean trip working time were calculated to lie in the range -0.84 to -1 with a mean value of -0.92. The analysis suggests that in Pakistan’s conditions, journey time savings following road investment are likely to be fully translated into extra trips. Freight vehicles were found to be used very intensively in Pakistan. The sumey found that vehicle running accounted for 40 per cent of the time and that with loading and unloading vehicles were in active use for over 12 hours per day. Vehicles were used intensively at night; the most active running period was between 4.00 pm. and 2.00 am. Even at the quietest time of day, 06.00 hrs, 37 per cent of vehicles were working. Although most loading and unloading took place during the “normal working day” 13 per cent of loading and 29 per cent of unloading was started between 8.00 pm. and 6.00 am. A statistical analysis of freight tariffs with trip times and distances found that either time or distance could provide a good explanation of tariffs. However the high degree of correlation between time and distance made it difficult to identify the separate explanatory power of either variable. The best analysis suggested that time and distance could account for 45 per cent and 55 per cent of Bedford truck tariffs respectively. 8. ACKNOWLEDGEMENTS The work described in this report forms pan of a programme of joint research between the Overseas Unit (Head: J S Yerrell) of the Transport and Road Research Laboratory, and the National Transport Research Centre (Head: M S Swati), Pakistan. 9. REFERENCES De Weille J. (1966). Quantification of road user savings. World Bank Staff Occasional Papers No. 2. Baltimore. Fleischer, G.A. (1963). Effect of Highway Improvement on Travel Time of Commercial Vehicles. Highway Research Record, Vol. 12. Hine, J L and A S Chilver (1991 ). Pakistan Road Freight Industry: An Ovewiew. Department of Transport TRRL Report RR 314. Transport and Road Research Laboratory, Crowthorne. Thomas, S (1983). The value of time savings in West Malaysia: commercial vehicles. TRRL Report SR 792. Transport And Road Research Laboratory, Crowthorne. Winfrey, R (1969). Economic analysis for highways. International Textbook Co. Scranton. Dawson, R.F.F. and Vass P. (1974) Vehicle operating costs in 1973. Department of the Environment TRRL Report LR 661. Transport and Road Research Laboratory, Crowthorne. 14 APPENDIX A: TABLE Al Vehicle Activity Suwey Data Code Make Axles Truc~type Stati Survey Trips Trips No. date days loaded empty No. No. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Total Bedford Bedford Mercedes Bedford Bedford Bedford Mercedes Bedford Bedford Mercedes Mercedes Mercedes Bedford Bedford Bedford Bedford Bedford Bedford Mercedes Mercedes Mercedes Bedford Bedford Mercedes Bedford Bedford Bedford Mercedes Bedford Bedford. Bedford Bedford Bedford Bedford Bedford Mercedes Bedford Mercedes Bedford Mercedes Bedford Bedford Bedford Bedford 2 2 3 2 2 2 2 2 2 3 3 3 3 2 2 2 2 2 3 3 3 3 3 2 3 2 2 3 3 2 2 2 2 2 2 2 3 3 3 2 3 2 2 2 2 Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Tanker Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Tanker Simple truck Simple truck Tanker Simple truck Simple truck Simple truck Simple truck Tanker Tanker Simple truck Simple truck Simple truck Simple truck Simple truck Simple truck Tanker Simple truck Simple truck Simple truck Simple truck Tanker 03107185 03/07/85 07110/85 03107185 01/1 1/85 17110/85 21/1 1/85 08/05186 07109/85 17/07/85 23/08/85 20/09/85 08/1 0/85 08/02186 19102186 06/04186 11/02/86 22/03/86 15/03/86 11/02/86 14103186 30/03186 01/04/86 16/04/86 08109/85 03/04/86 o~i09ia5 16/04/86 15/04/86 08/03/86 19/03186 16/05186 08/05186 09/071a6 18106186 06/05/86 25109/85 01lo51a6 02/09/85 121061a6 12102186 02110/85 151091a6 021031a6 26/05/86 13 7 16 10 16 13 11 7 19 12 17 18 19 6 0 5 8 4 6 6 7 7 15 33 8 29 16 16 16. 10 17 5 9 19 7 12 8 15 6 15 16 10 16 14 11 640 6 4 6 4 27 26 6 6 10 7 8 a 8 4 9 10 11 20 13 6 6 5 14 11 3 16 3 18 5 5 9 1 2 16 6 8 2 21 -2 3 a 13 15 12 2 405 4 3 6 3 3 12 5 6 10 6 a a 8 5 6 a 12 11 13 6 6 5. 14 6 2 17 4 19 5 4 10 1, 2 15 5 9 1’ 21 2 4 7 8 11 3 3 327 15 TABLE A2 Time Utilisation Data: Vehicle Activity Survey CodeNo. Total Loaded Empty Loading Unloading Rest Repair Loaded Total Moving Moving Woting Loaded Worhng time moving moving time time time time trips tips time time time time time worting time /day /day /loaded /all all time /loaded ttipa tips trips /loaded trips tips Hrs Hrs Hrs Hrs Hrs Hrs Hrs No, No, Hrs Hrs Hrs Hrs Hrs 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 2a 29 30 31 32 33 34 35 36 37 36 39 40 41 42 43 44 45 mean Total 2a7,a I 5a.7 3a4.o 210,0 356.0 303.0 241.a I 5a,3 421.9 3a6.5 3a4.a 449,5 447.0 132.8 449.3 327,3 422.1 322.3 36a.1 335,0 3a5.o 364.0 349.6 775.5 173.5 690,5 363.0 360.2 377.0 2ia.a 392.0 127,0 la5.i 430.a 131.5 304.2 I 7a.o 332.6 120,0 342.5 354.5 239a 34a,5 313.1 252.5 t4657.0 91,5 75.4 97.6 a8.5 107.0 aa.o a9.3 44a 162.5 127.3 131.a 174.0 168.5 5a.a 137,6 a2,0 ct4a.a ‘ao.9 90.9 109.8 106,5 76,0 71,2 332.3 46.8 1a6.7 109,5 50.5 95.5 a6.1 loa,6 31.2 61.a 101,5 32,5 74.4 4a.2 27.9 2a,7 1I 7.a 133.5 37.7 63.0 142.0 71.a 4376,5 22.8 11,0 29.4 11.5 5.0 2a.o 17.7 20.0 47.4 44.3 3a.a 47.0 35.0 16.2 16.9 9.8 41 .a a.a 29.3 34,3 29.5 40.0 27.0 47,3 Ii.a 55.5 59.0 29.4 20.5 12.2 32.4 36a 46.3 70.9 ia.1 49,4 10.0 24.1 16.7 97,0 32,0 20.5 36.5 3.a al,5 1422.5 6,5 7.9 10.2 3,5 64.0 4a.o 6.7 9.a 20.a 10.5 16.0 33.0 46.5 7.7 14.0 2a,o 25.1 4a.a 26,9 12.5 17.5 12.0 36.1 14.6 3.5 29,1 5.5 22.5 Ia.o 22.0 14.6 4,2 2.4 34.6 10.3 la.7 a.2 19,4 7,0 2.7 23.0 21.5 41,0 14.3 5.3 a56.o 6.3 5.6 7.0 4.5 49,0 53.5 11.6 a.7 17.6 15.5 I a.5 29,0 22,0 9.0 10.5 35.8 17,9 47.7 17.6 14,0 19.0 13,0 17,6 24.9 3.0 24.2 9,5 13.5 19.0 11.9 16.4 3.0 7.a 22.2 a.9 14.6 5.0 16.1 7.3 1a,a 1a.o 22.3 29.0 10,2 7.3 158.3 4a.5 225.0 97.0 99.0 71,5 129.1 73.0 164.6 165.3 169.3 153.5 136.0 40.4 232.0 169.2 186.7 120,6 i9a,7 159.5 194.0 216.0 197.6 347,4 I oa,4 349.7 170.0 244,3 213.0 a6.5 201.4 51.9 65.0 201.6 61.7 145.0 Iol.a 211,a 52.3 100.0 137.0 126,3 161.5 136.7 ai ,a 763,6 6759.9 2.5 10.1 14,7 5.0 32.0 14.0 5,4 2.1 9.1 25a 10,5 13.0 39.0 0.7 3a,5 2.5 1.9 15.5 4,7 5.0 la,5 7,0 0.0 9.1 0.0 45.4 9.5 0.0 11.0 0.0 la,6 0,0 I .a 0.0 0.0 2.1 4.a 33,a a,o 6.3 11.0 11.5 17,5 6.1 4.a 47a,5 0.50 0.61 o,3a 0.46 1.az 2.06 0.60 0.91 0.57 0.43 0.50 0.43 0.43 0.72 o.4a 0.73 0,63 1.49 o.a5 0,43 0.37 0.33 0.96 0.34 0.41 0.56 0.20 1,20 0.32 0.55 0.55 0.19 0.26 o,a9 1.10 0.63 0.27 1.51 0.40 0.21 0.54 1,30 1.03 0.92 0.19 0.67 o,a3 1.06 0.75 o,ao 2,02 3.01 1.09 1.a2 1.14 o.ao 1.00 0.85 o.a6 1.63 o,ao 1,32 1.31 2.31 1,70 o.a6 0.75 0.66 1.92 0.53 0.69 1,15 0.46 2.47 0,64 0,99 1.16 o,3a 0,52 1.73 2.01 1,34 0,40 3.03 o.ao 0.49 1.02 2.10 1.79 1.15 o.4a 1,21 15.3 I a.a 16.3 22.1 4.0 3.4 11.6 7.5 16.3 I a.2 I a.5 21 ,a 21,1 14.7 15,3 a.2 13.5 4.0 7.0 18.3 I 7.a 15,2 5.1 30.2 15.6 11.7 36.5 2.a 19.1 17.2 12.1 31.2 30.9 6,3 5.4 9.3 24.1 1.3 14,3 39.3 16.7 2.9 4.2 11.a 35.9 15,3 11.4 12.3 10.6 14,3 3.7 3.1 7.9 5.4 10.5 13.2 10.7 13.a 12.7 8.3 10.3 5.1 a.3 2.9 4.6 12.0 11.3 11,6 3.5 22.3 11,7 7.3 24.1 2.2 11.6 10.9 7.4 34.0 27.0 5.a 4.6 7,3 19.4 1.2 11.3 30.7 11.0 2.a 3,a 9,7 30.7 11.2 12.7 14.3 12,0 15,4 7.5 5.7 9.7 6.9 12.4 15.2 12,a 17.7 17.0 10,2 11.9 a.6 10,1 6,0 6.3 14,2 14.4 14.1 5.4 24,6 13.0 9.0 26.2 3,1 15.3 14.7 9.1 37.6 29.6 7.4 6,3 9,2 23a 2.1 14.9 33.7 I 3.a 4,9 6.5 11.4 33,2 13.6 17.4 22,3 19.2 24.1 a.1 7.3 14.9 10.5 20.1 21.9 20,a 29.5 29.6 la,9 18.0 14,6 17.4 a,9 10.4 22,7 23,a 20.2 8.9 33,a 17.a 15.0 41.5 4,a 26.5 24.0 15,5 3a.3 36.0 9.9 a.6 13.5 30.7 3,0 21.5 46.4 21.a 6.3 8.9 13.9 42,2 19.a 21.2 25,0 24.1 27.0 a.3 a.4 17.9 13.9 24.8 2a,2 25a 35.4 34.0 22.9 19.9 15.6 21,2 9.3 12.7 28.4 2a,a 2a.2 10.9 3a, I 21.7 1a.5 al,2 6.4 30.6 26.4 19,1 75.1 59.1 14,3 11.a 19.6 35.7 4,2 29a 7a.7 25a 7,a 11.3 14.2 a3.o 26.3 16 Printed in the United Kingdom for HMSO DdK50120 11/91 C5 G2516 10170