Fs Immaterial Costs Archived

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

SWOV Fact sheet

The valuation of human losses of road deaths

Summary
A considerable part of the total social road safety costs consists of human losses. This is damage in
the form of suffering, pain, sorrow, and loss of the joy of living of the casualties as well as their family
and friends. It is important for policy and its supporting research to be able to attach a value to this
damage in terms of money. Detailed study has been made in the Netherlands of the valuation of
human costs in fatal crashes. That study estimated the value of a 'statistical' life (VOSL) which mainly
consists of human losses, at 2.2 million 0.3 million. This amounts to nearly 1.8 million 0.3
million in human losses. Taking inflation into account, the value of a statistical life amounted to 2.6
million in 2009.

Background
Crashes lead to all sorts of social costs, such as material damage, loss of production, and medical
costs, but also human costs. Human costs consist of loss in the quality of life for casualties and their
family and friends. These are costs in the form of suffering, pain, sorrow, and loss of the joy of living.
These costs are separate from material costs such as not being able to consume anymore, etc.
Various studies indicate that the human losses resulting from crashes, for both fatalities and injured,
are a substantial part of the total social costs of road safety. Therefore, knowledge about human costs
is important for policy and its supporting research. This includes research into road safety costs and
cost-benefit analyses. Following various other countries, a detailed study was made in the
Netherlands of monetizing the human costs of (only) fatal car crashes. This fact sheet presents a
summary of this knowledge.

Why should we monetize human losses?


There are at least three reasons to express human losses due to road crashes in monetary terms. The
first reason is that the efficiency of traffic and road safety measures is increasingly determined by
using cost-benefit analyses. In these analyses, safety effects, including human losses, are expressed
in monetary values as much as possible (see SWOV Fact sheet Cost-benefit analysis of road safety
measures). This makes it possible to weigh the saving of road crash casualties against the costs of
measures and their effect on, for example, mobility and pollution.
The second reason is that human losses are a large cost item in the total of road crash costs.
Information about road safety costs is used in the preparation and assessment of national road safety
policy. It offers possibilities of comparing the cost of road crashes with other social costs, such as
those of traffic jams or pollution (see SWOV Fact sheet Road crash costs).
The third reason is that a fuller picture of the costs of a specific crash can be obtained if the human
costs are also known. This can be important for insurance claims. The Netherlands does not (yet)
have this possibility, but in other countries estimates of human costs are sometimes used for this
purpose (Smith, 2000).

How does valuation work? the term VOSL


In order to standardize the valuation of safety, including human losses, the concept Value of a
Statistical Life (VOSL) was developed. This 'statistical human life' can be illustrated by the following
example. A risk of a fatal crash of for instance 7 in 100,000 means that, statistically, 7 of every
100,000 people will die in a road crash each year. A decrease of the risk from 7 to 4 road deaths per
100,000 means that on a population of 100,000 inhabitants 3 'statistical lives' are saved.
The 'willingness-to-pay', which is the amount that people are prepared to pay for a certain decrease in
crash rate, is used to determine the value of the decrease in crash rate. This concept originates from
the economic theory of welfare, and makes it possible to put a price on risk reduction and thus on the
saving of 'statistical' human lives. Suppose that people are willing to pay 60 for a crash rate
reduction from 4 per 100,000 to 7 per 100,000. In that case 100,000 people are collectively prepared
to pay 100,000 x 60 = 6 million for an expected decrease from 7 to 4 fatalities. The VOSL is then 6
million 3 saved statistical lives = 2 million per statistical life.

SWOV Fact sheet 1 SWOV, Leidschendam, the Netherlands


January 2012
Reproduction is only permitted with due acknowledgement
A technical complication is that the VOSL comprises the valuation of human losses (suffering, sorrow,
etc.) as well as the material damage (not being able to consume anymore). To only determine the
human losses, the material damage as a result of death must be deducted from the VOSL. As a
measure, the loss of consumption during the years of life lost is used.
Please note that the VOSL does not concern the value of a specific individual's life, but the decrease
in crash rate instead. Most people would 'give anything' not to die. In the concept 'willingness-to-pay',
people make a trade-off between crash rate and money. Every day, people make decisions in which,
unconsciously or not, they make such trade-offs. Take for example, choosing food, choosing driving
speed, whether or not to take a smoke detector, or choosing whether or not to smoke.

Which methods are there for calculating the VOSL?


Various methods have been developed to measure 'willingness-to-pay'. The two most important
groups of methods are (De Blaeij, 2003a):
Revealed Preference (RP). These methods study how much money people actually spend on
safety provisions.
Stated Preference (SP). These methods use questionnaires in which people, directly or indirectly,
are asked about how much they are willing to pay for safety provisions.

In principle, RP methods are preferable to SP methods because RP methods are based on the actual
spending of real incomes. However, applicability of RP methods to road safety is limited. For example,
the purchase of an airbag is often not a separate choice; it is usually part of a set of accessories.
Furthermore, the purchase and use of many safety devices, such as seatbelts, is legally obligatory.
Another limitation of RP methods is that they assume that people are capable of correctly estimating
risks, and the risk reduction that is achieved by purchasing a particular device. However, the risks and
the risk changes in relation with fatal crashes are very small. It is well known that people are hardly
capable of judging them correctly. SP research can provide people with information about small risks
and risk changes in such a way that they are more capable of processing them adequately. This is not
possible with RP methods, which limits their applicability. SP methods also have a broader
applicability because they are not dependant on the availability of data about peoples' actual choice
behaviour. Therefore, SP methods are often used for the valuation of e.g. safety, travel time, or nature
and pollution. On the other hand, this research is so complicated that it needs to be carried out
carefully to obtain valid results.
In addition to RP and SP methods, use is sometimes made of the costs-per-life-saved method. This
method uses policy documents about proposed or realised measures to derive how much money is
invested per life saved. For each decision it is then possible to calculate how much one is prepared to
spend to save a life. Strictly speaking, however, this is not a VOSL because this method does not look
at the preferences of the population, but at the revealed preferences of the policy makers. The
usefulness of this method is limited because it produces a minimal safety valuation. The maximum
amount that policy makers would wish to invest in safety can be higher than the actual amount spent.
In addition, it is often not known whether there were any other (side) effects involved that played a role
in taking the decision about policy measures.

How much is the VOSL according to international research?


In various countries, research has been carried out into the 'value of a statistical human life', and a
number of surveys have been published. A Canadian study analysed the results of 85 VOSL studies
(Dionne & Lanoie, 2004). 28 of these studies were related to road safety and the others to other types
of safety, such as work safety. In the road safety studies, the average VOSL was more than $ 3.5
million (price level 2000). If the studies of other types of safety are included, the average VOSL
amounts to $ 5.6 million. One of the possible explanations for this difference is that the road safety risk
level is lower than that of other types of safety, especially work safety. In the last few years, there have
also been a number of meta-analyses of the VOSL within various contexts. Meta-analyses give an
overview of existing estimates and attempt to explain the differences between them, or to determine a
'best' VOSL estimate. An example of this is a recent OECD meta-analysis (Lindhjem et al., 2011). In
this analysis, 850 VOSL estimates from 80 stated preference studies were analysed. Around one third
of the VOSLs is related to road safety; the other VOSLs are related to environmental risks and other
health risks. The average VOSL for road safety amounts to almost $ 7 million (price level 2005). The
analysis shows that VOSLs that are related to traffic and other health risks are generally higher than
environmental VOSLs. The cause of this is still unknown. Another example is a meta-analysis by De
Blaeij et al. (2003) which analysed road safety VOSL estimates in a number of European countries
and the United States. One of the things this study wanted to determine is which factors influence the

SWOV Fact sheet 2 SWOV, Leidschendam, the Netherlands


January 2012
Reproduction is only permitted with due acknowledgement
VOSL estimate. The values found depend, among other things, on the method used and the initial
crash rate. This study found an average VOSL of $ 4.4 million (price level 1997). The calculations also
included studies that had used the costs-per-life-saved method; without them the average VOSL
would have been higher.
A meta-analysis carried out by Miller (2000) intended to determine a so-called 'value transfer function',
based on 69 VSOL studies in various countries. Using certain variables such as the net domestic
product per capita, this function can be used to estimate the VSOL of any country. Miller used studies
in work safety, fire safety, and road safety that used a variety of methods. For the Netherlands, Miller
calculated a VOSL of $ 2.9 million (price level 1995).

Which VOSL values are used in policy?


In various countries and at European level, official, government-approved VOSL values exist that are
recommended for use in policy support and assessment. In the European projects UNITE and
HEATCO proposals were developed for standard European unit values, including a standard VOSL
(Nellthorp et al., 2001; Bickel et al., 2006). Both projects propose a standard VOSL of 1.5 million
(price level 1998). This value is differentiated per country, taking differences in purchasing power into
account. The estimate that is made for the Netherlands amounts to 1.7 million. However, if a country
has performed a thorough VOSL-study itself, it is recommended that this VOSL specific for this
country is used. The European value of 1.5 million euro per fatality is based on a report of the
European Conference of Ministers of Transport (ECMT, 1998). This stated, on the basis of literature
studies, that the best scientifically based VOSL estimate was 2.4 1 million (price level 1990).
However, for policy purposes, a more conservative approach was followed, and a lower limit of 1.5
million (price level 1998) was recommended, as this would be more readily accepted as a minimum
estimate. In addition, this amount agreed with the average of the official values used in five European
countries which had carried out VOSL research. The ETSC (Jost et al., 2011) has recently made an
inventory of the most recent VOSL values for eight countries. The values vary from 1.4 million to 2.6
million, with an average of 1.7 million (price level 2009). Within the framework of the ROSEBUD
project De Blaeij et al. (2004) found values varying between 1.4 million and 2.6 million (price level
2002). Considering that different reference years were used, this is higher than the ETSC values. An
explanation for this can be that ETSC expresses the values in so-called factor costs which means
that a correction downward is used for indirect taxes (such as VAT).
A VOSL is also being used in the Netherlands. For calculating the social costs of road crashes the
Ministry of Transport (now the Ministry of Infrastructure and the Environment) uses a VOSL of 2.2
million (price level 2001; AVV, 2006), based on the study of De Blaeij (2003b). This VOSL has
previously been used in the Mobility Policy Document (VenW & VROM, 2004). Furthermore, the
Ministry uses the OEI guideline for cost-benefit analysis, but this does not yet contain a
recommendation about the VOSL (Eijgenraam et al., 2000); OEI stands for Overview Effects
Infrastructure). There will shortly be an addition to the OEI guideline that will deal specifically with the
safety effects and their valuation. Expectations are that this addition will make concrete
recommendations about the VOSL that should be used in cost-benefit analyses.

How much is the VOSL according to Dutch studies?


A thorough study of the VOSL which used SP methods was first carried out in the Netherlands in 2001
(De Blaeij, 2003b). The study used a questionnaire with two types of questions. The first type asked
drivers to make comparative assessments between safety, travel time, and money. Every subject was
asked ten times to make a choice between 2 routes from A to B, for which the toll (price), safety, and
journey distance were different each time. Table 1 presents an example of such a question. Analysis
of the data resulted in a VOSL of 2.2 million 0.3 million (a 95% confidence interval).

SWOV Fact sheet 3 SWOV, Leidschendam, the Netherlands


January 2012
Reproduction is only permitted with due acknowledgement
You can choose for route A or route B. Each year 16 people die in a car crash on route A. If you choose A your journey
lasts 1 hour, and the toll is Dfl. 5. Each year 4 people fewer die in a car crash on route B. If you choose route B your
journey is 10 minutes shorter, and the toll is Dfl. 5 more expensive, because it costs Dfl. 10. This information has been
summarized for you in the table below.

If you have the choice between route A and route B, which would you choose?
Route A Route B
Toll 5 10
Number of deaths each year 16 12
Travel time 60 mins 50 mins
I choose for: Route A Route B

Table 1. Example of a survey question from De Blaeij (2003b). Dfl. means Dutch florins, the former
currency of the Netherlands.
The second type of question requested respondents to make a choice between three different
versions of the same car model, the only differences being the price and the safety. Then they were
asked which maximum price they were willing to pay for one of the versions. From this answer, a
safety valuation could be derived. This led to a higher VOSL ( 5 million, with a 95% confidence
interval of 2 to 9.5 million). There are different explanations for this difference. For example,
people are generally prepared to pay more for safety if it concerns a personal purchase such as a car,
than if they think that others will also profit from their purchase. The manner of payment (toll v. car
price) can also be of influence. Another possible explanation is that people prefer paying for safety
one time only, rather than having to make a trade-off regularly (every journey).

How much is the VOSL in the Netherlands?


The study by De Blaeij indicates that, among other things, the VOSL depends on the kind of measure
(public or private) and the target group for which it is intended. However, it is less desirable (less
credible), but also impossible (because of a lack of knowledge) to use a different VOSL in each case.
According to Wesemann et al. (2005) it is better to use one standard value that can also be used in
cases of a low VOSL. Therefore SWOV recommends using the lower of the two values found, i.e. a
VOSL of 2.2 0.3 million (price level 2001), for cost-benefit analyses and research into the costs
of road safety. In the Netherlands, this value is already used in calculations of the social costs of road
crashes. After deducting the material damage of fatal injury (loss of consumption during the years of
life lost, an average of 450,000 per fatality), there is nearly 1.8 0.3 million left for the human
costs of a road death. Based on analysis of foreign studies in which a VOSL for the Netherlands is
determined, Wesemann et al.(2005) recommend to, if desired, use an even larger margin for the
VOSL: 1.6 - 3.0 million in sensitivity analyses (price level 2001).
Because the VOSL can vary over the years, SWOV recommends carrying out research into the VOSL
periodically, e.g. every 5 years, as is done for valuating travel time. At present, correction is only made
for inflation, for example in determining the total human costs of road traffic crashes (see SWOV Fact
sheet Road crash costs). For 2009 the calculations have been made using a 2.6 million VOSL. In
addition, the VOSL should be revised for 'real' growth in time. This means that the VOSL is adapted to
for instance economic growth, as is also done for valuating travel time (VenW & CPB, 2004). This is
necessary for using VOSL in cost-benefit analyses, because the effects of measures can stretch over
a long period of time. However, at present no method has yet been found to correct the VOSL for real
growth in future years.

Publications and Sources


(SWOV reports in Dutch have an English summary)

AVV (2006). Kosten verkeersongevallen in Nederland; Ontwikkelingen 1997-2003. Adviesdienst


Verkeer en Vervoer AVV, Rotterdam.

Bickel, P., Friedrich, R., Burgess, A., Fagiani, P., et al. (2006). Proposal for harmonised guidelines.
Deliverable 5 of the EU project HEATCO. European Commission, Brussels.

Blaeij, A.T. de (2003a). De monetaire waarde van een statistisch mensenleven in een
verkeersveiligheidscontext. Researchmemorandum 2003-20. Vrije Universiteit, Amsterdam.

SWOV Fact sheet 4 SWOV, Leidschendam, the Netherlands


January 2012
Reproduction is only permitted with due acknowledgement
Blaeij, A.T. de (2003b). The value of a statistical life in road safety; Stated preference methodologies
and empirical estimates for the Netherlands. Tinbergen Institute Research Series, Vrije Universiteit,
Amsterdam.

Blaeij, A.T. de, Florax, R.J.G.M., Rietveld, P. & Verhoef, E. (2003). The value of statistical life in road
safety; A meta-analysis. In: Accident Analysis and Prevention, vol. 35, nr. 6, p. 973-986.

Blaeij, A.T. de, Koetse, M., Tseng, Y., Rietveld, P. & Verhoef, E. (2004). Valuation of safety, time, air
pollution, climate change, and noise; Methods and estimates for various countries. Report for the EU
project ROSEBUD. Vrije Universiteit, Amsterdam.

Dionne, G. & Lanoie, P. (2004). Public choice and the value of a statistical life for cost-benefit
analysis: the case of road safety. In: Journal of Transport Economics and Policy, vol. 38, nr. 2, p. 247-
274.

ECMT (1998). Efficient transport for Europe; Policies for internalisation of external costs. Organisation
for Economic Co-operation and Development OECD, Paris.

Eijgenraam, C.J.J., Koopmans, C.C., Tang, P.J.G. & Verster, A.C.P. (2000). Evaluatie van
infrastructuurprojecten; Leidraad voor kosten-batenanalyse. Ministerie van Verkeer en
Waterstaat/Ministerie van Economische Zaken, Den Haag.

Jost, G., Allsop, R., Steriu, M. & Popolizio, M. (2011). 2010 Road safety target outcome: 100,000
fewer deaths since 2001. 5th Road Safety PIN Report. European Transport Safety Council ETSC,
Brussels.

Lindhjem, H., Navrud, S., Braathen, N.A. & Biausque, V. (2011). Valuing mortality risk reductions from
environmental, transport and health policies; a global meta-analysis of stated preference studies. In:
Risk Analysis, vol. 31, nr. 9, p. 1381-1407.

Miller, T.R. (2000). Variations between countries in values of statistical life. In: Journal of Transport
Economics and Policy, vol. 34, nr. 2, p. 169-188.

Nellthorp, J., Sansom, T., Bickel, P., Doll, C. & Lindberg, G. (2001). Valuation Conventions for UNITE
(UNIfication of accounts and marginal costs for Transport Efficiency). Funded by 5th Framework RTD
Programme. Institute for Transport Studies ITS, University of Leeds, Leeds.

Smith, S.V. (2000). Jury verdicts and the dollar value of human life. In: Journal of Forensic Economics,
vol. 13, nr. 2, p. 169-188.

Wesemann, P. (2000). Kosten van de verkeersonveiligheid in Nederland, 1997. D-2000-17. SWOV,


Leidschendam.

Wesemann, P., Blaeij, A.T. de & Rietveld, P. (2005). De waardering van bespaarde verkeersdoden;
Covernota bij 'The value of a statistical life in road safety'. R-2005-4. SWOV, Leidschendam.

VenW & CPB (2004). Directe effecten infrastructuurprojecten; Aanvulling op de Leidraad OEI.
Adviesdienst Verkeer en Vervoer AVV/Centraal Planbureau CPB, Den Haag.

VenW & VROM (2004). Nota Mobiliteit; Naar een betrouwbare en voorspelbare bereikbaarheid.
Ministerie van Verkeer en Waterstaat/Ministerie van Volkshuisvesting, Ruimtelijke Ordening en
Milieubeheer, Den Haag.

SWOV Fact sheet 5 SWOV, Leidschendam, the Netherlands


January 2012
Reproduction is only permitted with due acknowledgement

You might also like