High Volume Transport

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

Accident Costs in Indonesia: A Review. ICORS ’97, India, 24 – 27 July 1997


Publications with the same themes

View all

PDF content (text-only)

TITLE: by: Accident Costs in Indonesia: A Review A Downing Overseas Centre Transport Research Laboratory Crowthorne Berkshire RG45 6AU United Kingdom PA3303197 PA33303/97 DOWNING, A(1997).Accident Costs in Indonesia: AReview. ICORS'97, India, 24 -2 7 July 199 7. Republic of Indonesia Ministry of Public Works Agency for Research and Development Institute of Road Engineering Road Research Development Project Published Paper PA 4 ACCIDENT COSTS IN INDONESIA: A REVIEW by Andrew Downing Paper presented at the International Conference on Road Safety, 24-27 July 1997, Andhra University, Visakhapatnarn, India. - ASPf .SER. LASORATORY Trans port Research Lab~oratory, in association United Kingdom. Ivith PT Yodya Karya, Indonesia. ACCIDENT COSTS IN IND'NESIA: A REVIEW List of Contents Page No. 1. INTRODUCTION 1 2. THE NEED FOR ACCIDENT COSTING 1 3. ROAD ACCIDENT COSTING METHODS 2 4. ACCIDENT COSTING IN INDONESIA 3 4.1 Overall Approach 3 4.2 The Gross Output Method 4 4.3 Other Approaches 6 4.4 Casualties per accident 7 4.5 Issues in the costing methodology 8 5. ESTIMATING THE TRUE ACCIDENT/CASUALTY PROBLEM 8 6. ESTIMATES OF ROAD ACCIDENT AND CASUALTY COSTS IN INDONESIA 9 6.1 Average costs per casualty/accident type in 1991 9 6.2 Estimated total road accident costs in 1995 10 6.3 Comparison of Indonesian and UK costs 1 1 7. CONCLUSIONS 1 1 8. ACKNOWLEDGEMENTS 12 9. REFERENCES 12 ACCIDENT COSTS IN INDONESIA: A REVIEW by Mr A J Downing Transport Research Laboratory, United Kingdom ABSTRACT Indonesia has a serious road accident problem with over 10,000 deaths reported annually. These accdents not only cause considerable pain and suffering but they also lead to direct econom-ic costs and a large waste of the countries scarce resources. The Government and the public are showing increasing concern and Indonesia has taken a series of actions to reduce the number and severity of road accidents. However, in order to plan the management of the countries resources effectively in road safety and ensure that road safety attracts a reasonable share of funding, it is essential that Indonesia develops an acceptable approach to costing road accidents. International costing methods have been adapted for Indonesia and this paper reviews these approaches and makes some estimates for National Costs. 1. INTRODUCTION The objective of this paper is to: I) provide a concise review of accident costing methods developed in Indonesia including any key background information 2) show how these can be used to estimate the national cost of road accidents 3) highlight any problems and possible solutions with accident costing in Indonesia. The key documents used in this review are as follows: 1). Transport Research Laboratoiy (TRL), 1993. Accident Cost Study. 2). Jacobs, G D, 1993. The Inclusion of Benefits from Reduced Accident Rates in Highway Cost Benefit Analysis. In: Proceedings of Conference on Asian Road Safety, 1993, Kuiala Lumpur. 3). Swe Road, 1995. Highway Capacity Manual: Traffic Safety and Environmental Impacts. 4). Swe Road, 1995. Highway Capacity Manual: Accident Costs. 5). Transport Research Laboratory, 1995. Costing Road Accidents in Developing Countries. Overseas Road Note 1 0. 6). Dorsch Consult, 1996. Econormic Evaluation and Determuination of Priorities for Road Traffic Safety Enhancement Policies and Projects. This paper highlights the need for accident costing and then summaries the two most appropriate methodologies with reference to developments in Indonesia. It goes on to estimate the national cost of road acci'dents and indicates the problems and possible solutions for future costing exercises. 2. THE NEED FOR ACCIDENT COSTING All countries with well developed road transport systems experience the serious, negative side effect of road accidents. However, most developed countries have managed to bring about large accident and casualty reductions in recent years. For example, in the United Kingdom (UK) deaths were 35 per cent lower in 1995 than the yearly average for 1981 to 1985 in spite of an increase in traffic of 46 per cent. Dcv eloping countries with fewer resources and greater growth rates for traffic and population have not been able to achieve the same improvements. Thus in Indonesia road deaths have remained fairly constant between 1 0,000 and 1 1,000 deaths per year. Two of the main reasons for the slow take up of road safety initiatives are: 1). The government's and the public's lack of awareness of the magnitude of the problem. 2). A lack of knowledge about the likely cost effectiveness of various measures. Clearly costing road accidents can help the government and the public appreciate the real scale of the C: szife n. leors-97.wpdl 1.06.971 1 problem and enable decision makers to prioritise the allocation of resources more effectively. Thus the major benefits of costing road accidents are as follows. I). Accident rates only show part of the problem ie the tip of the iceberg of the real damage, pain and suffering caused. Costing approaches, in effect, allow for all the negative consequences to be valued and comparisons made with other national problems. 2). Putting a value on road accidents and casualties enables benefits of accident or casualty savings to be estimated. This ensures that road safety is ranked equitably in terms of investment in its improvement at the national resource planning level. For example, it can be demonstrated that national road safety plans can achieve accident savings of 5 per cent with Beniefit:Cost ratios of 1 0: 1. The possible economic benefits are therefore considerable and economic arguments can be used to make a powerful case for road safety improvements. 3). Including values for the safety effects of road improvement schemes in their economic appraisal ensures safety is considered in the decision making process and safer designs are chosen. Without accdent cost infrmation, schemes willbe biased towards inreased speed and capacity for motor vehicles and vulnerable road users will be disregarded. 4). Cost benefit analysis of alternative road safety improvement schemes enables the expenditure on road safety to be optimised, ie with limited resources the selection of improvements is optimised. 3. ROAD ACCIDENT COSTING METHODS Six alternative methods are detailed in the Overseas Road Note 10 by TRL (1995) but only two of these are recommended if the national objective is to maximise national output or pursue social welfare objectives eg minimise casualty rates. These two methods (extracts quoted from Road Note 10) are: I). The "gross output" (or human capital) approach. hin this method, the cost of a traffic accident involving a fatality can be divided into two main categories. Firstly there are the costs that are due to a loss or diversion of current resources and secondly there are the costs that are due to a loss of future output. Included in the former will be the cost of vehicle damage, medical treatment and police/adminiistration costs and usually there is little disagreement as to what should be included here. Determining loss of future output of the person(s) killed however is less clear cut. Usually average wage rates are used (gross of tax) to determine lost output both for the year in which death occurred and then for future years. Costs in future years that the casualty rmight have lived have to be discounted back to give present day values. This is not done separately for every individual killed (or injured) in a road accident: estimates are based on average (ie national) output or earnings data together with appropriately estimate damage, medical and police costs. In some variants of this approach, a significant sum is added to reflect the "pain, grief and suffering" of the accident victim and to those who care for him or her. 2). The "value of risk change" or "willingness to pay" approach. This approach is based on the fundamental premuise that decisions made in the public sector concerning the allocation of scarce resources should reflect the preferences and wishes of those individual citizens who will be affected by the decisions (Jones-Lee 1976, 1989). Accordingly, the value of a given improvement in road safety (ie a reduction in risk) is defined in terms of the aggregate amount that people are prepared to pay for it. Conversely the cost of a reduction in safety is defined in termis of the amount people would require in compensation for the increased risk. More specifically, the value of a particular safety improvement is defined as the sum of all the amounts that people (affected by the improvement) would be willing to pay for the (usually very small) reductions in risk provided by the safety improvement. Thus the value of prevention of one accident involving one fatality is defined as the total amount that all affected individuals would pay for the very small risk-reduction, both for themselves and for those they care about. Estimation of willingness-to-pay costs and values is far from straight forward. Various methods have been used and include an approach where estimates are obtained by observing situations C: -ife n. lcors-97.wpdl 1.06.972 2 where people actually do trade off wealth or income for physical risk. Another approach uses a complex questionnaire where samples of individuals are asked more or less directly how much money they would be willing to forfeit in order to obtain a small reduction in their own or other people's risk. For example, a detailed questionnaire might indicate that drivers were prepared to pay, on average UKL 5 for a risk reduction of one chance in 500,000 that they would be killed on aparticularly journey. Then the 'value of an average life' in this instance would be UKCL 5 x 500,000 ie UKL 2.5 million". The first approach, Gross Output, is well suited to the objective of maximising the wealth of the country but is not so appropriate for cost benefit analysis. The second method, Willingness to pay, meets social welfare objectives and is ideal for cost benefit analysis. This method is used in a number of developed countries including the UK, USA, New Zealand and Sweden. However this method is difficult to use and the reliability of the estimates has been questioned. The method involves asking road users to estimate how much they would pay for safety devices which reduce their risk of injury in road accidents. In the UK the study excluded children and non- motorists and assumed they valued safety at the same amounts. These assumptions have been criticised and they are perhaps even less applicable to developing countries where larger proportions of the casualties are children and non-motorists. Because of these difficulties with the "willingness to pay" approach, the Gross Output method was recommended by TR.L for use in developing countries (see Road Note 10). It was also recommended that the method was augmented by the addition of an allowance for pain, grief and suffering in order to 'capture somic of the human considerations of the "willingness to pay" approach. In the UK, this Gross Output method with an addition for pain, grief and suffering was used until 1988 when "willingness to pay" was introduced for fatal accidents and this increased the cost of these accidents by about 260 per cent. Later, in 1993 the method was unified across all injury accidents. It is now generally recognised that the "willingness to pay" based costs will substantially exceed their "gross output" counterparts but that the latter could at least be treated as the lowest reasonable estimate of accident costs. 4. ACCIDENT COSTING IN INDONESIA The first systematic survey of road accident costs in Indonesia was carried out jointly by TR.L, the Institute of Road Engineering (IRE) and the Institute of Technology Bandung (ITB) in 1992 and 1993 as part of the Technical Assistance and Research Training Project, Second Phase (TARP II). These accident costs were reviewed under the Highway Capacity Manual (HCM) project in 1995 and examined again in the Traffic Safety and Vehicle Weights and Dimensions Project (TSVWDP) in 1996. This section of the paper compares the methodologies proposed in the relevant reports issued by these Projects ie 1). Accident Cost Study. TRL/1RE (1993). TARP II. 2). Highway Capacity Manual. SweRoad/Bina Marga (1995). HCM. 3). Economic Evaluation and Determination of Priorities for Road Traffic Safety Enhancement Policies and Projects. Dorsch Consult/DGLC (1 996). TSVWDP. 4.11 Overall Approach All 3 studies recommend the Gross Output method for Indonesia at this time. A comparison of the key features of each study is shown in Table 4.1 below. The TARP II and HCM studies give estimated cost figures based largely on the original TARP II estimates made for Bandung and both give multipliers for converting the Gross Output values to Gross Output plus Pain, Grief and Suffering and to Willingness to Pay (see 4.3). The TSVWDP report focuses on describing a software programme which can provide cost estimates based on Gross Output with a facility for including multipliers to add elements for Pain, Grief and Sufferfflg. C:rl0ti.Iors,-97.wpd1 1.06.973 3 The TARP 11 study provides estimates for average costs per accident by severity for Bandung and Indonesia (1991 prices) using the Gross Output method, the HICM study report gives average costs per severity of casualty (not accidents) for all three costing approaches. The TSVW&DP report gives some examiples of accident costs using its software but warns that these should not be used as estimates of actual costs. Overall Costing Approach in the three Indonesian studies 4.2 The Gross Output Method There are four key elements of cost in the Gross Output method as follows. 1). 2). 3). 4). Value of lost production Medical costs Damage/repair costs Administration, police and legal costs Although the actual cost calculation approaches to each of these elements were based on TRL's Road Note 10 (1995) there were some differences between the three studies. A comparison of the methods used is shown in Table 4.2. The HICM study used the same estimates as made in TARP II except that a correction was made for the value of lost production for fatalities to allow for the non-productive years of children. This had the effect of reducing the value of lost production for a fatality firom Rp. 33 million to Rp. 26 million ie by just over 20 per cent. The TSVW&DP software allowed for more flexibility in the calculation and for more details to be entered. The main differences are shown in Table 4.2. Some of the key changes are: 1). Any average retirement age can be entered. 2). Medical costs are split into daily costs and per patient costs. 3). Repair costs are estimated per vehicle type with a facility for including costs of the vehicle befing off the road for repair. 4). Police and administration costs are based on police time costs rather than a percentage of the other resource costs. C: satfe~fn. [eors-97.wpdl1 .06.974 Table 4.1 TARPIUI(1 993) HCM (1995) JTSVW&DP? (1 996) I. Describes three approaches as 1. Describes three approaches as follows. 1. Offers software using fibllows. 1). Gross Output 1). Gross Output 1). Gross Output 2). Gross Output plus an amount for 2). Gross Output plus an amount for "pain, grief and suffering" using UK 2). Gross Output plus an amount for "lpain, grief and suffering" using UlK multipliers. "pain, grief and suffering". Any multipliers, multipliers can be used. 3). Estimated "willingness to pay" based 3). Estimated "willingness to pay" on scaled up Gross Output figures using 2. Recommends "Gross Output" based on scaled up Gross Output Swedish multiplier, approach based on GDP but advises figures using UK multiplier, preferred criterion for evaluating 2. Recommends "Gross Output" if investments in road improvements is 2. Identifies "willingness to pay" as objective is to maximise GDPR The other Net Domestic Product ie resources ideal approach for cost benefit analysis two approaches are recommended if the consumed by each saved fatality but not feasible yet for Indonesia. objective is to maximise economic should be taken into account. welfare. 3. Accident cost figures provided for 3. Hypothetical case studies only. B~andung and Indonesia using method 3. Casualty cost figures provided for 1). Indonesia using all 3 methods. 4 Table 4.2 A comparison of Costing Elements in the Gross Output Method in the 3 Indonesian studies I ~~~TARPII(1993) ] HCM (1995) TSVW&DP (1996)I a). Value of lost production 1). Gross Domestic Product excluding 1). Same as TARP 1I. Software allows for the following to be the mining sector per capita of the entered. working population (1 5 to 55 year old). 2). Average years lost weighted to exclude non 1). Gross Domestic Produced excluding the 2). For Fatalities average years lost = productive years of victims mining sector per capita (all population*)s by average retirement age (55) -average age under working age ie overall Province. otfffatality (27). value reduced. 2). Same as TARP 1I. Any average retirement 3). For Present Values of fatalities saved 3). Same as TAR-P II age can be entered. ain incomne growth rate of 4.6 per cent was used and a discount rate of 10 per 4). Same as TARP El 3). Any values can be input for growth and ecent. discount. 4). For non-fa~tal accidents, days lost were 4). Oft' work days are entered separately from based on hospital days only ie hospital days ie usually more. Serious injury =31 days Slight injury = 2 days *can change population to working age group. b). Medical costs 1). Average daily hospital rate (from Same as TARP fl. Software allows for the following to be large hospital)*. entered. 2). Average outpatient rate. 1). Daily medical costs -Accommodation & administration 3). Elemnent aidded for ambulance & -Drugs administration. 2). Per patient costs 4). Assumned 50% of slight casualties go -Tests/surgery to hospital for outpatient treatment. -Ambulance-Outpatient visits 5). No. of days in hospital estimated from survey NB. As with TARP fl estimates, it may be Fatal = 4 days difficult to get true costs. Serious 31 days in hospital + 4 visits as out patient. * Likely to underestimate real costs as state subsidises running cost of hospital. c). Damage/repair costs 1). Average insurance policy payment 1). Same as TARP II. The software allows for the following to be obtained from PT Asuransi Jasa entered. Indonesia. 2). Same as TAR.P UI. 1). The average repair costs per vehicle per 2). Used the following ratios from the 3). Same as TARP HI. accident severity. UK to calculate repair costs per severity of accident 2). The average number of vehicles i) Fatal accident = 1.8 involved/damaged per year per accident ii) Severe accident = 1.4 severity. iii) Slight accident = 1.0 iv) D~amage only ace. = 0.54 3). Vehicle off road costs based on days for v) Average injury ace. -1. 11 repairing and daily depreciation cost of vehicle. 3). The insurance payment in 1) was equated with v) the average injury NB. Surveys of garages have shown variations accident. of up to 4 times the cost for the same repair. C: safe n. lcors-97.wpdl1 t.06.975 Table 4.2 5 4.3 Other Approaches Both the TARP II study and the HCM report give multipliers for estimating the "Gross Output plus Pain, Grief and Suffering" and "Willin'gness to Pay" values. Both studies use the same values for the former ie the UK figures used in 1980's. However the TARP II study describes the UK, increases for "Willingness to Pay" ie a multiplier of 2.6 whereas HICM based its "Willingness to Pay" estimates on Swedish experience. In section 6 of this paper which presents actual cost estimates, the TARP II multipliers were modified for non-fatal casualty costs to match the increases found in 1992 when the UK adopted a "Willinlgness to Pay" approach for all seventies of accidents/casualty. The comparison between the studies is shown in Table 4.3. The TSVW&DP allows for any multipliers to be used for estimating the Pain, Grief and Suffering element but there is no facility for estimating "Willingness to Pay" values. Table 4.3 Multipliers for alternative methods in the three studies C: satfe n. leors-97.wpdl 1.06.976 Table 4.2 (Cont'd) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ [ ~~~TAR.P H(1 993) HCM(I995) ITSVW DIP(1 996)I d). Administration, police and legal costs 1 ). Costs estimated as percentages of total of Same percentages used hut final Software allows for the following to all other costs. amounts different because of be entered. different figure for value of lost 2). Percentages were based on UK figures ie production. 1). Number of police man days per Fatal accident x 0.2% accident severity. Serious accident x 4 % Slight accident x 14% 2). Labour cost per day. Damage only accident x 1 0% 3). Overheads as a percentage of 1) NB. These percentage allow for all adminis- and 2). tration, police and legal costs involved. TARP 11 (1 993) HCM (1 995) ] TSVW&DP (1 996)-_ ai). Gr~oss Output plus `pain, grief and suffering" 1). Recommends the following increases to the 1). Same percentage as TARP 11. 1). Software allows for non resource costs: 2). Same as TARP 11. resource costs to be entered as Fatal accident +38 per cent 3). Figures given in report. a percentage of the resource Serious injury acc. +1 00% costs. Any values can be Slight injury accident +8 % chosen. 2). Perce ntage based on 1986 UK figures. 3). No figures were actually given in the study. b). "Willingness to Pay" 1). Identifies UK multiplier as 2.6 times the 1). Calculates an equivalent value 1). The software does not make value for the Gross Output materials costs using multipliers based on Swedish any provision for "willingness plus pain, grief kind suffering. No figure was experience, to pay" calculations. presented in the report. In Tables 6.1 and 6.2, the latest UK cost revisions made in 2). The multipliers for casualties 1992 have been used for non fatal accidents were as follow: Fatality x 2.93 ie Serious injury acc. cost x 3.25 Severe injury x 5.46 Slight injury aecccost x 3.7 Slight injury x 1.97 2). The above multipliers apply to the 3). The multiplier for damage only accident cost plus pain, grief & suffering, accidents was 1. 11. 4). The above multipliers apply to the basic gross output cost only. 6 4.4 Casualties per accident When calculating the resources costs, the value of lost production and medical costs are estimated per casualty whereas repair costs and administration and police costs are estimated per accident. Clearly it is necessary to express these costs uniformally per accident and per casualty. In Road Note 1 0, it is recommended that an average conversion factor is used for all seventies of accident or casualty. In the TARP II study the Bandung data indicated that there were about 1.25 casualties per inu 'acdent and this figure was used to estimate the accident costs of lost production and medical treatment. It was recommended that a higher value of 1.45 should be used for Indonesia as a whole to allow for the more serious accidents on interurban roads and this was the figure used in the HCM study to convert repair costs per accident to costs per casualty (see Table 4.4). The TSVW&DP software calculates the number of fatalities and casualties per severity of accident and allows for conversions to be made based on actual data. Table 4.4 The conversion ratio for accident costs to casualty costs proposed by the three Indonesian studies. The conversion factor of 1.45 regardless of severity needs further investigation. The 1995 Indonesian Police road accident statistics indicate 42,45 3 casualties (including fatalities) for 16,5 10 accidents ie 2.5 7 casualties per accident (including a few damage only). Thus 1.45 could be too low. However the TSVW&DP also reviewed reported accident data for five Provinces 'm 1994 and the Project reported 4,532 casualties for 3 170 accidents ie a rate of 1.43 casualties per accident. The estimated numbers of casualties pe'acdent severit are shown in Table 4.5 below. T a l e4e ru m e oac s a l i s e r i j u y c i d n Accident5 Number of casualties per injury accident svrty Fatal Seriously injured Slightly injured Fatal 1.63 0.24 0.09 Serious 0 1.24 0.09 Slight 0 0 1.09 These estimates would indicate that multiplying casualty cost elements by 1.45 for all three seventies of accdent will underestimate the cost of a fatal accident whereas the costs of serious and shligt accidents will be overestimated. One problem with obtaining a reliable estimate of the number of casualties per accident is the high levels of under reporting of accidents especially the less serious ones and the variability of reporting levels in different parts of the country. Thus the lower number of casualties per accident in the 5 provinces compared with the Nation as a whole could be due to a higher than average proportion of urban roads (less serousaccdents) or to a higher level of reporting of less serious accidents which have fewer casualties in the smaller data set. The problem of under reporting is dealt with further in section 5 below. C: satfe h. lcors-97.wpdl 1.06.977 [ TARP II(1 993) I F[CM (1 995) [ TSVW&P (1 996) -~ I ). A figure of 1.25 casualties per injury 1). The second TARP 11 figure The software allows for the entry of. accident was used for converting casualty was used ie 1.45. costs to accident costs for Bandung. 1). The fatalities and casualties per year for each severity of injury accident. These can 2). A figure of 1.45 casualties per injury be based on National or Provincial figures accident was used for Indonesia in or other data. Conversions are based on general. This was based on the Cyprus actual numbers of casualties per severity of figure in Road Note 10. accident. 7 4.5 Issues in the costing methodology In order to cost accidents it is evident that a mnuber of assumptions have to be made. Some of these assumptions are questionable and they need to be investigate further by carrying out more detailed analysis or surveys. Some of those at issue are as follows. 1). The use of GDP per capita working population. How close is this to real average incomes'? What assumptions have been made about women working ? How can the 'black economy' be taken into account. 2). The retirement age of 55. This seems low and may result in an underestimation of the value of lost output. 3). The medical costs based on figures given by a state hospital. How much subsidy is provided to state hospitals and were their overhead costs excluded ? 4). Police costs were based on UK multipliers. In Indonesia there are a large number of police involved in accident investigation and a high proportion of cases (over 75 per cent) go to court. Therefore the costs may be proportionately higher. 5. ESTIMATING THE TRUE ACCIDENT/CASUALTY PROBLEM Table 5.1 shows the ratio of different casualty severity totals to the total number of deaths for Indonesia and other countries. According to the police reported accidents in Indonesia there are about twice as many non fatal casualties as deaths whereas in the UK even in 1937 the ratio was 34: 1. Although the pattern of accdents may be different in Indonesia especially with more motorbike accidents this huge difference is likely to be due largely to a massive under reporting of the non-fatal accidents by the public. Also, in Indonesia, the police statistics for fatalities are rarely updated if victims die after befing transported from the accident scene, whereas developed countries usually correct all reports up to 30 days after the accdent. It isgenerally accepted that the number of victims dyin after being moved from the scene i between 5 0 to 1 00 per cent of those who died at the spot. In Indonesia this is supported by the insurance statistics and in 1995 there were 15,080 claims for people killed in road accidents compared with only 10,990 deaths reported by the police (see Table 5.2). Therefore the Police statistics clearly under estimate the number of people killed in road accidents. In order to estimate the true magnitude of the accident problem in Indonesia it is proposed that the numnber of fatalities is corrected for the victims who die within 30 days of the accident and serious and slight casualties are estimated by multiplying this corrected figure for road accident deaths. Three estimates have becn given in Table 5.1 varying from 25:1 for injuries to deaths using a TSVW&DP figure based on very early UK statistics to 52:1 from the HICM study. In Table 5.2 these two ratios and the 1937 UK ratios have been used to estimate the 'true' numbers of casualties in Indonesia for 1995. It can be seen that even with the lowest multiplier (TSVW&DP) the estimated total casualties is 428,609 compared with only 32,804 reported to the police ie 92 per cent of the casualties are missing. Clearly these estimates must be treated cautiously and research is needed to compare hospital records with police records to develop realistic estimates based on factual information. C: satfe h. leors-97.wpdl 1.06.978 8 Casualty ratio SourcesT11 Fatal Serious Slight All injuries Damage ______________________________________________ (1) j (2) J (3) J (2)+(3) only [ndonesia 1995 Police 1 0.9 1.1 2 - Indonesia 1995 Insurance 1 - - 2- UK 1937 "I 1 8 26 34- UK 1981-1985 ()1 13 43 56- LUK 1995 (2) 1 13 72 85 - Sweden 1985-1990 Interurban 1 14 33 47 167 Sweden Urban 1 26 130 156 1400 Indonesia -All roads. HCM 1995 (estimate) 1 13 39 52 59 Indonesia -Bandung. F[CM 1995 (estimate) 1 20 116 136 559 Indonesia -1995. TS&VWDP (estimate) I 25 from Department of Transport (UK), 1977 ()from Department of Transport (UK), 1996 Table 5.2 Road Accident Casualties in Indonesia Number of casualties Sources J~ Fatal Serious Slight All injuries Damnage (1) [2 ()(l)+(2)+(3) J only 1). Police, 1995 10,990 9,952 118232,804 - 2). Insurance, 1995 15,080 29,680 44,760 - 3). TSVW&DP, 1995 16,485 10,01 309,093 428,609 - 4). tIK, 1937 16,485 1180 4860576,975 - 5). HCM -All roads, 1995 16,485 2435 6295873,705 889,720 Also there is a need to encourage and facilitate better accident reporting by the public. Making third party insurance compulsory in Indonesia could lead to better levels of reporting if drivers mivolved in accidents actually make insurance claims. However large increases in reporting will also require far more police resources for accident investigation. 6. ESTIMATES OF ROAD ACCIDENT AND CASUALTY COSTS IN INDONESIA 6.1 Average costs per casualty/accident type in 1991 The estimated average costs for different seventies of casualty and accident are shown in Tables 6.1 and 6.2. Estimates are given for 1991 prices using the three different valuation approaches and variations according to the methods used by the TARP II study and the HCM study. The main difference between the two studies is that the valuation of a fatality is about twenty per cent lower for the HCM Gross Output estimate and over thirty per cent lower for the Willingness to Pay estimate compared with the TARP II figures. Each table gives a choice of six figures but it is recommended that the minimum estimate for Indonesia should be the HCM Gross Output plus Pain, Grief and Suffering figures highlighted 'm bold print. C: s~fe h. leors-97.wpdl 1.06.979 Tahle 5.1 Casualtv ratios from different sources 9 Method Accident SeveityTf TARP 11: 1991 prices (Rp. 1,000.-) ]HCM: 1991 prices (Rp. 1,000.-) 1. Gross Output 1). Fatal 38,849 31,583 2). Serious 1,924 1,924 3). Slight 662 662 4). Damage only 308 300 2. Gross Output plus 1). Fatal 53,612 [1.1) x 1.38] 43,585 [l.l) x 1.38] Pamn, Grief and 2). Serious 3,848 [1.2) x21 3,848 [1.2) x2] Suffering 3). Slight 715 [1.3) x 1.08] 715 [1.3) x 1.08] 4). Damage only 308 300 3. Willingness to Pay 1). Fatal 139,391 [2. 1) x2.6] 92,538 [1.1) x2.93] 2). Serious 12,506 [2.2) x 3.2 51* 10,505 [1.2) x 5.46] 3). Slight 2,646 [2.3) x3.7]* 1,304 1.3) x 1.97] _________________4). Damage only 308 333 [1.3) x 1. 11] *based on UlK 1992 revision from Hopkin and O'Reilly, 1993 These figures can be used as an estimate for current and future accident costs in Indonesia by micreasing the costs according to national indicators of price increases. However this paper has highlighted some issues and limitations and it is recommended that the survey of costs is repeated in order to provide a wider basis for national and regionial costs. Also the key agencies involved should reach a consensus on which method to use for the short and medium term future. Ideally a simple system should be established for revising the figures every year. Also estimates should be provided for rural and urban accidents and casualties separately (HCM provides some estimates) and for an average injury accident. Tahle 6.2 Casualty cost. for Indonesia by severity tvne 6.2 Estimated total road accident costs in 1995 Table 6.3 shows estimates (not made in the original studies) for total road accident costs in Indonesia. Clearly these figures need to be treated with caution because of the difficulty in estimating the true number of casualties and accidents in Indonesia as well as the costing difficulties. The costs have been calculated for three different estimates of the number of casualties in 1995 ie the TSVW&DP proposed ratio of casualties to deaths of 25:1 and the higher ratios from UK, 1937 statistics and the HCM ratios based on Swedish experience. Figures are given for three costing methodologies using the HCM 1995 casualty costs rather than the higher TARP II figures. The 1991 costs have been increased by 40 per cent to bring them in line with 1995 prices. Again, assuming a National objective of maximising welfare, values should be adopted with an element for pain, grief and suffering. Therefore, depending on the true magnitude of the road accident problem, the minimum estimated cost for road accidents in 1995 ranges from 691 to 958 million US dollars. Adopting a "Willingness to Pay" approach is likely to double these costs. C: ssfe n. leors-97.wpdi 1.06.97 1 I Method - [Accident Severity JTARP 11:1991 prices (Rp. 1,000-J CM 199.1 prices (Rp. 1,000. -) 1. Gross Output 1). Fatal 26,792 21,800 2). Serious 1,327 1,300 3). Slight 456 460 2. Gross Output plus 1). Fatal 36,973 [1.1) x 1.38] 30,100 [1.1) x 1.38] Pain, Grief and 2). Serious 2,654 [1.2) x21 2,600 [1.2) x2] Suffering 3). Slight 492 [1.3) x 1.08] 500 [1.3) x 1.08] 3. Willingness to Pay 1). Fatal 96,130 [2. 1) x 2.6] 63,800 [1.1) x 2.93] 2). Serious 8,626 [2.2) x3.2 5] 7,100 [1.2) x5.461 ____________________3). Slight 1,820 [2.3) x3.71* 910 [1.3) x 1.1 1] *=based on UK 1992 revision from Hopkin and O'Reilly, 1993 Avvident t.n-tq liv severity tyne. for Indonesia TAM4, 6-1 10 Table 6.3 Estimated total road accident costs in Indonesia: 1995 6.3 Comparison of Indonesian and UK costs The estimated average costs per accident for Indonesia and the UK are shown in US dollars for 199 1. The UK figures vary from 12 to 68 times higher than the Indonesian figures. Table 6.4 A comparison of UK and Indonesia Accident Costs Road Accident Costs, 1991 AccidentSeverity Indonesia (1) UK (2) (2) : (1) HCM: --Gross Output plus Mixed "Willingness to Pay" and "Gross Ratio of UK costs to Pain, Grief and Suffering` Output + Pain, Grief & Suffering" Indonesia costs (US Dollars) (US Dollars) Fatal 18,160 1,243,364 68:1 Serious 1,603 43,195 27:1 Slight 298 4,091 14:1 Damage only 125 1,565 12.5:1 For 1992 (Hopkin and O'Reilly, 1993) the UK human costs were revised for serious and slight accidents and this had the effect of'increasing the values by more than three times thosein Table6A.4 However, even using the old UK costing methods for the non-fatal injury accidents, it is evident that the UK costs are much higher per accident than for the same severity in Indonesia. This is not unexpected given lower wages and GDP per capita in Indonesia but it is important that costs of countermeasures are similarly reduced in Indonesia if the same economic returns are to b civd 7. CONCLUSIONS This paper has: 1). Reviewed accident cost studies in Indonesia and proposed a minimum average figure for casualties and accidents. 2). Demonstrated the problems of the under reporting of accidents, proposed methods for estimating the true magnitude of the problem and estimated the national cost of road accidents. 3). Highlighted a number of issues in costing road accidents in Indonesia. 4). Recommended that the cost estimates which were based on Bandung data should be reviewed on a National basis and a system developed for updating these figures using simple procedures on an C: safe n. leors-97.wpdl 1.06.97 1 Road Accident Costs: 1995 (Rp millions and US$ millions) Method of estimating true number of Accident Cost Method (HCM, 1995 unit costs) ca~~ualtie~~ Gross Output IGross Output + Pain, Grief & Suffering Willingness toa TSVWDP: Rp. 1,263,377 1,659,644 3,305,140 US$ 526 691 1,377 UK. 1937: Rp. 1,392,851 1,848,431 3,744,164 USS 580 770 1,560 HCM-AII roads: Rp. 1,680,877 2,298,471 4,836,493 US$ 700 958 2,015 Table 6.3 1 1 annual basis. Also consensus should be reached on which costing approach to use. 5). Indicated that Indonesian accidents are valued significantly less than 'n the UK and, although this is to be expected, similarly reduced costs for road safety improvements are needed if they are to achieve equivalent economic returns. 8. ACKNOWLEDGEMENTS The author is grateful to the Institute of Road Engineering for the help given by its staff and the resources provided. Also the assistance with documents and comments provided by the Highway Capacity Manual and the Traffic Safety Vehicle Weights and Dimensions project teams are much appreciated. The draft report was quality reviewed by Dr B Hills of the Overseas Centre, TRL. 9. REFERENCES Department of Transport (UK), 1978. Road Accidents Great Britain 1977. HMSO: London. Department of Transport (UK), 1996. Road Accidents Great Britain 1995. HMSO: London. Dorsch Consult/Directorate General Land Transport, 1996. Economic Evaluation and Determination of Priorities for Road Traffic Safety Enhancement Policies and Projects. Working Paper No. 25, Directorate General Land Transport, Jakarta. Jacobs, G D, 1993. The Inclusion of Benefits firom Reduced Accident Rates in Highway Cost Benefit Analysis. Proceedings of Conference on Asian Road Safety, 1993: Kuala Lumpur, Malaysia. Jones-Lee, M W, 1976. The Value of Life: An Economic Analysis. London, Martin Robertson, UK. Jones-Lee, M W, 1989. The Economics of Safety and Physical Risk. Oxford, Basil Blackwell, UK. Hopkin, M and D O'Reilly, 1993. Revaluation of cost of road accident casualties: 1992 revision. TRL R-R 378, Transport Research Laboratory: Crowthorne, UK. Swe Road/Bina Marga, 1995. Highway Capacity Manual: Traffic Safety and Environmental lImpacts. Work Package Report A: 2. 1. Directorate General Bina Marga, Jakarta, Indonesia. Swe Road/Bina Marga, 1995. Highway Capacity Manual: Accident Costs. Work Package Report A: 2.2, Directorate General Bina Marga, Jakarta, Indonesia. Transport Research Laboratory/Institute of Road Engineering, 1993. Accident Cost Study. Technical Assistance and Research Training Project, Second Phase (TARP II) Project Report, Pusat Litbang Jalan, Bandung. Transport Research Laboratory, 1995. Costing Road Accidents in Developing Countries. Overseas Road Note 1 0, Transport Research Laboratory: Crowthome, UK. C: safe n. leors-97.wpdl 1.06.97 112