SWOV homepage
Nederlands  

Road Safety Developments in the Netherlands

 

Each year, SWOV analyses the road accident data in order to identify trends and to offer explanations. The analysis shows that over the last few years there has been a clear reduction in the number of road deaths in the Netherlands.

 

The number of in-patients also decreased, but far less rapidly. The policy that was aimed at sustainably safe traffic seems to be effective. However, extra efforts are necessary to ensure this will continue in the future.

 

Real number of casualties

With respect to accident data, SWOV has repeatedly carried out projects concerning their completeness and representativity. Together with AVV (the research body of the ministry of Transport) and CBS (the Dutch Central Bureau of Statistics) these efforts resulted in a combined approach to calculate the so-called real numbers of casualties, as from 1996.

 

Real number of fatalities

Until a few years ago, it was assumed that the numbers of traffic fatalities and fatal accidents as recorded by police were accurate. Then the results of a CBS-study were published, covering the year 1996, indicating that even these numbers were incomplete. In the study data from three different sources were compared: police registration, cause of death statistics, and juridical data concerning deaths. An extra 7% to 8% road fatalities complying with the definition used by the police (i.e. died as result of a traffic accident, on a Dutch public road, and within the 30-day limit) were detected.

This study also showed that completeness of the original police registration of fatalities differs with respect to type of road user. Underregistration of fatally injured cyclists is the highest (about 15%), while for most types of motor vehicle fatality underregistration is clearly less then average.

Since then, CBS and AVV annually publish these numbers as the official number of (traffic) fatalities and some tables (the distribution of type of road user, age, sex, etc.) are presented as well.

In practice, the less complete registered numbers of fatalities are often still used for time series and other analyses, since the real numbers only represent the years since 1996, and do not cover all possible distributions.

 

Real number of hospitalized

It has long been known that the registered number of hospitalized based on police registration is far lower than the real number of in-patients. This fact was established by at least two independent sources: regular enquiries and data from the Dutch national hospital data registration, called LMR (an continuous registration, owned by Prismant, an organisation working in the field of public health).

To establish the real number of hospitalized, SWOV has carried out several studies in which data from both police registration and LMR were statistically linked (matched).

Average completeness of the police registration appears about 60%. Here also, we see large differences with respect to type of road user, as well as collision type. Casualties from traffic accidents including motor vehicles, are far better registered than those including non-motorvehicles. The lowest degree of registration is found for hospitalised cyclists from single-vehicle accidents.

Based on these studies, SWOV developed a method to calculate the real number of hospitalized, based on both the LMR-data and the registered number. The method is used by AVV to determine the number of injuries.

The real annual number of hospitalized traffic casualties is about 18,000 (while the registered number is less then 12,000). The data concerning the real number of hospitalized is only available for a restricted number of variables and distributions, starting from 1985 up till now. For time series covering a longer period, and analyses on other variables, the registered numbers are still often used. It is clear that the results of those analyses should be used with care, considering the 40% average incompleteness, and the extra underrepresentation of cyclists.

 

Fatalities and risk

The last few years, the real annual number of traffic deaths in the Netherlands is slightly under 1100, while the registered number is already under 1000.

After a huge peak of more than 3250 road deaths in 1972, the number has decreased steadily.

Since 1950, the number of motor vehicle kilometres has increased by a factor 19. The death rate (deaths per billion vehicle kilometres) has decreased by a factor 20. This is illustrated in the graph.

 

Fluctuation of Motor vehicle kilometres, Road deaths and Death rate in the period 1950-2000

 

 

Short term prognosis

During the last seven years there has been a more rapid decrease in the number of road deaths and death rate than the average for the last two decades (see table). This is good news, especially if it means meeting the road safety target (which still is 750 deaths in 2010). Using our own prognosis model, SWOV has calculated that in 2010 there will be about 625 deaths, providing that the death rate continues to decrease by more than 5% per year (as it has during the last few years). To achieve this, efforts continue to be necessary. The picture looks less bright for the in-patients; with the present insights, the target of 14,000 in-patients in 2010 does not seem achievable, and much more effort is essential.

 

Period

Average annual fatality- risk change

1985-2000

1985 -1990

1990 -1995

1995 - 2000

-4,87%

-4,59%

-2,88%

-7,10%

 

Sustainably safe

It is actually too soon for an adequate evaluation of recently taken Sustainably Safe measures such as 'Moped on the Carriageway' and 'Priority Cyclists from Right'. In spite of this, the number of moped casualties seems to be declining since 1999.

Sustainably Safe has also resulted in a regional enforcement policy, with many more police traffic controls. This became clear from an extremely large increase in the number of traffic offences, mainly for speeding. A positive road safety effect is certainly to be expected from this if motorists' speeds really do get lower. The evaluation of the effect of these and other related enforcement activities is part of a current SWOV study.

 

Result of sustainably safe policy

On balance, we maintain that there are certain indications that Sustainably Safe, which made its entry in the 1990s, has had its results. We conclude this from the current faster decline in accidents and casualties than in the early 1990s.

Apart from the measures mentioned earlier, we advise the sustainably safe redesign of the urban and rural roads to further reduce the number of victims. This will result in a limited number of recognizable road types that 'invite' appropriate safe traffic behaviour. The application of roundabouts instead of regulated crossroads, the introduction of traffic calming 30 km/h areas, and other areas with a speed reduction are good examples. Motorways, and now through-roads, were already recognizable road types. The main features (a strict division of driving directions, no crossroads, only motor traffic) remain a guarantee for a traffic distribution that is as safe as possible.

Although the further sustainably safe redesign of the Netherlands road network still needs much investment, the decreasing death rate achieved up till now provides an emphatic indication that such investment is worthwhile.

 

The Dutch-language SWOV report R-2003-15 The road safety in the Netherlands up to 2002; Analysis of size, nature, and developments.
SWOV Research Activities 23 - September 2003

Knowledge base