Authors: Plema Seferagic Dymbe, Rune Undheim, Solveig Johannsdottir Wikstrand, Thusanee Tjørhom, Jouke Gaastra

Introduction

Last February the online paper Nettavisen Økonomi (Blaker, 2023) published an article describing the effect of a decline in the use of studded tires. Studless winter tires have become more popular in recent years because they cause less road wear, driving noise and pollution (Elvik and Kaminska, 2013). This shift doesn’t just revolve around how well studless tires perform on icy roads; it’s also about how they might make roads even more slippery, a situation referred to as “road polishing’’. In this article, we will delve deeper into the safety and behavior aspects related to driving with or without studded tires in winter conditions in Norway. We will pursue the question of whether drivers take more risk when using studded tires. A theory referred to as ‘Risk compensation’. The question we will try to answer is:

How are the number of accidents influenced by the use of studded tires compared to studless tires and to what extent is this effect offset by the risk compensation behavior of the driver?

In Norway, studded tires have been used since the 1960s (Wikse, 2022). Studded tires are intended to provide better traction and shorter stopping distances on icy roads compared to studless tires because of their studs that increase grip on icy roads. The use of studded tires is allowed in Norway from October 1st until the first Monday after Easter in Southern Norway, and in Northern Norway, their use is permitted from October 15th to May 1^st-  (Bruk av piggdekk – Lovdata, 2012). The adoption of studded tires has significantly declined since the 1990s, influenced by factors such as studded tire fees in major cities and the availability of improved winter and all-season tires.

Scheibe (2002) found that a “number of driver behaviour issues have been postulated to affect the judgment of studded tire effectiveness. There is no consensus on these points: 1) drivers with studded tires care more about safety, hence they drive more safely, 2) they drive faster (because of a false sense of security or confidence), and 3) drivers with studless tires avoid driving when weather is severe’’ (Scheibe, 2002, p.12). While Scheibe cannot conclude about driver behavior using studded tires, Elvik and Kaminska find that drivers adapt behavior to adverse weather (Elvik et al, 2013). They conclude that reduced use of studded tires in Norwegian cities is associated with an increase in the number of injury accidents. They also see that fewer people are using studless tires in big Norwegian cities, and if the decrease is too great, the accident rate on winter roads will increase. Icy roads will stay icy if fewer people drive with studs since studded tires roughen the ice and make it safer for studless drivers (Scheibe, 2002). The varying perspectives underscore a knowledge gap in understanding how drivers adapt their behavior in response to different tire choices under winter conditions. Consequently, the extent to which risk compensation influences accident rates remains an open question, necessitating further research to bridge this gap and provide more insights into the dynamics of studded tire usage and its implications on road safety.

Risk perception

“Human behavior is primarily driven by perception and not by facts” (Renn, 2008, p. 93). Perception is the way an individual understands his/her surroundings, which also includes the individual’s understanding of risk. This understanding is heavily influenced by personal experiences, culture, commonsense and communication between actors (Renn, 2008). It is also shown that individuals link certain emotions like being scared, feared, relaxed, and so on, with certain activities that have some sort of consequence and some uncertainty to it. So based on these elements of risk perception, different individuals understand risk in very different ways. This also includes how the individual perceives the degree of seriousness of the risk. We are talking about a level of acceptability when we perceive risk. A base jumper sees jumping off a roof with a parachute as a risk that he thinks is acceptable but lay people would deem this as an unacceptable risk. The acceptability and seriousness of the risk are also affected by some of the elements we mentioned above; feelings and values, but also include knowledge and judgements. These mental images are in constant change. What individuals deem as risky or not will change during their lives, as their knowledge increases, as they get to see or experience more events that could be perceived as risky, and as their feelings change that could alter their judgement about certain aspects of life (Renn, 2008).

Two factors that influence our risk perception are of particular importance for the current paper. The first one being the voluntariness (Renn, 2008, p. 94). Meaning that we perceive the risk as less serious if we have the opinion that we can control the risk. The second is the familiarity of a risk, which creates a naturalness over the situation. As more one participates in an event, the safer one would feel participating in said event. For example, almost every adult person drives a car, which is a statistically risky activity to participate in, but there are not many that are directly scared of driving a car, compared to flying a plane for example (Renn, 2008, p. 107).

Risk compensation and risk homeostasis in relation to driving

For many years researchers have investigated the reasons why people make choices involving risk. Various theories attempt to explain how people make decisions based on their risk perception and acceptance. In the present article, we will dwell further into the theories of Risk compensation by Samen Peltzman, (1975) and the Homeostasis theory by Gerald J S Wilde (1998). Both theories try to explain behavior adjustments following the introduction of safety measures.

The theories propose that people seem to maintain a certain level of risk acceptance based on their empiricisms and personalities. The degree to which this behavior adjustment, often referred to as offsetting, occurs is contested. However, there is consensus on the existence of the phenomena (Hause, 2006). The two theories have different names, but according to Pless (2016), both Wilde and Peltzman speak about the theories in an interchangeable way. This paper is not about differentiating the minor differences between the two theories, we will – as many others, speak about the two theories interchangeably.

Risk homeostasis theory (RHT) will guide us to understand drivers’ processes around behavioral changes during driving. Risk homeostasis is a psychological approach towards how individuals change their behavior to gain health and safety benefits. As mentioned earlier we are exposed to all sorts of risks every day and in almost every activity. In RHT we define subject risk as how much risk an individual feels to be exposed to in an event or activity.

There are two main elements which affect a driver’s attitude during driving according to Wilde. The first one is the Target level of risk, and this indicates the amount of risk a driver is willing to tolerate to maximize their benefits in driving (Wilde, 2001). For example, how fast a driver is willing to drive to get earlier to his/her destination. The next element is the perceived level of risk, also called subjective risk, which is based on a driver’s experience with traffic in the road environment, vehicle-handling skills and decision-making skills. These three components differ from one individual to another, and therefore will also the perceived level of risk varies from one to another (Wilde, 2001, s.40). Persons who have low vehicle-handling skills will have a higher perceived level of risk. Wilde explains RHT in the following way; “People alter their behaviour in response to the implantation of health and safety measures, but the riskiness of the way they behave will not change unless those measures are capable of motivating people to alter the amount of risk they are willing to incur.”(Wilde, 2001. p.6).

Figure 1 shows how Wilde compared the target level of risk and perceived level of risk through a comparator. After calculating the amount of risk the driver is willing to take, then he/she changes their driving behavior from desired adjustment to adjustment action. The process itself is what RHT is all about, not the outcome of an event (Wilde, 2001. p.34).

Research in Finland studied drivers’ speed and gap distance compared to studded or without studded winter tires. This study concluded that drivers without studded winter tires drove significantly more carefully and slowly than drivers with studded tires (Wilde, 2001. p.119). According to this study, the subject risk is higher when a driver drives without studded tires during winter. Similar research in Iceland supports the Finnish research by specifying that the behavioral changes during driving are more important than the equipment of the vehicle (Wilde, 2001. p.120). This is yet another point which supports this theory towards our research where previous studies have used risk homeostasis as a fundamental theory to describe behavior changes while using studded- and studless tires.

Challenges when trying to establish risk compensation behavior
There are several challenges when trying to establish risk compensation behavior. James Hedlund (2000) describes three reasons for this. First of all, we don’t know which behavior is changed when using risk compensation as a theoretical base. Changes in behavior might occur in places that are not foreseen. For instance, when trying to establish if risk compensation occurs after cars are equipped with more safety equipment, we might look at speed or tailgating to establish whether risk compensation is present. However, it is almost impossible to foresee all the possible behavior changes after implementing a safety measure. For instance, drivers might be willing to accept longer commutes or are more willing to drive in severe weather conditions. In addition to this, there might be a delay between the implementation of the safety measure and the risk-taking behavior, which could for instance be the case in an educational setting. If military personnel are safeguarded from dangers during training, then they might be less able to act adequately when faced with actual combat. If we don’t know which behavior changes to measure, then we can’t use risk compensation as a theoretical framework. The second reason is that if we would know which behavior changes to measure, the changes might be subtle and difficult to operationalise. Speed and tailgating are measurable and objective measures, but more subtle changes like paying attention, are harder to measure. Lastly, even if we know which behavior changes to measure and how to accurately measure them, we cannot infer with certainty that these changes lead to more risk. An increase in driving speed might lead the drivers to be more alert which compensates for the increased risk when driving faster.

Method and Results

In order to answer our main research question we have to answer two sub-questions:

Sub question 1: ’To what extent does the use of studded tires impact the frequency of road accidents in winter conditions in Norway?’’.  

Sub question 2 “How does drivers’ behavior change when driving with studded tires compared to studless tires?”.

To answer sub-question 1, we must first find out how many accidents occur with studded and studless tires.

Table 1 In December 2022, January 2023, February and March 2023 (SSB, n.d.)

Figure 3 Number of accidents divided by wheel type in December 2022, January 2023, February and March 2023. (SSB, n.d.)

Using this information we can formulate and test two hypotheses:

H01: Proportion of drivers driving in winter conditions using studded tires are involved in the same proportion of accidents as drivers with studless (winter) tires.

HA1: Proportion of drivers driving in winter conditions using studded tires are involved in a lower proportion of accidents than drivers with studless winter tires.

To this this hypothesis we have performed a two-proportion Z-test. The results show that there is no significant difference between the proportion of accidents of drivers who use studded tires compared to drivers who use unstudded tires (Z=1.411 p=0.079). We therefore reject hypothesis HA1.

The second question we will have to answer is: Sub question 2) “How does drivers’ behavior change when driving with studded tires compared to studless tires?”. To answer this question we have conducted a survey consisting of N=283 participants. We asked the respondents the following questions related to their driving behaviour.

Figure 8  Results for the question: ‘’Would there be differences in your (driving pattern) when using studded tires?

Figure 8 describes the results for the question ‘’Would there be differences in your (driving pattern) when using studded tires compared to studless tires, especially on icy or snowy roads?’’ Most participants (54%) answered yes to this question. The error bars show that this is significantly more than the other categories. 74 participants (26%) answered that they don’t use studded tires. 57 participants (20%) answered that they do not feel safer when using studded tires.

Figure 9 Results for the question: ‘’Imagine driving with studded tires compared to studless tires in winter, on average you drive:’’ With error bars describing the p=0.05 error margin. N=283.

Figure 9 shows the answers to the question ‘’Imagine driving with studded tires compared to studless tires in winter, on average you drive:’’. As discussed in the limitations chapter, it is unclear what these results represent because the question is ambiguous. It can either be interpreted as ‘’When I drive with studded tires I drive’’ or ‘’When I drive with studless tires I drive:’’ After reflecting upon this question we expect that most participants answered the question from a point of view where they imagine themselves driving with studded tires in winter conditions. If we interpret the results with this assumption in mind, we can conclude that most people adhere to the speed limit. About an equal amount of people drive under and over the speed limit.

Figure 10 Results for the question: ‘’Imagine driving with studded tires compared to studless tires in winter, do you keep:’’ With error bars describing the p=0.05 error margin. N=283.

Figure 10 the descriptive statistics for the question ‘’Imagine driving with studded tires compared to studless tires in winter, do you keep:’’. This question has the same shortcomings as the question about adherence to the speed limit. After reflecting upon this shortcoming, we expect that people answer this question from the viewpoint of driving with studded tires in winter conditions. If we interpret the results with this assumption in mind, we can conclude that most people answer that they keep the three-second rule or keep more distance than the three-second rule when driving with studded tires in winter conditions.

Figure 11 Results for the question: ‘’On a scale from 1 to 5, where 1 indicates “Much Less Safe” and 5 indicates “Much Safer,” how safe do you perceive studded tires to be compared to studless tires in icy or snowy conditions?’’ With error bars describing the p=0

Figure 11 shows the answers to the question: ‘’On a scale from 1 to 5, where 1 indicates “Much Less Safe” and 5 indicates “Much Safer, “How safe do you perceive studded tires to be compared to studless tires in icy or snowy conditions?’’. We can conclude from the results that most people feel either safer or much safer when using studded tires. The non-overlapping error bars indicate that a significant amount of people feel either safer or much safer when using studded tires (N=283; P=0.05).

Discussion

In this chapter, we want to switch back the attention to our theory and answer the research questions we have presented before. Our main research question is ‘’How are the number of accidents influenced by the use of studded tires compared to studless tires and to what extent is this effect offset by the risk compensation behavior of the driver?’’ We will answer this question by answering the two sub-questions.

Sub Question 1

In this paragraph we will formulate an answer to question 1) “To what extent does the use of studded tires impact the frequency of road accidents in winter conditions in Norway?” We have written earlier that Scheibe (2002), finds that studded tires roughen up the ice and make better traction for studless tires. The fewer drivers using studded tires, the accident rate will rise. According to Statens vegvesen (Statistisk sentralbyrå, 2023), the use of studded tires does not impact the frequency of road accidents in winter roads in Norway compared to studless tires, and as we can find in the earlier studies by Elvik and Kaminska (2013) and King et al. (2022), the accident rate difference is near to non-existing today. Still, the accident rate for studless tires will rise if fewer and fewer drivers drive with studded tires because studded tires will not bring better traction than studless tires. Nevertheless, a Finnish study shows that studless drivers drive more carefully

Sub Question 2

In this chapter we will formulate an answer to question 2) “How does drivers’ behavior change when driving with studded tires compared to studless tires?”.

We can see the existence of behavioral change in a driver driving studded tires compared to studless tires. Studded tires are safety equipment which makes a driver safer during driving, and therefore they are willing to take a higher amount of risk. When a driver feels safer due to a better tire, he/she will also alter their behavior by increasing their risk to gain a better benefit. On the other hand, drivers driving with studless tires act more carefully on icy roads to avoid accidents. The handling skills and decision-making skills are different between those two categories, because the risk perception – based on cultural and psychological aspects, are different. The one has better equipment which leads to taking more risk, and the other acts more cautiously.

Psychological impacts of driver’s risk perception

What then would change a driver’s perception of the risk event they are facing? We know from RCT that decisional skills and vehicle handling skills are a big part of the individuals reasoning for adjusting their course of action, or put differently, how their behavior changes. For this adjustment action – as Wilde (2001) calls it, will be affected by the individual’s perception of their ability to handle the situation  (Wilde, 2001, p. 33). This relates to what Renn (2008) explains as personal involvement. The driver has something to gain from the situation, and therefore they to a higher degree accept the risks involved.

It is claimed that when people are involved and have personal control over the risk event, they have a more favourable perception of the risk they are exposed to (Renn, p. 109). This is what is mentioned as situation-related patterns of perception. Here is also voluntariness included as a second factor which increases the risk tolerance level, which is also highly applicable to our case, where everyone driving a car is doing so out of their own free will (Renn, 2008). This notion of control is very interesting in our case, as what Wilde (2001) describes as an important factor for risk perception is the vehicle handling skill. Just imagine a fictional person as a reference called Knut. Knut has been doing the same exercise for 20+ years and has never been involved in a crash or anything similar. This will – to a very high degree, create an illusion of being in a high degree of control over your vehicle. This will raise the bar for the risk tolerance that one has over the situation, which will result in a perception of being more in control than one is in reality. This of course is an assumption, as we cannot measure that amount of control, certainly not from our data. But from that assumption and using our collected data the driving behavior change when using studded tires – see Figure 8, we can explain that change in behavior is a result of the personal control which Renn speaks about. Also to refer to the other question in our survey – see Figure 11, people do feel safer when using studded tires in winter conditions, so we would draw the same assumption that because of personal control, people feel safer because they are in personal control of the vehicle, and therefore their driving behavior change in a certain manner.

Another factor is the familiarity of the risk source. “The known and familiar risk is much less fearsome than the unknown and less familiar risk” (Renn, p. 107). We know that driving a car is a very regular activity for most individuals in Norwegian society, and because of said risk familiarization, the risk tolerance regarding this activity is perceived as increasing. This means individuals driving a car, feel they have a higher degree of control because of their familiarity with the situation, and therefore they perceive the situation they are in as safer. Going back to our friend Knut, he’s been driving a car for 20+ years now, and for Renn (2008), driving of a car for Knut has a sense of naturalness over it. Anything that is perceived as natural for a person, is then perceived as familiar, which increases the risk tolerance, and therefore the naturalness of this activity will shape Knut’s risk perception (Renn, 2008, p. 109). According to RCT and Renn, our survey, depicted in Figure 8 and Figure 11, suggests that people alter their driving behavior with studded tires, feeling safer possibly due to the natural familiarity of the activity. 

We have in our theory chapter explained the principles of risk perception through a psychological approach. RCT tells us that people change their behavior in some sort after a risk-reducing implementation has been implemented. The risk-reducing implementation in our paper is using studded tires when driving in winter conditions. Because of the limitations discussed in our limitation chapter, we cannot make any statistical inference related the sub-questions 2A and 2B. We can however answer sub-question 2 by using data collected by other questions from the survey. The survey showed that a significant amount of people answered that they would change their behavior when driving with studded tires compared to studless tires. Besides this, the survey results showed that people feel safer when using studded tires compared to studless tires. In concordance with the RCT theory, we can make a case that drivers will take more risk and therefore (partly) offset the benefits of using studded tires.

Conclusion

Question 1

“To what extent does the use of studded tires impact the frequency of road accidents in winter conditions in Norway?”

Drivers using studded tires only have a very small advantage over drivers using studless tires. Research shows that braking distance and handling only improve by a small margin (Elvik et al., 2013). The use of studded tires does however increase road safety for all drivers because studded tires help break up the ice on the road. This causes roads to be less slippery overall, decreasing the accident rate for all drivers (Scheibe, 2002). We can therefore conclude that studded tires decrease the frequency of road accidents in winter conditions in Norway for all drivers combined, but not for individual drivers. The effect of using studded tires on the frequency of road accidents is mediated by the roughening of the ice on the road.

Question 2

“How does drivers’ behavior change when driving with studded tires compared to studless tires?”

Driver’s behavior is influenced by risk perception. For each activity, we accept a certain level of risk. When the activity becomes safer because we implement measures, in our case the use of studded tires, we are still willing to accept the same amount of risk. Therefore we are willing to compensate to increase the benefits of the activity. In our case driving faster and keeping less distance to the driver in front of us. Because drivers perceive studded tires as being safer than studless tires, drivers using studded tires take more risks. This has a negative effect on safety.

How are the number of accidents influenced by the use of studded tires compared to studless tires and to what extent is this effect offset by the risk compensation behavior of the driver?

We conclude that studded tires decrease the frequency of road accidents in winter conditions in Norway for all drivers combined, but not for individual drivers. The effect of using studded tires on the frequency of road accidents is mediated by the roughening of the ice on the road. This effect is at least partly offset by risk compensation behaviour. Future research will have to determine to what extent risk compensation affects offsetting the increase in safety of studded tire usage.

References:

Blaker, M. (2023, February 21). Advarer: Norge trenger flere bilister med piggdekk. Nettavisen. https://www.nettavisen.no/5-95-916352

Bruk av piggdekk—Lovdata. (2012). https://lovdata.no/artikkel/bruk_av_piggdekk/57?fbclid=IwAR1Ld7oa0XYSk341dSCvi3_8LlcUX0mc0ntT4mwIl5olrFduy0-11SEWUD4

Dyvik, E. (2023). Norway: Population by age group 2023. Statista. https://www.statista.com/statistics/586378/total-population-by-age-group-in-norway/

Elvik, R. (2023). Driver mileage and accident involvement: A synthesis of evidence. Accident Analysis & Prevention, 179, 106899. https://doi.org/10.1016/j.aap.2022.106899

Elvik, R., Fridstrøm, L., Kaminska, J., & Meyer, S. F. (2013). Effects on accidents of changes in the use of studded tyres in major cities in Norway: A long-term investigation. Accident Analysis & Prevention, 54, 15–25. https://doi.org/10.1016/j.aap.2013.02.004

Picture 1 “Piggdekk” By Erik Johansen. (https://www.faktisk.no/artikler/jdx4z/bilar-med-piggfrie-vinterdekk-er-oftare-involverte-i-trafikkulykker-med-personskade-enn-bilar-med-pi)

Hause, J. C. (2006). Offsetting Behavior and the Benefits of Safety Regulations. Economic Inquiry, 44(4), 689–698. https://doi.org/10.1093/ei/cbj041

Hedlund, J. (2000). Risky business: Safety regulations, risk compensation, and individual behavior. Injury Prevention, 6(2), 82. https://doi.org/10.1136/ip.6.2.82

Johansen, E. (2009) Piggdekk. (https://www.faktisk.no/artikler/jdx4z/bilar-med-piggfrie-vinterdekk-er-oftare-involverte-i-trafikkulykker-med-personskade-enn-bilar-med-pi)

King, D. R., Phillips, K. B., & Krauss, D. A. (2022). Knowledge of state-recommended following-distance rules. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 66(1), 878–882. https://doi.org/10.1177/1071181322661414

Lupton, D. (2013). Risk (2nd ed). Routledge.

Pei, X., Wong, S. C., & Sze, N. N. (2012). The roles of exposure and speed in road safety analysis. Accident Analysis & Prevention, 48, 464–471. https://doi.org/10.1016/j.aap.2012.03.005

Peltzman, S. (1975). The Effects of Automobile Safety Regulation. Journal of Political Economy, 83(4), 677–725.

Renn, O. (2008). Risk governance: Coping with uncertainty in a complex world. Earthscan.

Rosenlund, L., & Prieur, A. (2006). Det sociale rum og livsstilenes rum—Og La Distinction. In Pierre Bourdieu: En introduktion (p. 260). Reitzel.

Scheibe, R. R. (2002). AN OVERVIEW OF STUDDED AND STUDLESS TIRE October 2002 TRACTION AND SAFETY. U.S. Department of Transportation, Federal Highway Administration. https://wsdot.wa.gov/research/reports/fullreports/551.1.pdf

Sikre anonymitet i Nettskjema—Universitetet i Oslo. (n.d.). Retrieved 14 November 2023, from https://www.uio.no/tjenester/it/adm-app/nettskjema/hjelp/tiltak-for-a-sikre-anonymitet.html

SSB. (n.d.). 05247: Kjøretøy i trafikkulykker med personskade, etter statistikkvariabel, kjøretøytype, dekktype, ulykkestype og måned. Statistikkbanken. SSB. Retrieved 14 November 2023, from https://www.ssb.no/system/

Tanishita, M., & van Wee, B. (2017). Impact of vehicle speeds and changes in mean speeds on per vehicle-kilometer traffic accident rates in Japan. IATSS Research, 41(3), 107–112. https://doi.org/10.1016/j.iatssr.2016.09.003

Wikse, K. A. (2022). Vinterdekk. In Store norske leksikon. https://snl.no/vinterdekk

Wilde, G. J. S. (1998). Risk homeostasis theory: An overview. Injury Prevention, 4(2), 89–91. https://doi.org/10.1136/ip.4.2.89

Wilde, G. J. S. (2001). Target risk 2: A new psychology of safety and health; what works? What doesn’t? And why (2. (rev.) ed). PDE Publications.