A wide variety of vehicle safety ratings have been developed since the 1970s and these have evolved largely independently of each other. Predictive systems provide information on the performance of new cars and equipment in various crash tests, whereas retrospective systems inform about cars already on the road on the basis of crash data. Predictive systems provide a more objective assessment of vehicle safety, but only for the conditions tested, whereas retrospective rating systems, when controlling for external factors, offer useful information on performance across the range of crash conditions and for all seating positions. Each system has been shown to usefully contribute to the provision of safety information to the consumer .
Predictive systems aim to assess a car's safety performance before it is used on the road. The predictions are based on controlled whole car crash tests of individual models; tests of components of the car which have been proven to be important in crashes; and/or visual inspections and rating of the interior of cars.
Consumer information based on crash tests started in Europe in the late 1980s with German motoring organisation and magazine publication of results of frontal crash tests. In the early 1990s in the UK, the WHICH? Magazine published the results of the Secondary Safety Rating System in Cars - a mix of visual inspection and component testing . This system later became the European Secondary Safety System which was used by the EU-wide umbrella organisation – the European Consumers organisation (BEUC) and International Testing .
New Car Assessment Programmes (NCAPs)
New Car Assessment Programmes (NCAPs) assess a new car's safety performance before it is used on the road. They have been established in the US, Australia, Japan and Europe and can be an important catalyst for improving vehicle safety. While tests vary over different NCAPs, predictions can be based on controlled whole car crash tests of individual models; tests of components of the car which have been proven to be important in crashes; and/or visual inspections and rating of the interior of cars. The aim of this information is to provide objective data to highlight the maximum level of protection available to car buyers and to complement regulation which, in EU Whole Vehicle Type Approval, should stipulate a high but minimum level of protection.
European New Car Assessment Programme (Euro NCAP)
Euro NCAP Established in 1997, Euro NCAP is the most sophisticated of all the new car assessment programmes. Euro NCAP provides motoring consumers, with an independent assessment of the safety performance of some of the most popular cars sold in Europe. Through its stringent vehicle crash testing, Euro NCAP has rapidly become a catalyst for encouraging significant safety improvements to new car design. Euro NCAP is backed by five European governments, the European Commission as well as motoring and consumer organisations in every EU country. Euro NCAP is acknowledged as the most advanced of all the current NCAP programmes, and the Australian New Car Assessment programme has aligned its protocols to it. Euro NCAP provides star ratings of the performance of new cars using state of the art crash tests and inspection protocols:
- Frontal impact tests using an offset deformable barrier intended to represent the most frequent type of road crash, resulting in serious or fatal injury. This test is a severe test of the car’s ability to survive the impact without suffering passenger compartment intrusion Example of a EuroNCAP crash test
- A side impact test addresses the second most important crash configuration of car to car side impact although the lower end of severe and fatal crash severity
- A pole test addresses head injury in side impact which is the most frequently seriously injured body region in side impacts;
- A child protection protocol is used to encourage manufacturers to take responsibility for protecting children and to provide suitable facilities for the fitment of child restraints
- Pedestrian protection sub-system tests based on those devised by the EEVC to assess protection afforded to the lower leg by the bumper, the leading edge of the bonnet and child and the bonnet top area. These replicate crashes resulting in severe injury involving child and adult pedestrians where impacts occur at 40kph. In general, the car industry has still to respond well to these tests in their designs
- Electronic Stability Control could drastically reduce crash occurrence, yet there are huge differences in the extent to which it is offered to car buyers across the EU. A rating is provided giving level of fitment for different models in different countries.
The European programme also uses visual inspection in addition to crash testing in determining the safety rating assessment. Star ratings are provide for adult occupant protection (5 stars), child protection (5 stars) and pedestrian protection (4 stars) for nine classes of vehicle from supermini to large off-road 4x4 vehicle. According to EuroNCAP, the highest ever score achieved for adult occupant protection is the Nissan Qashqai (5 stars 37 points), for child protection Toyota Prius (43 points) and for pedestrian protection, the Citroen C6 (28 points).
Retrospective vehicle safety ratings
Retrospective safety ratings can be of particular help in assisting buyers of used cars, which have the lion share of the car sales market . In retrospective systems, safety ratings are based on the actual performance of cars in real crashes. The frequency and severity of injury to car occupants in individual model cars are determined by examination of police crash statistics and/or insurance injury claim data. The earliest ratings to back to 1975 to those published based on insurance claims data by the Highway Loss Data institute . In general, they have been in use over the last 15 years.
While the general approach is the same for all systems, there are many differences in the exact methodology, such as the types of crashes included in the analyses, whether seat belt usage is accounted for, how the effects of exposure are controlled and whether or not the rating also takes into account the effects on other road users outside the vehicle. Aspects of the different methodologies and the adjustments made for exposure have been summarised [SARAC II]. The more these potentially confounding factors are controlled, the better the rating system .
Folksam Car Safety Rating System (Sweden)
Since the 1980s, Folksam has published injury risk ratings based on paired comparisons of car-to-car crashes from police reports where the injury outcome in both vehicles is considered.
Folksam’ colour coded safety rating:
Green: At least 30% higher safety than the average car
Blue: At least 15% higher safety than the average car
Yellow: At least as safe as the average car
Orange: At most 15% lower safety than the average car
Red: Worse than the “orange” group
In the 2005 rating, cars are allocated to one of four size groupings based on weight. For all cars an average crash safety rating is calculated. Early Folksam ratings indicated that if all cars were designed to be equal to the best current car in each class, 50% of all fatal and disabling injuries could be avoided . An analysis of Folksam data on car to car crashes in Sweden between 1994 and 1996 showed a decrease of 35% in the relative risk of fatal and severe injury associated with 'new' car designs compared with 'old' designs .
University of Oulu Passive Safety Ratings (Finland)
Since 1987, the Traffic Safety Committee of Insurance Companies (VALT) in Finland have regularly published ratings compiled by the University of Oulu comparing crash performance of cars in two-car collisions between passenger cars on Finnish roads. The rating concluded that if the crash protection of all the car models in the same weight class matched the best then 27% fewer drivers would be injured in urban car to car collisions .
Car and driver: injury accident and casualty rates (UK)
In 1991 in the UK the first edition of “Car and Driver: Injury Accident and Casualty Rates” was published giving information on comparative accident involvement and injury risks of popular makes and models of car . The rating, based on the risk of driver-only injury in car-to-car injury accidents reported to the police showed that if the safety of all models were improved to the level achieved or exceeded by the safest twentieth of models then the number of drivers injured in car to car accidents would fall by 12% and the number killed or seriously injured by 22%.
Used Car Safety Ratings (UCSR) (Australia)
The UCSR were developed by Monash University´s Accident Research Centre MUARC based on records of over 2.8 million crashes on Australasian roads. The UCSR rates cars according to their on-road crash performance and how well they protect drivers in a crash. Also rated is the risk each vehicle presents to other drivers involved in a crash with that particular model. The ratings are presented in governmental websites e.g. VicRoads Used Car Safety Ratings (Australia), the Transport Accident Commission and Land Transport New Zealand as well on websites of the Australasian motoring organisations.
Retrospective ratings: recent evaluations and future data needs
The Safety Rating Advisory Committee (SARAC)is an international forum initiated by the German insurance organisation GDV and the European Comité Europeén des Assurances (CEA). It brings together experts from the crash research community, government agencies, universities and automobile manufacturers. Research was undertaken in the SARAC 1 and SARAC II projects between 1999-2006 funded by the European Commission and the Comité Europeén des Assurances (CEA). In SARACII, safety ratings from around the world were examined to identify and develop advanced methods to assess crashworthiness and aggressivity and other aspects of statistical reliability, presentation of results and areas requiring further research.
SARACII indicated that an ideal retrospective rating should have:
- A measure of impact severity
- A range of variables that provide good proxies for impact severity if no measure is available
- Good data on non-vehicle variables that affect injury outcomes and differ from vehicle to vehicle
- Full reporting on injury and non injury crashes
None of the existing data sets on which rating systems are based meet these requirements in full. No existing rating has a measure of impact severity and it is not clear how well the available proxy measures represent impact severity. In addition to the need for action on assessing and recording impact severity, SARAC also highlights the need for action on the recording of vehicle annual kilometrage/mileage, the Vehicle Identification Number (as required in the US) and the availability of Event Data Recorders all of which would improve the retrospective rating data sets.
Apart from the the Folksam ratings, there is less promotion in Europe of retrospective ratings these days and this deficit needs to be addressed.