Behavioural research in road safety: tenth seminar proceedings

Back to contents

6 A study of the accidents and behaviours of company car drivers

Peter Chapman, Katherine Roberts & Geoffrey Underwood, Accident Research Unit, School of Psychology, University of Nottingham, Nottingham NG7 2RD


Research has suggested that non-professional drivers who drive for business purposes may be at especially high risk of accident. We describe a questionnaire study which obtained accident reports, self-reported driver behaviours and additional details from more than 500 employees within a large organisation. Five employee types were sampled, differing in employee status and business mileage requirements. Preliminary results suggest that these groups differ in self-reported driving behaviours, accident frequency, and in the types of accident reported. The broad differences in accident frequency accord well with those reported in other studies, but this research sheds new light on the increased accident rates in company car drivers by exploring variations in the types of accident experienced by these drivers and the differences between different groups of employees.

The accident liability of company car drivers

Many researchers have recently concluded that company car drivers are at an increased risk of accident relative to the general population (eg Dimmer & Parker,1999; Grayson,1999; Lynn & Lockwood,1998). Lynn and Lockwood suggest that the company car driver is 49% more likely to be involved in an accident than an ordinary driver, even after demographic variables and their relatively high exposure (in terms of mileage) is taken into account. Dimmer and Parker (1999) conducted a recent survey of 441company vehicle drivers within a single large organisation and found that 27% of these respondents reported at least one accident while driving over the last three years, compared to a level of 18% from the wider driving population. These kind of estimates accord well with the general public perception that company car drivers represent a real road safety problem.

The reasons for company car drivers' increased accident liability are poorly understood and interventions aimed at reducing risk to company car drivers have a relatively poor history of success when they are rigorously evaluated. Adams-Guppy and Guppy (1995) suggest that company car drivers may be especially subject to strong time demands. They conclude that such demands may seriously impact on the driver's decision making, particularly with respect to speed choice and decisions about overtaking. This may typify the conflict between haste and safety identified by Wagenaar and Reason (1990) as highlighting system failure types such as incompatible goals and conditions promoting unsafe behaviour. Adams-Guppy and Guppy found that company car drivers who reported frequently exceeding the speed limit on motorways viewed speeding as a less important risk factor and viewed being on time for appointments as particularly desirable.

Dimmer and Parker (1999) explored the attitudes and self-reported behaviours of a sample of company car drivers more formally using the Driver Behaviour Questionnaire (DBQ) (Reason, Manstead, Stradling, Baxter & Campbell,1990). They used a shortened DBQ intended to reflect the additional distinction between aggressive and other violations as well as the traditional error and lapse factors (Åberg & Rimmö,1998; Lawton, Parker, Manstead, & Stradling,1997; Parker, Reason, Manstead & Stradling,1995; Parker, West, Stradling & Manstead,1995). Interestingly, they found that their sample of company car drivers produced a six factor solution when the reduced DBQ was subjected to a factor analysis. In this new factor structure the basic error (or mistake) factor emerged and not only were violations broken into the anticipated two types (aggressive and other), but lapses were additionally divided into two types (which Dimmer and Parker termed action slips and inattention lapses) and a final sixth factor emerged which they tentatively identified as relating to the driver being under pressure. This final factor appeared to be of particular interest in the context of company car drivers, since it might theoretically relate to the demand issues identified by Adams-Guppy and Guppy (1995). Moreover, this sixth factor was the only factor in their study to significantly differentiate between accident-involved and accident-free drivers.

Lynn and Lockwood (1998) found that drivers with high work mileages also drive significantly elevated non-work mileages and that the increased accident liability of this group appears to extend to non-work driving. If anything, it appeared that their drivers were actually safer when driving for work purposes, but it is difficult to adequately take into account differences in mileages and road types when making this comparison. Nonetheless, this suggests that part of the company car driver problem may be to do with a habitual driving style adopted by these people rather than the pressures of work suggested by Adams-Guppy and Guppy (1995). Clearly, this is an important distinction to understand because these different theoretical explanations for the elevated risk of company car drivers suggest very different intervention strategies at an organisational level.

Lynn and Lockwood's (1998) survey analysis found no significant evidence that attending a training programme reduced a driver's accident liability, although there was some suggestion that companies which offered small financial rewards for accident free driving had slightly lower accident rates. Survey research is obviously limited in its ability to evaluate training and incentive schemes because of self-selection effects - drivers who voluntarily undertake training may often be those least in need of it. To rigorously evaluate the effectiveness of different intervention strategies, carefully controlled experimental studies are required. Gregersen, Brehmer & Moren (1996) evaluated four different measures (individual driver training, group discussions, campaigns, and bonuses for accident free driving). The latter two measures, while costly and benefiting from considerable face validity, produced only small benefits, compared to a control group. Nonetheless, the other two measures, driver training and group discussions, did produce significant reductions in accident rates, Gregersen (1999) explores possible reasons for these benefits, concluding that a focus on risk awareness may be a particularly important feature in interventions aimed at company car drivers.

One issue in much of the above work is the definition of a company vehicle driver. Lynn and Lockwood (1998) restricted their sample to 'drivers of company owned or company financed cars who drive regularly for business purposes'. Dimmer and Parker (1999) recognised that company car drivers include a wide range of road users from senior executives provided with a company car as a perk of the job, through those who drive non-liveried company owned vehicles both for work and non-work purposes, to those employed to drive fleet cars, vans and other specialist vehicles. It is clear that these groups cannot be treated as a homogenous whole. Not only may different interventions be required for the different groups, differences in behaviour and accident involvement between these groups may provide important new understanding of the reasons behind company car drivers' increased accident liability.

A questionnaire study

The study involved sending a questionnaire to a wide selection of employees within a single large organisation. The aim was to sample a variety of different types of company vehicle drivers, with the hope that differences in accidents between different types of driver would shed additional light on the reasons for any heightened risk of accident in company car drivers and provide suggestions for interventions to reduce individuals' risk of accident.



The participants in this study were a variety of employees from a single large organisation. Employees were selected to all be regular drivers of cars or vans and were divided into five broad classes based on the ownership and type of the vehicle most commonly driver, the reasons that the vehicle is supplied to the individual, and the likely journey purposes when the vehicle is used:


Own Car Drivers (OCD) - a group of employees within the organisation who drive predominantly in a vehicle owned by themselves or their spouses; most of this group will regularly do business mileage in this vehicle.



Company Car Drivers A (CCDA) - These are middle and senior managers who have the option of receiving a company car as part of their remuneration package. They are not required to meet a minimum business mileage.



Company Car Drivers B (CCDB) - These are first and second line managers who receive a company car if it is required for business purposes.



Company Car Drivers C (CCDC) - These are sales staff within the organisation who receive a car to allow them to drive for business purposes.



Liveried Company Vehicle Drivers (LCVD) - These are a group of employees who drive a liveried car, van, or 4x4 vehicle for work purposes. Recent policy within the organisation allows these drivers to take these vehicles home when it is convenient.


Two thousand questionnaires were sent out to the employees within the organisation. The questionnaire contained sections addressing a wide selection of driving and organisational issues. In this paper we will report preliminary data from three sections of the questionnaire: Background information, self reported driving behaviours, and accident histories. The background information section of the questionnaire was concerned with the types of vehicle driven by the individual, the driving conditions they most commonly faced and the purposes of the journeys that they made. The self reported driving behaviour section consisted of the 28 item version of the Driver Behaviour Questionnaire as used by Dimmer and Parker (1999). The accident history section started with the following details:

'We are interested in the type of road accident you have been involved in as a driver over the past three years. By 'accident' we mean any incident which involved injury to another person or yourself, damage to property, damage to another vehicle, or damage to the vehicle that you were driving. Mention only those accidents in which you were involved as a driver, not as a passenger. Please include all accidents regardless of how they were caused or how slight they were.'

Respondents first reported the total number of accidents they had been involved in over the three year period, and then gave detailed descriptions of these accidents (up to a maximum of three). For the accidents where detailed descriptions were obtained respondents also answered a series of specific questions to determine possible contributory factors for these accidents.


By the end of February 2000, 635 questionnaires had been returned (a response rate of 31.8%), 592 of these respondents indicated that they were male, and 41that they were female. Not all respondents answered all questions. In the tables and analyses that follow individuals who failed to answer particular questions are excluded from analyses that require those particular data (eg the two respondents who failed to indicate gender). Table 1shows the distribution of respondents in terms of the vehicle type that each respondent drove predominantly. It should be noted that although only12.6% of the sample predominantly drive their own vehicle, many of the other respondents were still driving their own vehicles for a substantial part of their everyday driving. Table 2 shows the actual driving done by each group divided up by the journey purpose. The last three groups, CCDB, CCDC, and CLVD, show a similar pattern of driving, all doing relatively high mileages with more than 70% of their driving being done for business purposes. The own car drivers and CCDA groups are notable, both in their generally lower mileages overall, but also in the higher proportion of their driving which is done for purposes of commuting and leisure.

Table 1: Respondents divided by principal vehicle driven

Driver Group


















Back to top