Professional Documents
Culture Documents
September 2008
Acronyms
TABLE OF CONTENTS
List of Tables
List of Figures
1. It is now well established that many developing countries have a serious road
accident problem. Fatality rates (per licensed vehicle) are high in comparison with those in
developed countries and whilst in Europe and North America the situation is generally
improving, many developing countries face a worsening situation. For example, over the
period 1969 to 1986, the number of people killed in road accidents in twelve European
countries combined actually fell by about 20 per cent. In sixteen Third World countries
combined there was, over the same time period, an increase in the number killed of over 150
per cent.
2. The global cost of road accidents in developing and emerging nations is about $70
billion each year1. An estimate of the total national cost of road accidents will help
Governments realize the heavy economic losses (typically between 1% and 2% of gross
domestic product [GDP] being incurred annually). This will encourage them to invest in road
safety improvements to reduce these losses and to see expenditure on road safety as an
investment and not as a cost2.
3. Whilst these trends give cause for concern in developing countries, road safety is,
but, one of the many problems demanding its share of funding and other resources. Even
within the boundaries of the transport and highway sector, hard decisions have to be taken
on the resources that a Third World government can devote to road safety. In order to assist
in this decision-making process, it is essential that a method be devised to determine the
cost of road accidents and the value of preventing them.
First need for cost figures is at the level of national resource planning to ensure that
road safety is ranked equitably in terms of investment in its improvement. Fairly
broad estimates are usually sufficient for this purpose, but those must be compatible.
A second need for road accident cost figures is to ensure that the best use is made
of any investment and that the best (and most appropriate) safety improvements are
introduced in terms of the benefits that they will generate in relation to the cost of
their implementation.
4. A very large number of high income countries (HICs) have been estimating the costs
of road traffic crashes1 over the past three decades. The methods used and costs allocated
have generated a great deal of discussion and debate, in particular because of the difficulty
of putting monetary values on pain and suffering.
5. Calculation of direct and indirect costs of injuries, deaths and damage due to road
traffic crashes3 started in the 1970s and many such analyses have been done in USA and
Europe The main objective of assessing costs has been to provide an objective tool for help
in selecting more cost-effective countermeasures for road safety and also to justify
expenditures for the same.
6. There are many methods of estimating costs from the death of a citizen of working
age. In their papers on the cost of traffic accidents and evaluation of accident prevention in
developing countries, Hills and Jones-Lee (1981, 1983) identified six different methods that
have been proposed for placing a cost on road accidents. Under any of these methods,
Nepal loses vast amounts because too many people of working age die untimely deaths.
7. One standard method of figuring out such costs is to estimate lost output - if a 25-
year-old dies, we may assume that he or she would have worked another 33 years on
average and amortise the estimated loss of productivity. Another method is to extrapolate
costs from insurance (life, medical and vehicle) claims. A third method is to estimate
"willingness to pay": how much would the dead person's dependants have paid to keep him
or her hale and hearty? That's a proxy for output loss.
8. How much does it cost? Apart from "output loss", "willingness to pay" (WTP) and
insurance claims, there are other direct costs that are difficult to estimate. A road is blocked
for an hour after an accident: how great is the loss of productivity? What about paramedic
and police deployment costs? This method has been adopted by several developed
countries, viz. UK, USA, New Zealand and Sweden. As per the TRL (1995) Report, obtaining
reliable empirical data was difficult. The method used for cost of fatal accident, e.g. in UK,
cannot be used for non-fatal accidents. Even while estimating the fatal accidents, wide range
of empirical estimates was obtained from different studies including considerable element of
judgment. The values often related to adults, not children, whereas, in developing countries
children are highly affected in RTAs. WTP method used in UK covered drivers and
passengers of motor vehicles only. In developing countries, there are many road users are
affected including high proportion of pedestrians and using non-motorized vehicles. These
limitations of the WTP method for the developing countries, including Nepal, restrict the
application of the WTP method for the developing countries, including Nepal to estimate the
cost of accidents.
9. To simplify the process of identifying the sources costs, the Human Capital
Approach, otherwise known as the Gross Output Method, classifies accident cost into three
main components as shown in Figure 1. Victim-related costs are directly associated with the
resources lost of the casualties. To be humane and considerate of the social impacts of
accidents, a notional sum is also added to quantify the pain, grief and suffering (PGS) of the
victim’s families. Property damage consists of vehicle repair, lost of economic productivity of
public utility vehicles and cost of towing services. The first two components make the largest
portion of property damage. The third and last component of accident cost consists of costs
associated with police investigation, legal activities and insurance administration.
10. Human Capital Method (Gross Output) of costing road accident is generally used for
developing countries. Costing will, invariably, include costs incurred by the victim, the
society, and the government in general. Measuring emotional damages translated to social
costs puts values to non-monetary terms such as PGS of the victim and the affected family.
These are things, not bought nor sold, otherwise known as non-market values.
11. Miller [10] summarises the burden of injury losses into the following categories:
(iii) Work Loss Costs value productivity losses. They include victims' lost
wages and the replacement cost of lost household work, as well as fringe
benefits and the administrative costs of processing compensation for lost
earnings through litigation, insurance, or public welfare programs like food
stamps and Supplemental Security Income. As well as victim work losses
from death or permanent disability and from short-term disability, this category
includes work losses by family and friends who care for sick children, travel
delay for uninjured travellers that results from transportation crashes and the
injuries they cause, and employer productivity losses caused by temporary or
permanent worker absence (e.g. the cost of hiring and training replacement
workers).
(iv) Quality of life includes the value of pain, suffering, and quality of life loss
to victims and their families.
Damage to Property
12. Costs of vehicle damage can be obtained from insurance companies and workshops
based on property damage claims and costs of repair of damaged vehicles, respectively.
Since the costs of vehicles damaged in accidents can vary depending on the degree of
damage sustained, it is only possible to obtain estimates based on accident severity. It is not
always the case that the severity of injuries sustained by accident victims correlates with the
degree of damage done to vehicles. For instance, in the case of a pedestrian accident that is
categorized as a fatal accident, the vehicle may have little or no damage.
13. When getting information from insurance companies, it is possible to get a sample of
cases of accident claims based on accident severity. The process of estimating costs will be
possible, but it will not be accurate. This is because the comprehensive process of collecting
information is not an easy task. Several things need to be considered when dealing with
information from insurance companies. These include uninsured commercial and public
service vehicles and vehicle damage costs not fully paid by the insurance company.
14. Another way of getting estimates of vehicle damage costs is to gather information
from automobile repair workshops, but this can be time-consuming and often does not get
the expected cooperation from workshops, due to varieties of damage to the vehicles
involved in accidents. The average repair cost for vehicle damage and the average number
of vehicles involved in each severity of accident are obtained.
15. Finally, the cost of damage to government property, such as streetlights, railroads,
and other street furniture can usually be obtained from the Public Works and Services
Department and the police.
Administrative Costs
16. Administrative costs involve the costs of court proceedings, insurance claim
processing, and time spent by the police dealing with or investigating accidents. It is
particularly very difficult to attach monetary value to these services. Some countries choose
to ignore police and administration costs because apportioning costs can be difficult, unless
detailed records are maintained in which such records are not yet likely to be established by
most countries. Further, such effort may not be necessary since the contribution of
administration costs to the overall accident costs is considerably low.
17. One suggestion is to adopt the method used by the United Kingdom and developing
countries when estimating these costs, that is police and administration costs valued at 0.2%
of total resource cost, due to loss of resources in fatal accidents; 4.0% in serious accidents;
14.0% for accidents involving slight injury; and 10.0% for no– injury accidents (property
damage only accidents).
The loss of the value of the work that injured people would
Lost Output have produced during the time they were absent from their
jobs.
Pain, Grief, An allowance for the loss of quality of life and the pain, grief,
and Suffering and suffering incurred by injured people and their families
and friends.
The cost of treating people injured in accidents, including
Medical Costs the cost of doctor’s fees, medicines, and (if required) time
spent in hospitals and receiving long term care.
The cost of repairing vehicles and other property damaged,
Property
including the costs paid by people and
Damage
Their insurance companies
The costs incurred by the police and other
Administrative
services in the process of attending to accidents and
Costs
associated activities and by insurance companies
administering insurance claims.
Medical Costs
18. Costs on medical treatment include hospital treatment and the use of ambulance
services. Hospital treatment refers to both inpatients and outpatients. For accident victims
who are hospitalized (inpatient), the total costs would involve the average length of
hospitalization; average cost of treatment per day; ambulance services (if sent by
ambulance); and, after released, additional costs of medical treatment that need to be taken
into account. For outpatients (usually those with slight injury) only the average number and
average costs per outpatient visit are considered. In the case of fatalities, we can ignore
costs arising from outpatient treatment.
Some information needed to estimate costs of medical treatment can be obtained from
hospitals. Data on costs of hospital treatment per day are usually not published. However,
estimating costs of treatment per day is still possible by making use of information on
available treatment costs. Since medical services and facilities in governmental hospitals in
Nepal are likely to be subsidized by the Government, records of costs of medical treatment
from road accidents are not likely to be available. Nevertheless, the cost figures can still be
obtained from fee-paying patients.
Lost Output
19. When an accident involves a fatality, it will cause a loss of future output for the
country because a resource, in this case labour (one of the economy's factors of production),
is lost. Loss of output is estimated by first calculating the number of days or years lost. This
is obtained by subtracting the average age of accident fatalities from the retirement age. If
the accidents involve serious injury, the number of days of hospitalization and any recovery
time required should be taken into account.
20. For accidents resulting in slight injuries, the number of days absent from work each
time an outpatient visit is made is considered to be the number of days lost. The number of
days or years lost is then multiplied by the national wage rate, to estimate the cost of lost
output in the case of accidents involving serious and slight injury. For fatality cases, a
discount rate would have to be used to discount the loss in future years to a present value.
21. Information on number of days lost can be obtained from hospital records,
employers, or the victims themselves. The average age of accident fatalities is available
from police, transport department, and hospital records, whereas the source for information
on wage rate would be from the Statistical Yearbook.
22. When accidents involve serious injuries and fatalities, victims (seriously injured),
relatives, and friends might suffer mentally and physically. Thus, it is deemed humane to
attach costs for pain, grief, and suffering. There is no way to attach specific values to pain,
grief, and suffering associated with road accidents.
23. An alternative would be to adopt the weights used by the United Kingdom:
24. The experience of poor communities in coping with medical catastrophes is very
different than that experienced by economically well-off communities. The special problems
faced by poor families can include the following:
25. None of the above issues are factored in the standard economic calculations done
for estimating the cost of road crashes in poor societies. When someone in a poor family is
injured and is bed ridden at home or the hospital, the following circumstances take place.
The whole family gets involved to take care of the patient. This results in the
reallocation of labour of all family members – those on daily wages lose their income;
Children may not go to school; and older family members may spend less time in the
care of children and infants.
The household has to cope with the time and financial demands of the situation and
this can have a permanent affect on the health of children and infants in the family.
This can be the result of loss of income, less attention, worsening hygiene at home,
etc.
Since a very large number of poor households depend on daily wages and temporary
jobs, don’t have health insurance, or the assistance of social welfare schemes, a
serious injury can result in permanent reduction of income.
In cases of prolonged treatment or death of the victim, the family may end up selling
most of their assets and land and getting trapped into long-term indebtedness.
Investment in treatment of a seriously ill family member stops only when all assets
get sold. A study from Thailand shows that 60% of involuntary land sales were to
finance treatment of a family member. Death of a male head of household creates a
household headed by a woman. Such families have to suffer serious social and
economic hardships and can have negative health effects on children.
26. It is clear that the outcome of a serious injury or death of a family member in poor
communities has many long-term effects, socially, economically and psychologically on all
the other family members and the community. Many of these outcomes are permanent and
soul destroying for individuals and possibly for the larger community. There is very little work
done to understand these issues. Therefore, we must not stop at the calculation of losses in
purely monetary terms. For poor communities, our methods do not even capture the
economic losses in all their complexity. The effect of injury and death on the family structure,
crushing of hopes and aspirations of future generations, and the psychology of the
community are just not factored in. We’ll have to take these issues much more seriously in
the future and not neglect them just because they cannot be monetised.
27. A research study was undertaken to work out the unit accident cost of the various
categories of road accidents in Nepal, Experts from the Transport Research Laboratory, U.K.
worked on the issue under the project and came out with the report “Road accident Costing
in Nepal” [1] in 1996 under the project “Road Safety Component of the Road Maintenance
Project”, Departments of Roads, Government of Nepal.
28. The above mentioned study used the approach of the Gross Output Method, and
estimated in case of fatalities the lost productivity resulting from the road accident rather
than attempt to put a price on the value of a life. It used data from both inside and outside of
the Kathmandu valley and from insurance companies and hospitals as well as the police for
the study year i.e. fiscal year 2051/52 (1995/96).
29. The study work included work undertaken in estimating the three main components
of the study viz., lost output, vehicle damage, and medical costs. In its calculations for road
accident costing, it also considered administrative costs, a component for pain, grief and
suffering, and damage only accidents. The report came out with following figures for average
accident costs:
30. No matter what methods are used, the economic costs of traffic crashes turn out to
be so high that it becomes easier for professionals to justify higher expenditures in
promoting road safety.
31. A recent report commissioned by the Global Road Safety Partnership [2],
summarised the efforts in the area and concludes that overall it does appear that in most
countries, costs exceed 1 per cent of GDP which may now be considered to be an
underestimate of national accident costs.
32. Table 3 shows recent estimates of economic costs of road crashes summarised by
Jacobs, Aeron-Thomas and Astrop [11]. Expressed as a percentage of GDP, the costs
range from 0.3 per cent in Vietnam to almost 5 % in USA. According to a more recent
estimate from OECD the total annual economic loss resulting from road deaths and injuries
is estimated to at around $US 450 billion or about of 2% of GDP in OECD countries. The
policy makers regard this as a cost that society should regard as unacceptable.
33. The values shown in the Table 3 indicate that the estimates for LICs as a per cent of
GDP are in general lower than those in HICs. We need to be careful in drawing conclusions
from such numbers as that would mean that road safety measures have a higher justification
in HICs than in LICs. The more recent estimates for HICs are based on more detailed and
comprehensive calculations including the willingness to pay, QALYs and DALYs, etc. On the
other hand, to the best of our knowledge, such concepts have not been used in making
estimates in LICs.
34. It has been mentioned earlier that the official estimates for traffic crash injuries in
India could be underestimated by an order of magnitude. If the willingness to pay concept,
effect on quality of life, etc. is properly accounted for, it is likely that the road crash costs in
India would also be 2 per cent of the GDP or greater.
35. As per Prof. Dinesh Mohan [17], while talking in the context of India says that,
generally, the road traffic injuries are underestimated by an order of magnitude and if the
same, along with the willingness to pay concept, effect on quality of life, etc. is properly
accounted for, it is likely that the road crash costs in India would also be 2 per cent of the
GDP or greater. At the intuitive level, this makes sense for developing countries for the
following reasons:
India (and other developing countries) has a higher rate of road crashes than
HICs.
Since cost of life is reflected in the per capita income of the country and its GDP,
the proportion of costs due to loss of life should be similar in HICs and LICs.
As a proportion of per capita income, costs of similar levels of medical care are
higher in LICs as compared to those in HICs.
Because of a scarcity of good rehabilitation care facilities and lack of aids for the
disabled, road crash victims suffering permanent disability would suffer greater
lack of access and employment opportunities in LICs.
Owing to lack of welfare functions provided by the state and health care facilities,
families of injury victims have to spend much more time looking after injury
victims in LICs. This causes greater time and economic losses overall.
36. If the above issues are taken into account, it is possible that the figure of one per
cent of GDP for LICs calculated by Jacobs, Aeron-Thomas and Astrop of the TRL [11] are
also underestimates for LICs for costs of road traffic crashes. They have under estimated the
ratio of fatalities to total injuries at 1:20-30 as against of 1:15:70 for fatalities, serious injuries
and minor injuries, respectively.
37. There have been a few attempts in India to estimate the costs of road traffic crashes
over the past few decades. But, these have followed very simple economic models to
include actual expenses and direct and indirect loss of income, etc. One of the early studies
which attempted to evaluate road accident costs was conducted for Delhi for the year 1968,
and another study [13] calculated accident costs based on insurance company data for
Chennai (Madras) for the year 1978. Both studies used rudimentary methods for cost
analysis. The first major road user cost study (RUCS) [14] was published in India in 1982.
This study, sponsored by the World Bank, included a section on accidents as a component
of road user cost. The costs included were: medical expenses, legal fees, property damage,
insurance costs, and loss of output due to death (future consumption as one-third of income
and future output calculated up to the age of 55 years). The latest study on evaluation of
road accident costs was sponsored by the Ministry of Surface Transport and conducted by
M/s Tata Consultancy Services [16]. A summary of the values calculated in the studies
mentioned above is given in Table 4.
*Source : Jacobs, G., Aeron-Thomas, A. and Astrop, A. (2000). Estimating Global Road Fatalities. TRL Report 445, Transport Research
Laboratory, Crowthorne, U.K.
* In this study, two categories, serious and major injuries were used
1. Srinivasan, N.S., Hingorani, D.V. and Sharma, B. M. (1975). Economic Costs of Road Accidents, Journal of
the Indian Roads Congress, 36-2.
2. Natarajan, T. (1980). Costs of Road Accidents. M.E. Thesis, Department of Traffic and Urban Engineering, College
of Engineering , Guindy, Madras.
3. Road User Cost Study in India (1982). Final Report. Central Road Research Institute, New Delhi.
4. Evaluation of Road Accident Costs - Research Digest (2000). Indian Highways. 28:2, 27-44.
5. Mahajan, B.M.(1991)," Cost of road accidents-A case Study in India" International Conference on traffic and Transport Safety
38. The Road Accident Costing in Nepal made an important recommendation that DOR
should update accident costs annually for the next few succeeding years till that time when a
reliable national data base was established. It further recommended that once that was done
the accidents costs should then be reinvestigated with emphasis given to determining
accurate damage only accident frequency. In order to facilitate annual update recommended
in the 1996 study, an accident costing update guide along with a spreadsheet was
suggested by the study team. The same is given as Annex 1 to this Report. The accident
costing instructions outline the procedure that may be followed in order to update the
accident data, lost output, medical costs, vehicle damage, administration costs and costs for
pain, grief and suffering.
39. The 1996 Cost Study [1] suggested that apart from conducting annual casualty
surveys at hospitals, which the report strongly recommended the remaining effort can be
limited to current medical costs at Patan Hospital and average semi-skilled wage rate and
updating the spread sheet suggested for updating the accident costing. These
recommendations were by and large adopted for this Update together with some
refinements. The following, thus, describes the approach and the results of this updated
costing.
40. This update of the accident costs for Nepal adopted the HC or the gross-output
method in line with the 1996 Costs and for the obvious reasons compared to other methods
available (refer to earlier section).
41. Analysis for the update is based on data from the Traffic Directorate of Nepal Police,
selected hospitals and insurance companies within Kathmandu. However, Consultant also
refined some of earlier parameters for the analysis, based on recent data in context to 2005
study by DOR [3].
42. Average cost of vehicle-damage of a RTA from various insurance companies was
based on breakdown of such costs derived from Rastriya Beema Sansthan, details of which
are given later in this Report.
43. Some of the parameters to the accident-costs were simply extrapolated based on the
national inflation between 1997 and 2007.
44. The scope of this Updated Costing is limited, in comparison to the previous 1996
Accident Study by DOR. A few hospitals, insurance companies and auto-repair workshops
based in Kathmandu were surveyed in this context.
45. The existing accident records maintained by Traffic Police are not adequate to
conduct comprehensive accident costing. There have not been any comprehensive
programmes dedicated to road-safety in Nepal, after the completion of UK funded Road
Maintenance Project during the late nineties except for the DOR 2005 Safety Audits cited
earlier. Many multilaterally-funded projects have conducted road safety audits of their road-
projects, but these studies are confined to their project areas and the scope for
comprehensive research is limited with those data.. Nevertheless, this Study extensively
used the data and findings of the 2005 DOR Road Safety Audits [3], wherever appropriate,
and supplemented by related statistics maintained at the Traffic Directorate and secondary
unofficial accident data.
46. For this Updated Costing, the following assumptions were necessary.
Where latest data on some parameters was not available, the same was extrapolated
from the 1996 Costs using the consumer price index between 1997 and 2008.
The consolidated RTA statistics maintained at Traffic Directorate of Nepal Police,
classified by Development Regions4 and nationally were assumed to fairly represent
the some of the parameters required in absence of more refined data.
47. Hospitals and insurance companies were unable to furnish data in the form required
for accident-analysis. These details are elaborated later in the subsequent sections.
48. Table 5 shows the total number of accidents and casualties that were recorded in
various fiscal years in Nepal.
49. Table 5 shows that the portion of the RTA injuries, as opposed to RTA fatalities, has
gradually increased over the years. This is attributable more to better reporting of injuries,
particularly the minor ones by the Traffic Police. Nevertheless, the average fatal to injury
ratio of 1:7 still signifies under-reporting of the total injuries in Nepal5.
50. Past studies conducted in India and Sri Lanka indicated the fatalities to injuries ratio
to be 1:13 and 1:10, respectively. This would indicate that the average fatalities to injury
factor in the Indian Sub-Continent is roughly 1:12. In contrast, the fatalities reported by Nepal
Traffic Police are presumed to be fairly accurate, based on the relative seriousness that
Traffic Police accords to the fatal accidents as compared to the injury accidents. Though,
there are national statistics on the total number of patients treated at hospitals, there are no
4
Nepal is divided into five development region for administrative purposes.
5
Based on global experience, a fatal to injury ratio less than 1:10 is said to clearly indicate under-reporting.
such data on the total number of RTA-casualties treated at hospitals throughout the country.
A more realistic source of data for number of RTA injuries would have been from hospitals
but attempt to gather such information on national scale was not successful for this Update.
51. Therefore, the total injuries in fiscal year 2006/07 were assumed by multiplying the
total fatalities (953) obtained from the Police data by factor of 12. Based on this
assumption, it is estimated that there were 953 fatalities and 11,436 injuries from RTA
in FY 2006/07.
52. Table 6 shows the distribution of RTA casualties along three national highways of
Nepal during 2004 and including unofficial accident data. The unofficial accident data include
records of accident posted or reported in newspapers and includes all types of roads in
Nepal.
53. Table 6 shows that the average number of all casualties (fatalities, serious injuries,
minor injuries) per fatal accident increases when the unofficial data is included compared
with the corresponding average of the three highways listed above. However, there is no
marked changes in the average casualties per injury accidents between the two data sets.
As the inclusion of unofficial data comprise RTAs in other roads with lower design standards
than highway, the average fatalities of 2.6 per fatal accident and 1.4 serious injuries per
injury accidents was taken for this Update.
54. With regards to the injuries in fatal accidents and distribution of serious and minor
injuries in injury accidents, earlier Section had highlighted severe under reporting of injuries
particularly those pertaining to minor accidents. Previous discussions had also referred to a
study by Prof. D. Mohan in India, where a more realistic distribution of casualties per fatal
accidents in India was cited to be 1 fatality, 15 serious injuries and 70 minor injuries per
fatal-accident. This assumption would also be true in the case of Nepal but to be
conservative it was assumed that the ratio of the casualties per fatal accident was 1:2:6 in
terms of fatalities, serious and minor injuries, respectively. The distribution of the serious
and minor injuries was also assumed to be 1:3 as per the assumption in the 1996 Costs.
55. Based on the above, the distribution of casualties per every fatal accident was
assumed to be 2.6 fatalities, 5.2 serious injuries and 15.6 minor injuries. Similarly, every
injury accident was assumed to result in 1.4 serious injuries and 4.2 minor injuries.
56. Alternatively, when distribution of 1:15:70 is taken for fatal accident; corresponding
ratio between serious and minor assumed valid for injury accident but fatality rate assumed
constant, every fatal accident incurs 2.6 fatalities, 39 serious injuries and 182 minor injuries.
Similarly, every injury accidents will incur 1.4 serious and 6.5 minor injuries.
57. In order to assess the vehicle-damage cost in RTAs, average value of such cost was
sought from five insurance agencies and six auto-workshops, based in Kathmandu.
58. From the insurance companies, only Rastriya Beema Sansthan (RBS) was able to
furnish average cost of damage per type of vehicle involved including breakdown of the
insurance claims in other heads (e.g. third party property or vehicular damage, medical
compensation, etc), from amongst the insurance companies contacted for this Study. The
remaining insurance companies contacted could either furnish only some limited details of
vehicular damage or only the total amount and claims they processed during the latest fiscal
year. However, these companies confirmed informally that the RBS breakdown of their
claims by type of vehicle involved roughly represented the share of the breakdown of their
total claims as well. Therefore, distribution of the total amount and number of claims for
vehicular damage processed at all insurance companies contacted were assumed to be as
shown in Table 7.
59. In addition, the total claim by type of vehicle involved was again segregated into
different cost-components, based on the average of these components from the data
furnished by the insurance agencies and were assumed to be distributed as shown in Table
8.
60. All the contacted auto-repair workshops provided costs incurred for both minor and
major repairs for vehicle damaged in accidents. Though, they provided a tentative estimate
of average repair-cost, none of them was consistent but the minor and major repair-estimate
was more or less uniform. In order to calculate the average vehicle damage, the extent of
damages observed on three national highways based on DOR 2005 Study was extrapolated
(see Table 9) and weighted average of damage by type of vehicle was calculated.
61. The above pattern of vehicle damage was assumed to fairly represent the whole
road-network in the Country and was adopted to arrive at the average cost of damage using
estimate provided by the auto-repair workshops. Table 10 shows the average cost of repair
for vehicle-damage resulting from RTA.
62. From Table 10, it is evident that some of the average costs of repair for particular
type of vehicles were unrealistic and were, thus, ignored. On a comparative basis, the
average cost of repairs for heavy vehicles (bus and trucks) is lower than the 1996 value. But
Table 10 shows that the corresponding cost based on insurance data are also lower than the
1996 figure and, therefore, the new average calculated was taken for this Study.
63. After calculating the average cost of vehicle repairs, the net cost of vehicle (which
takes into consideration other costs such as insurance surveyor’s fee, lost business and
deductions of duty, taxes on spare-parts to convert the financial cost of vehicle damage into
economic cost) was also calculated. The steps, as underlined in the 1996 Study, were
adopted for this purpose. Figure 2 shows the steps involved in calculating the net vehicle
6
Three-wheelers, tractors and non-motorised vehicles were classified as other vehicles.
damage cost and national average vehicle damage cost per accident, as stipulated in the
1996 Study.
64. For this report, cost of spare-parts was assumed to be two-thirds of the total repair
cost as per the 1996 Report. However, duty-tax and value-added tax (VAT) of respectively,
25% and 13% of cost of material and duty was assumed. Surveyor’s fee for insurance-
claims was based on inquiry of these rates from insurance companies (which are pro-rated
based on the damage involved).
65. For commercial vehicles, amount of business lost due to vehicle-damage should also
be considered. This Study assumed the amount of revenue lost by each commercial vehicle
damaged over the same period assumed in the 1996 Study i.e. 15 days. The amount of
revenue lost, however, did not consider the operator’s profit as it was difficult and not
possible to assess this within the limited time available for this Study. The vehicle-operation
cost (VOC) per km and daily travel data for different heavy vehicles were based on
corresponding data from the 2007 Road Users’ Satisfaction Survey for Nepal [5]. Tentative
amount of business lost for other vehicle categories (which includes three-wheelers, tractors
and rickshaws) was based on VOC value for these vehicles obtained using HDM4 RED7
software in an another project [4] and other source. Table 11 shows the net cost of vehicle
damage for each type of vehicles.
7
HDM= Highway Development & Management Model (World Bank) RED= Road Economic Decision
(HDM4 Vehicle Operating Cost Module v.3.2)
66. To calculate the average cost of vehicle repairs per every vehicle damaged in the
urban and rural areas, the distribution of various types of vehicles involved in RTA in these
areas is necessary. For this Update, these figures were extrapolated from the data derived
from the DOR 2005 Road Safety Audits of the three national highways. Table 12 shows the
distribution of vehicle involved in RTA based on above data source.
67. These distributions were assumed to fairly represent the whole country and average
cost of vehicle repairs per every vehicle damaged in urban and rural areas was calculated
taking the weighted average of the cost of repairs for various vehicles based on the
distribution in Table 11.
68. Based on recent data, 40% and 60% of accidents occur in urban and rural areas,
respectively. This distribution was, thus, considered to arrive at the national average cost of
vehicle repairs per every damaged vehicle.
69. Not all RTAs involve vehicle damage (for example in many accidents involving
pedestrians) and, therefore, one needs to adjust the national average cost of vehicle
damaged per damaged vehicle obtained above to reflect the same. While existing police
records do not always enter these details, this Study extrapolated the average of such no-
damage RTAs based on DOR 2005 Road-Safety Audits of the three national highways. (See
Box 1 )
Box 1
No-Vehicle Damage RTAs
(tentative)
% of RTA
EWH 33%
PRM 16%
ARM 21%
Wt-avg= 30%
70. Based on above extrapolation, the average national cost of vehicle repair per vehicle
damaged was multiplied by 0.7 to arrive at the average national cost of vehicle damaged per
each vehicle involved.
71. The total national cost of vehicle damage per each casualty accident was finally
derived after multiplying the average national cost of vehicle repairs per each vehicle
involved with the average number of vehicles involved in all casualty accidents in Nepal
(which was extrapolated to be 1.3 – Table 13).
72. Table 14 shows the final results of the calculation of the national average cost of
vehicle damage per casualty-accidents.
73. Table 14 shows that the national average cost of vehicle-damage cost per casualty-
accident is NRs. 64,967
74. The lost output based on gross-output method is the assessment of earning lost by
victim of a RTA casualty. The average annual income of a RTA casualty and the number of
years or days of income lost due to road-accident are the two main components required for
calculating the lost output.
75. Inquiries, regarding average age of RTA casualty at the three hospitals and past
study8 Nepal, revealed that the average age was within 20 – 30 years. Analysis of RTA data
from the Traffic Police, as well as unofficial data, revealed that the average age of RTA
fatality of 29 years has not changed over the years since the 1996 Study. This figure was,
therefore, taken as the average age of a RTA fatality. The lost earning for a fatality is also
based on the national average age, when adults are still economically active and a proxy for
this figure is the government retirement age in the absence of more refined data. Therefore,
the government retirement age of 58 is taken as the time when adults in Nepal are still
active. For this Study, the number of years of lost earning of a fatality was the difference
between the average age of a RTA fatality (29 years) and the retirement age (58 years).
76. In the case of injury-accidents, the lost income is based on the length of stay in
hospital for treatment, during which time the injured will have to forsake their earnings. The
1996 Study considered 47 days (17 in-patient days and 30 home-recovery days), as number
8
Epidemiological Study of RTA Cases: A Study from Eastern Nepal; N. Jha & C. S. Agrawal
of work lost due to a serious injury and latest inquiry at Kathmandu Teaching Hospital
confirmed that the average length of in-patient days for a RTA was still 17 days. Though, a
recent case study in Bangladesh showed the total lost time for a serious injury was 35 days
[2], the previous data of 47 days was still adopted for this Study considering accessibility
problems in the remote and rugged terrain of the hill areas of Nepal. For minor-injuries, the
number of working days lost derived from the Bangladesh Study (five days) was adopted in
lieu of the corresponding figure of two days adopted in the 1996 Study.
77. While national statistics on the average income of a RTA casualty is not available, a
1997-98 Study by Jha-Agrawal et-al [18] in East Nepal, profiled the RTA casualties in this
Region, based on the educational background and occupation of the victims. Table 15
shows the breakdown of the RTA casualties based on the above Study.
78. While trend in RTA casualties may have changed since the time of the Study in East
Nepal and distribution itself may not truly represent the whole Country, the distribution of this
Epidemiological Study was assumed as fairly accurate picture of the socio-economic status
and breakdown of the RTA casualties in Nepal in the absence of more accurate statistics.
The socio-economic breakdown from the above Study was, therefore, adopted to calculate
the weighted average of the income of a RTA casualty in Nepal, together with the average
wage rates for different types of manpower in Nepal. The average wage of various level of
manpower (unskilled, skilled, semi-skilled, etc) was calculated based on wage rate fixed by
various districts for fiscal year 2007/08. This approach marked a departure from the 1996
Costs, where the average income for semi-skilled worker was used as a basis. Based on the
adopted approach, weighted average wage of a RTA casualty worked out to be NRs. 260
per day or NRs. 93,600 per year.
79. For a serious injury resulting in disability, the average lost output was assumed to be
half of the corresponding value for a fatality in line with the global practice in accident-costing
for developing countries.
80. Table 16 shows the result of the calculation for the average lost output per casualty
in a road-accident.
81. The above rates were, thus, used to calculate the total lost output per accident, using
the distribution of RTA-casualties as shown in Table 6. The earlier assumption of 10 percent
of the serious-injury resulting in disability, as was adopted in the 1996 Costs, was retained
for this Update as well, which generally conforms to ones adopted in another study in the
United Arab Emirates [19]. Table 17 shows the total lost output per road-accident in Nepal.
Total Cost of Lost Output per fatal-acc. = NRs.328,335 Total Cost of Lost Output per injury-acc. = NRs.27,008
Source: Specialist’s Estimate
82. Table 17 shows that the total cost for lost output per fatal and injury accident is NRs.
328,335 and NRs. 27,008 respectively.
83. While data on average medical cost was sought from three main governmental
hospitals, based in Kathmandu Valley, none were able to give a reasonable estimate for a
RTA-casualty as such data was not maintained at these centres. In this respect, only Patan
Hospital had a system of recording the RTA-casualties administered in its emergency rooms
and, therefore, corresponding data at the remaining two hospitals could only be roughly
assumed. These hospitals, however, provided tentative costs of various injuries at their
hospitals. Based on those range of costs provided, the tentative cost of medical treatments
for RTA-injuries including disability was assumed. The average daily medical cost was, thus,
extrapolated as NRs. 265 for minor-injury and NRs. 1,550 per day for serious-injury (with
allowance for long-term disability as well). As per previous discussion, ten percent of the
serious-injuries was assumed to result in disability. Based on these daily costs, the average
medical cost per RTA casualty was calculated and is given below in Table 18.
84. As Table 18 shows, no cost was assigned for RTA-fatalities due to negligible records
of such deaths at the hospitals surveyed9. Based on this table, the average medical cost for
serious-injury including disability, and minor-injury were respectively, NRs. 27,138 and NRs.
530.
85. The total medical cost for each type of RTA was calculated by multiplying the
average medical cost per each type of accident with the corresponding average numbers of
casualties in these accidents as shown in Table 19 below.
86. The 1996 Costs did not assign any medical cost to the number of fatalities involved in
a fatal accident based on the arguments that very few fatalities ever reach hospital and even
amongst those fatalities, few use ambulance service for transport. For this Update, there
was no data available from the hospital surveyed regarding the number of fatalities referred
to in these institutions. The earlier assumptions regarding medical costs involved for fatalities
in a fatal accident could still be valid from a national perspective. Therefore, no medical cost
was assigned to fatalities involved in a fatal accident. This aspect should be revisited in
future update to the accident costs or/and when required data are available. Table 18 shows
that total medical cost for each fatal and injury accident in Nepal is NRs. 149,383 and NRs.
40,219, respectively.
9
In Nepal, any death within 35 days of injury resulting from a RTA is legally defined as RTA-fatality.
However, there were no data to corroborate number of such deaths from either the hospitals or other sources.
87. For this Study, administration cost for road-accidents is based on the extrapolation of
the 1996 Study, based on inflation rates since that date. This cost is not assumed to be
significant in Nepal, as only a few RTA cases are insured or filed in judicial courts in Nepal.
Therefore, the administrative cost for casualty-accident is assumed to be NRs. 884.
88. In accident-costing, a certain value is added to reflect pain, grief and suffering
resulting from a RTA. This practice was adopted so as to reflect society’s value to suffering
resulting from a road-accident beyond and above the economic value attached to the victim
concerned. For this Study, the Indian practice of assigning 20 percent of the total lost output
for each type of accident (fatal/injury) was adopted for this Update in line with the 1996
Costs. Based on this assumption, NRs. 65,667 and NRs. 5,402 are the costs for pain, grief
and suffering for each fatal and injury accident, respectively.
89. Table 20 summarises the average cost of casualty-accident in Nepal after adding all
the cost components mentioned above.
90. Therefore, average cost for a fatal-accident and injury-accident cost in Nepal is NRs.
609,236 and NRs. 138,479, respectively.
91. Table 21 shows the national total cost of all casualty-accidents in Nepal based on the
above figures. It is important to note that the number of fatal or injury accidents denote the
number of these events and should not be confused with the number of personal fatalities or
injuries cited in earlier section.
92. Therefore, in fiscal year 2006/07, the total national cost for casualty accidents was at
least NRs. 1.59 billion.
93. The above analysis only considered the total cost resulting from casualty accidents.
However, there are also significant numbers of road-accidents which do not involve any
casualties but vehicle damage (or damage only RTA). Globally the number and cost of
damage only RTAs is significant and cost can often exceed RTA fatalities cost.
94. The 1996 Costing had assumed damage only RTA to be five and two times greater
than the casualty accidents at urban and rural areas, respectively. ADB Road Safety
Guidelines recommends a factor of 5 to 6 times injury accidents for damage-only accidents
and UK ODA10 Overseas Road Note 10 recommends a factor of 6 and 4.5 times the injury
accidents in urban and rural areas, respectively. Based on these statistics and in absence of
country-specific data to reflect this, the earlier assumption of damage-only accidents to be
five and two times the injury accidents in urban and rural areas, respectively, was retained
for this Update as well following discussion with concerned authorities (Traffic Police and
DOR). The status-quo regarding research on damage only accidents has remained
unchanged. Table 22 shows the assessment of the cost of damage only accidents in Nepal
for FY 2006/07.
95. Table 22 therefore shows that the national cost for damage only RTA was at least
NRs. 1.11 billion in FY 2006/07.
96. The sum of the total national casualty and damage only RTA costs gives the gross
total national RTA cost. Table 23 shows that total national RTA cost of Nepal for FY
2006/07.
10
Overseas Development Agency of UK (now renamed the Department for International Development).
98. The national average cost for a road-accident is required for any economic analysis
on road infrastructures or road safety projects. Based on Table 23, the national average cost
for a road-accident inclusive of damage only accidents in FY 2006/07 is NRs. 305,396.
99. Based on alternative assumption of the distribution of RTA-casualties as per every
accident to be 1:15:70 for fatalities, serious and minor injuries, respectively, and fatality to
injury ratio of 1:20, the national average cost of a RTA inclusive of damage-only accidents is
NRS. 359,597 and the national total cost jumps to NRs. 5.11 billion.
100. Copy of the spreadsheet analysis used for this Update is enclosed in Annex 2.
8. CONCLUSIONS
General
102. The latest accident statistics maintained in the Traffic Directorate of Nepal Police and
the 2005 DOR Safety Audits lacked some details necessary for a comprehensive accident
costing. This Study therefore assumed or extrapolated some of these data from the 1996
Costs.
103. Hospitals and insurance agencies (except for Rastriya Beema Sansthan or RBS)
were unable furnish data in the form or detail required. The RBS data was therefore taken as
main basis to work out vehicle damage costs from RTA.
104. This Study adopted the gross output method for cost analysis as it was the most
appropriate method for Nepal and in line with the 1996 Accident Costing. This Update
therefore adopted the guide and spreadsheet recommended by the 1996 Cost with some
minor refinement.
105. As per the gross output method, analysis for costing comprised assessment of the
four components of a RTA – (i) vehicle damage cost; (ii) lost output costs; (iii) medical cost
of treatment (iv) cost for pain, grief and suffering and (v) administrative costs. All these
components were then added to derive the average cost for a casualty accident in Nepal for
FY 2006/07.
106. In addition to above components, damage-only cost was also calculated based on
certain assumptions to arrive at the total national cost for RTAs in Nepal.
107. The scope of this Update was limited compared to the 1996 Study and data sources
was limited to selected hospitals, insurance agencies and auto-repair workshops based in
Kathmandu Valley.
108. The latest statistics on the number of RTA fatalities provided by the Traffic
Directorate of Nepal Police was adopted but the number of RTA-injuries was extrapolated
assuming a fatality to injury ratio of 1:12 as per statistics in other South Asian countries to
offset under-reporting of injury accidents in Nepal. Based on this assumption, it was
estimated that there were 953 fatalities and 11,436 injuries from RTA in FY 2006/07.
109. Average number of fatalities per fatal-accident and serious injuries per injury-accident
was based on RTA data derived from DOR 2005 Road Safety Audits of four national
highways and unofficial sources. The serious injuries in a fatal accident was calculated
assuming a factor of two times the fatalities while minor injuries was assumed to be 75 per
cent of the total injuries as per the 1996 Costing.
110. Only RBS provided detailed information of insurance claims for vehicle damage and
data from other insurance companies were extrapolated using RBS data as basis. The
estimates from auto-repair workshops were further modified based on severity of vehicular
damages experienced at three national highways of Nepal to arrive at the average cost of
repair.
111. Based on above and taking into consideration the duties, taxes, fees, etc, the net
cost of vehicle damage per type of vehicles was calculated. Based on these figures, the
vehicle distribution in urban/rural areas and the average vehicle involved per road accident;
the total national cost for vehicle damage per casualty accident for FY 2006/07 was
estimated to be NRs. 64,967.
112. The updated average age of a RTA-fatality was found to be still 29 years, as was the
case 1996, following inquiries at hospitals surveyed and analysis of RTA data. The
retirement age of 58 years set by Government was taken as period when an individual is still
considered economically active. For RTA-injury, the total lost time was assumed to be 47
and 5 days for serious and minor injuries, respectively. Average income for a potential RTA
casualty was assumed to be NRs.93,600 annually. This income level was calculated as the
weighted average of the wage earned by the various categories of RTA causalities based on
a 1999-99 study conducted in East Nepal. Lost output for RTA-disability was assumed to be
half of the lost output for RTA-fatality.
113. Based on distribution of the RTA casualty, the total lost output for a fatal and injury
accident in Nepal was estimated to be NRs. 328,335 and NRs. 27,008 respectively in FY
2006/07.
Medical Costs
114. Acquiring reasonable estimate of medical costs for road accidents from the three
hospitals surveyed was difficult, in the absence of a system of maintaining such data at
these centres. The average cost was, therefore, only a rough estimate of cost involved for a
range of treatments at hospitals including long-term treatments.
115. Based on the above, previous 1996 assumptions regarding the number of in-patient
and out-patient visits and RTA-casualty distribution, total medical cost for each fatal and
injury accident in Nepal was estimated to be NRs. 149,383 and NRs. 40,219 respectively.
Administration Costs
116. NRs. 884 was finalized as the administrative costs involved in RTA, based on
extrapolation of corresponding amount in the 1996 Costs.
117. To reflect this cost, 20 per cent of the total lost output for each type of accident was
assumed, in line with the 1996 Costs.
118. Based on this assumption, the costs for this component were estimated to be NRs.
65,667 and NRs. 5,402 for each fatal and injury accident respectively.
119. Though the magnitude of damage-only RTA is significant, there was no statistics
available in Nepal on the above. Therefore, earlier assumptions of assuming the number of
damage-only accidents in urban and rural to be five and two times the injury accidents,
respectively, was also adopted for this Study following reference with other literatures.
Similarly, the earlier assumption, for the cost of damage-only accidents being half of the cost
of the vehicle-damage in respective areas (urban/rural), was also adopted for this Update.
120. Based on the above, it was estimated that the total national costs for damage only
RTA were at least NRs. 1.11 billion in FY 2006/07.
121. Summing up all the cost-components, it was estimated that the total national cost for
all road-accidents in Nepal, including damage-only accidents, in FY 2006/07 was at least
NRs. 2.70 billion, yielding a national average cost of NRs.305,396 for a RTA including
damage-only accidents.
122. Alternatively, based on higher distribution of injuries (serious and minor) in a RTA,
the total national cost comes around at least NRs. 5.11 billion, resulting in an average cost of
RTA with allowance for damage-only accident to be NRs. 359,597. Table 24 summaries the
major findings of the present work.
9. RECOMMENDATIONS
123. Based on this update and till the time better Country-specific data are available, the
best estimate of the total national cost for all RTAs in Nepal should be taken as NRs. 2.70
billion for FY 2006/07. Similarly, NRs. 305,396 should be assumed as the national average
cost per road accident in Nepal with allowance for damage-only accidents for FY 2006/07.
124. DOR should update and refine many of the parameters assumed, when these
become available. It is also important for DOR to update the accident costs annually. In this
context, personnel at DOR RTU should be properly trained to regularly update the accident-
costs and statistics.
125. DOR should conduct research on damage-only accident accidents in collaboration
with the Traffic Police. This would yield better estimate of the average cost of a RTA in
Nepal.
126. There should be coordination between DOR and Traffic Police to ensure that vital
details such as type of collision (head-on, front-end or side-swipe, etc), severity of injuries,
extent of damage, no-vehicle damage and damage-only accidents, etc, are invariably filled in
the Accident Report Forms, maintained by the Traffic Police to facilitate comprehensive
analysis of accidents including costing.
127. As suggested in the earlier Costing, DOR should conduct annual surveys at hospital
to accurately assess the RTA casualties and publish such statistics. This exercise will also
help in assessing the actual magnitude of under-reporting of RTA-injuries in Nepal.
128. Despite the staggering economic losses resulting from RTAs in Nepal, there is no
effort to systematically collect data pertinent to road accidents from different organisations
such as hospitals, insurance companies. Therefore, to facilitate better road safety planning
and management, there should be better coordination among stakeholder organizations with
a lead role by DOR, which should also be entrusted with the statutory responsibility to
maintain database related to RTA under a collaborative arrangement.. Such a system will
provide better understanding of factors leading to road accidents as well as help to estimate
more accurate costs of road accidents
129. In this regard, insurance agencies should be required to record the number and
amount of claims for vehicle-damage from road accidents classified by type of vehicle
involved. Likewise, hospitals should also maintain database related to RTA casualties, as
recommended in 1996 Study. Hospitals should also be instructed to record the number of
RTA-cases which could not be accommodated at their hospitals due to lack of hospital-beds
or treatment facilities. For this purpose, the Ministry of Physical Planning & Works should
coordinate with the Ministry of Health and other stakeholders.
References
Annex 1
Annex-1
For the next few years, the DOR should adopt the following system to annually update accident costs (see spreadsheet attached).
Apart from conducting annual casualty surveys at the hospitals ( which is strongly recommended ), the remaining effort can be
limited to collecting the current Patan medical costs and average semi-skilled wages rate and updating the spreadsheet attached.
1. TESU will be responsible for updating all accident rate assumption such as casualty distributions and average number of
vehicle involved
2. RTA injures should be estimated by hospital counts with survey conducted on all the hospital recommended by Department
of Health Services ( see section 3.1 of 1996 Report for list ) by DOR field staff. Only simple totals of emergency room and
RTA casualties and RTA in-patients by month need to be collected. Data collection is expected to take less than two hours
at each rural hospital each year.
1. Vehicle damage cost can be inflated by the change in average wage rate and / or medical costs.
2. When accident costs are revised (as opposed to update), RTA survey reports should be used as they provide easy
identification of accident location, causalities involved , and the distribution of repair costs in spare parts, labour towing
charges, etc. More importantly they offer an estimate of damage incurred regardless of the extent of insurance coverage.
They are limited however to only those vehicles insured.
LOST OUTPUT
1. Lost output should be represented by semi-skilled labour wage rate multiplied by the amount of work foregone with future
lost earnings discounted at the government discount rate.
MEDICAL COSTS
1. Medical cost should be updated annually according to Patan Hospital cost figures.
2. No updating of average in-patient stays or recovery period is thought to be necessary.
VEHICLE DAMAGE
1. Vehicle damage costs should be updated by the average of increase experienced between lost output and medical costs.
2. No revising of vehicle type or area (urban/ rural) distribution should be made until national statistics are more reliable.
3. As damage only costs thought to be both substantial and very difficult in the present time to accurately determine, their
estimate should continue to be included in the total RTA costs calculations but as a separate component. If further work on
determining property damage only accident frequency was desired , data could be requested from such vehicle fleet
companies as Lama rental cars in accident would must likely be needed to be collected from primary surveys.
ADMINISTRATION COSTS
1. The 20 percent lost output factor is recommended to be continued is use to reflect the cost of pain, grief and suffering.
Annex 2