Fact sheet

Road crash costs

Summary

Road crashes result in all kinds of social costs, such as medical costs, production loss, human losses, property damage, settlement costs and costs due to congestion. Studies into road crash costs and their trends are carried out quite regularly. In 2009, the costs amounted to € 12.5 billion, or 2.2% of the Gross Domestic Product (GDP). Insight into these costs is used for policy preparation and evaluation, and makes it possible to compare them with costs in other areas. Another important application is the use in cost-benefit analyses which use the costs per crash or casualty saved to assess road safety investments.

Background and content

In 2009, there were 720 fatalities in the Netherlands, and 18,580 serious road injuries. In addition, each year many more crashes result in less severe injury or in property damage only (PDO). This leads to all sorts of social costs, such as medical costs and production loss. Improving road safety has priority in mobility policy, not in the least because of these social costs [1]. Both in the Netherlands and in other countries research is regularly carried out into the size of these crash costs as well as into cost trends. The first study in the Netherlands was done in 1985 [2]. Commissioned by the Transport Research Centre (AVV) in the Netherlands, SWOV carried out the first update of this study in 1995, in which various aspects of the method were improved [3]. Since then, SWOV and the Centre for Transport and Navigation (WVL, previously called DVS and AVV) periodically study the costs of crashes in the Netherlands. Over the years the method has been improved further, for example, by including human losses and congestion due to crashes. Similar developments also took place in other countries.

This fact sheet discusses the different types of costs due to road crashes and the ways to calculate these costs. In addition, the most recently calculated costs of crashes in the Netherlands are presented and placed in (international) perspective.

Facts

What is the purpose of research into the costs of road crashes?

Information about the size and development of road crash costs is important to policy and research for two reasons. Firstly, this information is regularly used to prepare and assess the national road safety policy, like the realization of the Mobility Policy Document [4]. The information is also useful for comparing costs of road crashes with costs in other policy areas. These can be other sectors within traffic and transport or sectors in other areas, for example environmental policy, public health or other sectors concerned with safety policy. Information about the social costs provides insight into the possibilities for cost reduction and can be used for setting policy priorities. International comparisons of road crash costs are also made. Secondly, information about the costs of road crashes is used in cost-benefit analyses. See the (archived) SWOV Fact sheet Cost-benefit analysis of road safety measures. These analyses use the costs per crash or per casualty to express the road safety policy effects in terms of money. For example, the guideline Overview Effects Infrastructure (OEI), a guideline often used for cost-benefits analyses in the Netherlands, uses core road crash cost data [5]. Research into the costs of road safety is also carried out to determine prices for mobility. Estimates of traffic costs are used to determine the extent to which current pricing policy is efficient and which improvements are possible. CE Research and consultancy [6] and the European Conference of Ministers of Transport [7] investigated this subject. Their studies also examined other traffic costs than those of crashes, more particularly costs of infrastructure, environment and congestion.

Which types of costs are there and how should they be measured?

In 1994, an extensive international comparative study was published about the way in which 14 European countries determine the costs of traffic crashes: the COST 313 project [8]. The report made recommendations about the categories of costs that should be studied and how the costs in each category should be measured. In brief [9], this resulted in five main categories, each with its own measuring methods:

  1. Medical costs. These result from the treatment of casualties, e.g. costs of hospital, rehabilitation, medicines and adaptations for the handicapped. These costs must be determined by the 'restitution costs method'. All countries do this in practically the same way.
  2. Production loss. These costs result from the temporary or permanent disability of the injured, and the complete loss of production of fatalities. These costs have to be determined by the 'human capital method', which determines the value of the production that would have been realized by the casualties if they had not been injured or killed. This is also done in more or less the same way in all countries. However, important differences can occur: some countries also put a value on domestic work, some countries correct the production loss of fatalities for their consumption loss, and there are various assumptions regarding the growth of incomes [10] [11].
  3. Loss of quality of life of casualties and their families and friends ('human costs' or 'human losses'). These are the immaterial costs through suffering, pain, sorrow and loss of the joy of living. It is recommended to determine these costs with one of the 'willingness-to-pay methods'; these methods determine the amount of money people are prepared to pay to prevent human suffering. In general, most countries recognize this cost category, but the calculation methods differ widely, and only a few countries use a willingness-to-pay method. The willingness to pay for a reduction in the risk of dying also includes the economic value of consumption loss. If this is also part of the production loss, the willingness-to-pay must be corrected to avoid counting it twice.
  4. Property damage. These costs result from damage to goods such as vehicles, freights, roads and fixed roadside objects. For this the 'restitution costs method' is recommended. This is used in the same manner everywhere.
  5. Settlement costs. These are the costs that result from the assistance with and settlement of crashes. These expenses are incurred by organizations such as the fire brigade, police, law courts and insurers. Also for these costs the 'restitution costs method' is recommended. This is already generally used.

This state of affairs in 1994 is mainly the result of developments in the knowledge about valuing 'imponderables': cost elements for which there are no market prices. Elvik [10] distinguishes four phases in the developments since the 1950s: it was only during the last phase in the 1980s that a number of countries began to value 'human losses' based on a willingness-to-pay method. This approach is, theoretically, very well justifiable. However, empirical research into the willingness to pay for traffic risk reduction is complicated and provides results with many uncertainties. Since 1994 an increasing number of countries have nevertheless used a willingness-to-pay method (see the archived SWOV fact sheet The valuation of human losses of road deaths). This was sometimes accompanied by improvements in this method.

How are the crash costs measured in the Netherlands?

Nowadays, SWOV and DVS carry out the crash costs research according to the recommendations of the COST 313 report in practically all aspects. A sixth category was added: the costs resulting from congestion; these costs are caused by, among other things, loss of time due to traffic-jams resulting from crashes. In the Netherlands, the 'real' numbers of crashes and casualties are used, i.e. allowing for those crashes and casualties that were not registered.

  1. Medical costs. Various data sources are used to determine these costs, including data from the National Medical Register (LMR; inpatients), the Injury Information System (LIS; Accident& Emergency), the Accidents and Exercise in the Netherlands (ObiN monitor; treatment by GP) and Statistics Netherlands data (cost price). This includes, for example, the average number of days that a casualty is hospitalized, the average costs per day of hospital or nursing home treatment and the annual number of ambulance trips.
  2. Production loss. The potential loss of production is calculated, i.e. the monetary value of the contribution somebody would have made if he or she had not been injured or killed. Here, it does not matter if the individual casualties actually were employed before the crash, or would have been employed in the future. In the case of fatalities, the total value of the production over the lost productive years is estimated and the present value is calculated, i.e. the production is weighed over those lost years. As yet, no allowance has been made for unpaid work, such as domestic or voluntary work. However, the consumption loss of fatalities is included in the production costs.
  3. Loss of quality of life. To determine the human losses a survey was held in the Netherlands about the amount of money people are willing to pay for a certain reduction in crash rate [12]. This study determined the so-called 'Value Of a Statistical Life' (VOSL) which is used to calculate the human losses. The VOSL is corrected for the consumption loss of those killed, because these costs have already been included in the category production loss [13].
    So far little is known about the human losses as a result of non-fatal injury; studies have only been carried out in Great Britain [14], Sweden [15] and Belgium [16]. Based on the British study, the human losses per serious road injury are estimated at 12% of those of a fatality [17].
  4. Property damage. The estimation of these costs is based on insurance data, such as damage claims paid, and estimates of the damage not claimed and the damage not compensated.
  5. Settlement costs. Statistics Netherlands data, data provided by the Research and Documentation Centre (WODC) of the Ministry of Security and Justice, studies into the time spending by the police, and insurance data are among the sources used to estimate these costs.
  6. Congestion costs. Research data into the total traffic jam costs and the share of lost time due to crashes is used to estimate these costs. The time loss due to traffic jams as a result of crashes is based on data about congestion intensity. In 2009, about 11% of the congestion intensity was the result of crashes [18]. Furthermore, a multiplication factor is used for the other costs resulting from congestion, such as costs caused by the unreliability of travel time (see below).

The above method of calculation and data gathering was used in the most recent studies into the costs of road crashes in the Netherlands [17] [18]. This method differs in several aspects from the methods that were used previously [19].

The reasons are that data sources that were used previously are no longer available and/or that better calculation models have become available. Furthermore, the definitions of the different categories of casualties and crashes have been changed, as a consequence of the new definition of serious road injuries [i]. Therefore, in many ways the outcomes of the most recent cost calculations (discussed in the following section) cannot very well be compared with the results of previous studies. The most recent calculations have recalculated the costs in previous years, so that a reliable picture is created of the development of costs in time.

In the Netherlands, research has also been done into prevention costs, such as those for infrastructural measures, enforcement, vehicle safety and research [20]. They are, however, not included in crash costs, as they are meant to prevent crashes.


[i] A serious road injury is a road casualty who has been admitted to hospital with a minimum MAIS injury severity of 2 and who has not died of the consequences within a period of 30 days after the crash. MAIS is an international measure used to indicate injury severity.

How high are the crash costs in the Netherlands?

In 2009, the last time such research was carried out, the total crash costs in the Netherlands were estimated to amount to € 12.5 billion [18]. The largest categories are human costs (€ 5.8 billion), and property damage costs (€ 3.9 billion). The settlement costs are € 1.3 billion) and the production loss amounts to € 0.9 billion. Medical costs and congestion costs are a relatively small proportion of the costs (see Table 1).

Table 1. Social costs of road crashes (€ million, current prices) [18].

 

In 2009, the costs are about the same as in 2003, whereas the costs in 2006 are approximately € 1.5 billion lower. One of the explanations for the lower costs in 2006 is a smaller number of casualties. This causes a reduction in most cost categories, even though the costs per casualty or crash have increased as a result of (among other things) a general increase in costs. Part of the considerable decline in 2006 of the production loss can be attributed to changes in legislation which caused a reduction in casualties receiving disability benefits. However, this does not mean that casualties who had previously been declared to be disabled, are now more rapidly (fully) productive. The production loss in 2006 has probably been underestimated and this is also the case for 2009, as the disability benefit data do not give a complete picture of the production loss [18]. In 2009, the property damage and the human costs in particular are higher than in 2006. The increase in human costs is related to an increase in the number of serious road injuries. In 2009, three quarters of the human costs are due to the serious road injuries. This has caused an increase in the total amount of human costs, despite a decrease in the number of fatalities. The increase in property damage is attributed to a rise in the number of damage claims between 2006 and 2009. The average amount claimed, on the other hand, shows a declining trend, which results in a decrease of these costs between 2003 and 2006.

 

What are the costs per road crash casualty?

The recent study also calculated how the costs are distributed among the different categories of casualty and crash severities, and how much the costs are per casualty or crash [18] (see also Table 2). The majority of the costs can be attributed to  serious road injuries (€ 5.2 billion) and PDO crashes (€ 3.6 billion; not included in the table). The costs for road deaths amount to € 1.9 billion. Furthermore, the costs have been estimated of slightly injured (MAIS0-1) casualties who have been admitted to hospital or received treatment as Accident & Emergency (A&E) patients, and of other casualties with less severe injuries who have not received any hospital treatment. The costs of slightly injured casualties who received some form of hospital treatment are € 0.9 billion and the costs of other slight injury are also estimated to be approximately € 0.9 billion (not included in the table). The costs per fatality were by far the highest (€ 2,6 million), especially because the human costs are important in this category.

Table 2. Number of casualties, and costs by crash severity and per casualty in 2003 (in € million) [18].

 

The decline in the number of road fatalities reduced the total costs for fatalities in 2009, despite higher costs per fatality, by more than 25% compared to 2003 when these costs amounted to € 2.6 billion [18]. The costs for serious road injuries, on the other hand, are about 10% higher than in 2003 due to an increase in the number of serious road injuries. The share of the total costs for road fatalities therefore declined from more than 20% in 2003 to 15% in 2009, whereas the share for serious road injuries increased from 38% to 42% during the same period.

The costs per casualty are for instance used in cost-benefit analyses to express the road safety effects of measures.An amount per fatality is sometimes used to make a rough estimate, in which the costs of (all) casualties and PDO crashes are also included. This amount (€ 17 million per fatality in 2009) is calculated by dividing the total costs of road crashes (€ 12.5 billion) by the number of fatalities (720). However, this is a very rough estimate, because only the effect of the measure concerned on the number of fatalities is taken into account, not the deviating effects on injuries and PDO crashes. For this reason, SWOV, together with DVS, has fine-tuned the method, among other things as an input for an addition to the OEI guideline for (road) safety [21]. This improved method distinguishes fatalities and serious road injuries. The costs per fatality are used (€ 2.6 million in 2009) as well as an amount per serious road injury (€ 0.53 million in 2009). The latter amount also includes the costs of slight injury and the costs of PDO crashes. The amount is calculated by dividing the total costs of (all) injuries and PDO crashes (€ 9.9 billion) by the number of serious injuries (18,580). These figures can be used in a cost benefit analysis if the effects on the number of fatalities as well as on the number of serious injuries are known. This method can only be used when time and money for a cost-benefit analysis of all individual measures on crashes of different severities are lacking.

What does comparison with other traffic-related costs show?

In 2009 the costs of road crashes amounted to 2.2% of the Gross Domestic Product (GDP). This was 2.6% in 2003 and 2,0% in 2006. A comparison with other social costs also gives insight into the relative magnitude of the amount. This shows that the costs of road crashes are high in comparison with other traffic-related costs like those of congestion and environmental costs.

The Netherlands Institute for Transport Policy Analysis (KiM) [22] estimates the costs due to congestion on the main road network at € 2.4 to 3.2 billion in 2009.These are the costs due to direct loss of travel time and unreliability of travel times, costs that are caused by people adapting their behaviour to traffic jams (detour, another time of day, etc.), extra fuel costs, and indirect costs like the effects on public transport. Recent KiM data shows that in 2011 the congestion costs are at about the same level as in 2009, (€ 2.3 to 3.0 billion). This is a considerable reduction (18%) in comparison with the 2010 congestion costs caused by a decrease in the number of traffic jams [23].

The environmental costs of road traffic are estimated at € 2.0 billion to € 8.5 billion in 2009 [22]. These are the costs of air polluting emissions, CO2 emissions and noise nuisance.

What are the results of international comparison?

Many countries do research into crash costs and there are various international comparative studies. A sound comparison requires sufficient knowledge about the following issues: Have the same cost categories been studied? Have the same measuring methods been used? Is the data used of comparable quality?

The COST 313 study provides outline answers to the first two questions. With exception of the human losses and their measuring methods, the fourteen countries used the same cost categories and measuring methods. In spite of this, determining the production loss (by valuing unpaid work and correction for consumption loss of those killed) can still result in large differences [10]. COST 313 did not answer the question whether the data was of a comparable quality. Answering this question is practically impossible because of the large number of divergent sources used for these studies. However, one of the comparative studies [24] investigated whether the cost estimates referred to all crashes, i.e. also property damage only crashes and crashes not registered in official statistics. In the light of all this, the results of international comparisons must be used carefully. In addition, a distinction should at the very least be made between the countries that do and those that do not include human losses.

Gezien al het voorgaande moeten de uitkomsten van internationale vergelijkingen met de nodige reserves gebruikt worden. Verder zou op z'n minst onderscheid gemaakt moeten worden tussen landen die wel en die geen rekening houden met immateriële kosten.

Elvik [24] compared the costs of eight European and four non-European countries. All these countries had estimated human losses in one way or another. The data of the eight European countries referred to one single year in the 1990-1997 period. For these countries their costs, including human losses, varied from 1.3% to 3.2% of the Gross Domestic Product (an average of 2.1%). Excluding human losses, this was 0.5% to 2.8% (an average of 1.3% of the GDP).

Trawén et al. [11] studied the costs per fatality in nine European and two non-European countries. They indicate that the average costs in these countries had increased by about 70% in the period 1990-1999. This increase can for example be attributed to changes in calculation methods and adding cost categories such as human losses. Although the costs and methods that are included increasingly resemble each other, large differences remain between the estimates in the different countries. This is illustrated by an overview of costs per casualty in ten European countries that was made in the Road Safety and Environmental Benefit-Cost and Cost-Effectiveness Analysis for Use in Decision-Making (ROSEBUD) project [25]. For the five countries (other than the Netherlands) that included human losses, the costs per fatality ranged between more than € 1.9 million in Switzerland to € 3.0 million in Norway (price level 2002). The costs per seriously injured casualty ranged between € 169,000 (Switzerland) to € 474,000 (Norway). These differences can in part be explained by the use of different calculation methods and cost categories.

In 2014, SWOV made an international comparison of the methods used in different countries to determine the costs of road crashes [26]. This study compared the calculation methods used in eight different countries, among which the Netherlands, Belgium, Germany, the United Kingdom and the United States. The study showed that different countries use different cost items. At present some costs that are generally used in other countries are underestimated or not included in the Netherlands. These are the costs of minor injuries and production loss due to injuries, and a share of the property damage. These cost items may in the future also be used in the calculation method for the Netherlands.

Conclusions

In 2009, the road crash costs in the Netherlands were € 12.5 billion, or 2.2% of the Gross Domestic Product. The majority of these costs, € 5,2 billion, can be attributed to serious road injuries. The share of the total costs for road fatalities has been decreasing in recent years; these costs amounted to € 1.9 billion in 2009. With € 5.8 billion, human costs are the largest cost category, whereas property damage costs (€ 3.9 billion) are also a considerable share of the total costs. Insight in these costs is important for preparing and evaluating (road safety) policy.

Publications and sources

1. VenW (2008). Strategisch Plan Verkeersveiligheid 2008-2020: Van, voor en door iedereen. Ministerie van Verkeer en Waterstaat, 's-Gravenhage.

2. McKinsey (1986). Naar een slagvaardig verkeersveiligheidsbeleid. Eindrapport. McKinsey&Company, Amsterdam.

3. Muizelaar, J., Matthijssen, M.P.M. & Wesemann, P. (1995). Kosten van verkeersonveiligheid in Nederland, 1993. R-95-61. Stichting Wetenschappelijk Onderzoek Verkeersveiligheid SWOV, Leidschendam.

4. VenW & VROM (2004). Nota Mobiliteit. Deel I: naar een betrouwbare en voorspelbare bereikbaarheid. Ministerie van Verkeer en Waterstaat/Ministerie van Volkshuisvesting, Ruimtelijke Ordening en Milieubeheer, 's-Gravenhage.

5. 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, 's-Gravenhage.

6. CE (2004). De prijs van een reis; Maatschappelijke kosten van het verkeer. CE Oplossingen voor milieu, economie en technologie, Delft.

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

8. Alfaro, J.-L., Chapuis, M. & Fabre, F. (1994). Socioeconomic cost of road accidents. Transport Research COST 313. Commission of the European Communities, Brussels/Luxembourg.

9. Elvik, R. (1997). A framework for cost-benefit analysis of the Dutch road safety plan. TØI 380/1997. TØI, Institute of Transport Economics, Oslo.

10. Elvik, R.(1995). An analysis of official economic valuations of traffic accident fatalities in 20 motorized countries.  In: Accident Analysis and Prevention, vol. 27, nr. 2, p. 237-347.

11. Trawén, A., Maraste, P. & Persson, U. (2002). International comparison of costs of a fatal casualty of road accidents in 1990 and 1999. In: Accident Analysis and Prevention, vol. 34, nr. 3, p. 323-332.

12. Blaeij, A.T. de (2003). The value of a statistical life in road safety; Stated preference methodologies and empirical estimates for the Netherlands. Research Series. Vrije Universiteit, Amsterdam.

13. 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.

14. Hopkin, J.M. & O’Reilly, D.M. (1993). Revaluation of the cost of road accident casualties: 1992 revision. TRL Research Report 378. Transport Research Laboratory, Crowthorne.

15. Persson (2004). Valuing reductions in the risk of traffic accidents based on empirical studies in Sweden. PhD Thesis. Lund Institute of Technology, Lund.

16. Brabander, B. de (2006). Valuing the reduced risk of road accidents. Empirical estimates for Flanders based on stated preference methods. PhD Thesis. Hasselt University, Diepenbeek.

17. Wijnen, W. (2012). Bouwstenen voor berekening van de kosten van verkeersongevallen 2003-2009; Materiële en immateriële kosten en kosten van afhandeling. D-2012-4. Stichting Wetenschappelijk Onderzoek Verkeersveiligheid SWOV, Leidschendam.

18. Wit, M. de & Methorst, R. (2012). Kosten verkeersongevallen in Nederland; Ontwikkelingen 2003-2009. Directoraat-Generaal Rijkswaterstaat, Dienst Verkeer en Scheepvaart DVS, Delft.

19. AVV (2006). Kosten verkeersongevallen in Nederland; Ontwikkelingen 1997-2003. Directoraat-Generaal Rijkswaterstaat, Adviesdienst Verkeer en Vervoer, Rotterdam.

20. Wijnen, W. & Stroeker, N.E. (2009). Uitgaven aan verkeersveiligheid; Een schatting voor 2007. R‑2009-17. Stichting Wetenschappelijk Onderzoek Verkeersveiligheid SWOV, Leidschendam.

21. Linde, G. van der, Donkelaar, J., van, Witmond, B., Devillers, E. & Spit, W. (2012). Veiligheid in maatschappelijke kosten-baten analyse; Werkdocument. Directoraat-Generaal Rijkswaterstaat, Dienst Verkeer en Scheepvaart, Rotterdam.

22. Jorritsma, P., Bakker, P., Berveling, J., Derriks, H. et al. (2010). Mobiliteitsbalans 2010. Kennisinstituut voor Mobiliteitsbeleid, Ministerie van Infrastructuur en Milieu, Den Haag.

23. Savelberg, F., Bakker, P., Derriks, H., Francke, J., et al. (2012). Mobiliteitsbalans 2012. Kennisinstituut voor Mobiliteitsbeleid, Ministerie van Infrastructuur en Milieu, Den Haag.

24. Elvik, R. (2000). How much do road accidents cost the national economy? In: Accident Analysis and Prevention, vol. 32, nr. 6, p. 849-851.

25. Hakkert, S. & Wesemann, P. (eds.) (2005). The use of efficiency assessment tools: solutions to barriers. Workpackage 3 of the European research project ROSEBUD. R-2005-2. SWOV Institute for Road Safety Research, Leidschendam.

26. Wijnen, W. (2014). Kosten van verkeersongevallen in internationaal perspectief. R-2014-6. Stichting Wetenschappelijk Onderzoek Verkeersveiligheid SWOV, Den Haag.

Printer-friendly versionSend by email

Updated

01 Mar 2014