Risky traffic behaviour among young adolescents

Adolescence is the period between childhood and adulthood. Its beginning corresponds with the beginning of puberty. This starts with hormonal changes that eventually result in sexual maturity and adult sexual characteristics. For girls, puberty starts a little earlier (approximately at the age of ten) than for boys (around the age of twelve). It applies for both genders that puberty begins at an increasingly younger age compared to the past.

Whereas the beginning of adolescence is thus determined by physiological changes, until recently, the end of adolescence was mainly determined by socio-cultural characteristics, the most important being 'adopting adult roles'. This concerned, for example, issues such as the choice of a companion in life, getting children, and earning a living. Just as in other western countries, young people in the Netherlands adopt these roles at an increasingly later age. Therefore, and as a result of the increasingly early start of puberty, adolescence would become longer and longer. However, recently the social criterion has been complemented with a physiological one: the adult structure of the brain. Some decades ago it was assumed that, in terms of structure, the human brain was fully developed around the age of five. Recent brain research, however, has shown that the brain, in adolescence in particular, undergoes large changes which start at the beginning of puberty. The process in which the changes take place only finishes when people are in their mid-twenties. Around the age of 24, the brain achieves its final adult structure. On the basis of these physiological criteria we therefore consider young people between 10 and 24 years old as adolescents.

As said, the first question in this fact sheet is whether young adolescents will also behave more risky in road use because of their physical and mental development. To answer this question, we have looked whether cause-of-death statistics give reason to such an assumption. A relevant distinction is death due to illness (natural death) or death due to injuries (unnatural death). The essential difference between the two categories is that illness has mainly an internal cause, whereas injuries are the result of external causes. Within the category of 'unnatural death' we mainly focus on unintentional injuries: crashes, and we assume that the risk of crashes partly corresponds with risky behaviour.  

Figure 1 shows the average number of fatalities among adolescents per 100,000 (mortality rate) in terms of age group and cause of death. To keep the graph easy to read, the real value of the mortality rate as a result of illness among children of 0 to 4 years of age has been reduced by a factor of 10. It shows from the graph that young adolescents have a very low mortality rate in comparison with young children and adults. This is mainly the result of death by illness being lower than in any other age group. Once they have survived the high mortality risk as a result of illness during their childhood, adolescents are then the healthiest members of Dutch society. However, this favourable picture is contradicted by the high increase in the category of 'unnatural' death (suicide, murder, road crashes and other accidents), and then mainly because of the increase in the category of road crashes.

Figure 1. Fatalities due to illness and unnatural causes per 100,000 residents in the period 1999-2008. With a view to readability, the figure for mortality due to illness among children of 0 to 4 years of age has been reduced with a factor of 10 (CBS cause-of-death statistics).

To provide some insight into the backgrounds of this rise, Figure 2 presents the development in fatalities due to unnatural causes in terms of gender. This shows that a unnatural death occurs particularly among men and to a far lesser extent among women. This also applies for road casualties. This difference begins to become visible only at the start of adolescence (10‑14 years old) for that matter.

It shows from a further distinction in terms of mode of transport that of the road fatalities among the 10 to 17 years old, 26% die as car passengers, (see also archived SWOV Fact sheet Young drivers and their young passengers, 27% as moped riders (see also SWOV Fact sheet Moped and light-moped riders) 35% as cyclists (see also SWOV Fact sheet Cyclists) and 8% as pedestrians.

Figure 2. Unnatural death per 100,000 residents in terms of cause, age and gender in the period 1999-2008.

Naturally, the next question in this fact sheet is how the increase in casualties and its gender differences can be explained. The three most important explanations that play a role in road safety are: a) an increase in 'deliberate' risky behaviour, b) an increase in the exposure to hazards, for instance, because more kilometres are travelled or a 'more dangerous' mode of transport is used, or c) greater lack of experience with new traffic situations and modes of transport. The remainder of this fact sheet summarizes the knowledge about young adolescents for each of these explanations.

Not much research has been conducted into the frequency of risky behaviour as children and adolescents become older. It is therefore not known whether risky behaviour increases after childhood and whether it decreases during developing adulthood. It is known that health-related risky behaviour (such as alcolhol consumption and smoking) highly increases during early adolescence (Van Dorsselaer et al., 2007; Monshouwer et al., 2004). With respect to risky road usage behaviour, more information is available for this age category about the differences between males and females. In questionnaires young men state more often than young women that they exhibit risky behaviour in traffic (Feenstra, Ruiter & Kok, 2010, Twisk et al., forthcoming). Peers also have an effect on risky behaviour. As it becomes increasingly important for them to belong to a group, adolescents tend to behave more hazardlously in the company of peers than when they are on their own (Crone, 2008; Nelis & Van Sark, 2010). Hardly any research has been conducted of the role of peer pressure in road user behaviour in the Netherlands. Research in the US among 16-17-year old car drivers shows that they behaved more risky when peers were passengers than when they were on their own. Peer pressure is not the only explanation for this behaviour. An alternative explanation is the increase of distraction due to the presence of passengers. See for the description of these studies archived SWOV Fact sheet Young drivers and their young passengers about this subject.

Although we actually have little insight into the exact nature and amount of deliberate risky road user behaviour of young adolescents, we do have knowledge about some of its contributing factors.  These factors are: brain development, perceived invulnerability, social environment, impulsiveness, behavioural intentions and behaviour routines. These factors do not each determine an independent phenomenon, but address the risky behaviour from various perspectives.

Brain development

On account of recent insights into structural changes in the brain of young adolescents (Giedd, 2008), neuro-psychological research has focussed on the issue whether the rise in  risky behaviour might be a result of this. Indeed, research showed that the areas of the brain that are responsible for impulse control, planning and integration of information (the control system) are still very much being developed during early adolescence. At the same time, the parts of the brain that are sensitive to stimuli, rewards, new experiences and the satisfaction of needs (the limbic system) are stimulated by the puberty hormones. The early increase of risky behaviour might be the result of this a-synchronic development, whereas its later decrease is the result of the increasing adult structure of the control system. Because of this, the activation from the limbic system is better regulated (Steinberg, 2008). At approximately 24 years of age, the brain reaches its ultimate adult structure. Also the difference in risky behaviour between boys and girls is occasionally traced back to this a-synchronic development. Due to the effect of the male sex hormone (testosteron) on the limbic system, the beginning of puberty is more intense, whereas the development of the control system is quite slower with boys than with girls (see Vlakveld, 2011 for an overview of the literature about this).

Aside from this negative effect on the safety of young adolescents, the structural changes and the greater plasticity of the brain have in fact a positive effect on the learning skills of young adolescents. This way, young adolescents, far more so than adults, can learn new skills, feel more challenged and can therefore reach levels of excellence more easily (Keating, 2007). This might also apply to learning the necessary skills for safe road user behaviour.

Perceived invulnerability

Other explanations for risky behaviour focus less on the physiological and more on the psychological factors, such as lack of knowledge, 'dangerous' notions, and (too) much confidence in one's own skills. An important factor in this context is estimating the risk that a hazard may occur, and the risk that it may happen to you. It is often assumed that young adolescents underestimate both and that, as a result, they behave more risky than adults. This would thus imply that interventions should particularly be directed towards correcting these 'wrong' notions. However, when adolescents are asked about the risk of dying because of risky behaviour, their estimates turn out to be unrealistically high. Considered against this background, providing the 'correct figures' or providing 'hard realistic' information would only be counter-productive (Reyna & Farley, 2006). See also SWOV Fact sheet Fear appeals and confronting information campaigns.

Social environment

During adolescence the social environment becomes more and more important. Whereas during childhood parental behaviour modelling is most effective, during adolescence the experiences with friends seem to guide behaviour in particular. This impression actually needs some correction, because although parents often have the impression that 'they are no longer listened to', research shows that they have more influence during adolescence than they are aware of (Bokhorst, Sumter & Westenberg, 2010; Nelis & Van Sark, 2010)

It is this interaction between adolescents and their environment that protects or, on the contrary, exposes them to hazards, not only in road usage, but also in other areas, such as health (unsafe sex or smoking and alcohol consumption; Jessor, 1992). In the Netherlands, indications have been found for multi-problem behaviour among adolescents. Junger, for example, found a connection between road crashes and criminality (Junger, Mesman & Meeus, 2003; Junger, Terlouw & Van der Heijden, 1994; Nieuwenhuijzen et al., 2009) and research by Wurst (2002) showed that a small part of the Dutch adolescents, often young men, behave risky in many areas. This implies that studies and interventions for the purpose of explaining and preventing risky behaviour should not only focus on isolated problem behaviour, but especially focus on the interaction between personal development and social environment.

Behavioural intention, impulsiveness and routine

People consider themselves as rational beings showing 'systematic' behaviour: we decide on something and we then carry it out. For this reason, in many behavioural theories behavioural intention is considered as the most important 'predictor' of behaviour. Indeed, many studies show that changes in behavioural intention lead to changes in actual behaviour, but also that this correlation is less strong than would be expected on the basis of the theory. We often have an intention, and sometimes we also carry it out. This happens, among other things, because 1) a great many of this kind of decisions develop less consciously than was earlier assumed, 2) a large part of the actions are more or less automatic (habits), and 3) the control over this behaviour is limited (Dijksterhuis, 2008; Webb & Sheeran, 2006). In fact, our behaviour is driven by two different systems: an intuitive system in which decisions are made quickly and automatically and a control system in which decisions develop systematically, consciously, but slowly (Kahneman, 2003).

The correlation between behavioural intention and actual behaviour is probably even weaker with adolescents than it is with adults. Firstly, this is due to the greater impulsiveness of adolescents (acting, not thinking) and, secondly, because their behaviour, to begin with, is less automatic and intuitive due to lack of experience and routine. Adolescents still consider things consciously, whereas adults practically immediately know without thinking what is the correct decision for them, based on earlier experiences. For instance, it shows from research in which adolescents and adults had to judge as quickly as possible whether certain behaviour was wise - for example, 'swimming with sharks' or 'setting your hair on fire' - that adolescents, more so than adults, have a longer reaction time and that their brain is more active before they judge this behaviour as unwise (Reyna & Farley, 2006). Moreover, with adolescents the outcome of the decision is more often situation-related than it is with adults. Examples are: adolescents that, once in the pub, cannot resist drinks offered, even though at home they had the intention not to drink alcohol, and adolescents that in a state of excitement more often make more risky decisions.

Although it is difficult to predict behaviour from behavioural intentions, attitude and knowledge, they do offer insight into differences between male and female. Thus, research shows that boys, more often than girls, find it less important, for instance, to stick to the rules, do not mind it so much to drink and drive and feel less responsible for others' safety (Feenstra, Ruiter & Kok, 2010; Twisk et al., forthcoming).

In addition to explanations in terms of physiological and psychological development, increased exposure to hazards and lack of experience are also explanations for the higher road user mortality among young adolescents. At this stage in life, adolescents travel increasingly large distances as road users and they can make use of new modes of transport. For instance, they are permitted to ride a moped at the age of 16, once they have passed their moped driving licence. A few figures as an illustration. During secondary school age, the number of yearly cycling kilometres increases up to four times as many as during primary school age. Secondary school pupils travel almost as many kilometres by bike as by car. This increase in mobility highly contributes to the increase in the number of cycling casualties. When correcting for the exposure and calculating the number of fatalities per kilometre travelled, we see that the risks are particularly high for moped riders and that the risk for young cyclists remains remarkably similar between childhood and adolescence. Neither do we see differences between boys and girls among cyclists (Twisk et al., 2013a). This could indicate that experience (with cycling in this case) acts as a protection against the consequences of the previously mentioned increase in risky behaviour. Another possible explanation is the low speed of cycling. In contrast with riding a moped or being behind the steering wheel of a car, adolescents on bicycles do not ride noticeably faster than adults. Although the above-mentioned mobility is mainly seen as exposure to hazards, this is only one side of the picture. The positive side is that adolescents gain experience by more fully participating in road usage. As a result they become increasingly more skillful in dealing with complex traffic situations: practice makes perfect. In the Netherlands, secondary school pupils have already achieved a great many kilometres of cycling experience. For this reason, interventions rarely focus on skills, but mainly on 'motivation', that is to say, the 'willingness' to safe road user behaviour. That this approach is too one-sided shows from one of the few studies into competence levels. This research has studied the behaviour of young cyclists (11-13-year olds) close to trucks and the way in which they took into account the blind spot of the trucks. It turned out that the majority of the adolescents could indicate the blind spots correctly, but that the transposition of this knowledge into practical choices was very limited, especially in complex traffic situations (Twisk et al., 2013b). Boys actually did not differ from girls in this respect. This research illustrates that lack of experience can be an important component in risky behaviour of adolesents in complex traffic situations, and that, next to the skill of operating a mode of transport, especially the higher order skills, such as risk perception, are essential in this context. These higher order skills develop far more slowly than those for operating a vehicle. The lack of experience of this age group probably also plays an important role in the lack of safety of mopeds riders, in particular in respect to higher order skills (see also fact sheet Mopeds and light-mopeds riders).

Various intervention programmes are possible for the prevention of risky road usage. The above-mentioned characteristics of young adolescents have the following implications for such interventions:

• The simultaneous increase of the various categories of 'unnatural' fatalities among adolescents indicates that there is a more general problem than simply a problem of road usage. Interventions could focus on these general causes, for instance, the diminishing of parental supervision and the prevention of 'impulsive actions'.
• Before assuming that risky behaviour is deliberate, it should be excluded as a rule that lack of experience in traffic situations plays a role. Furthermore, programmes focussed on stimulating road user experience can - in theory - have a moderating effect on later risk during adolescence.
• Although the underlying causes are not yet fully clear, the statistics clearlly show that particularly boys are the victims of a unnatural death. Therefore, the development and implementation of interventions should mainly focus on the effects for this group.
• As pushing limits is essential in the development of the adolescent and a prerequisite for learning 'lessons for life', sufficient challenges should be available in the social and physical environment of adolescents so that they can explore their limits without running the risk of serious injuries.
• Parents still have greater influence on this age group than they are themselves aware of. Activating parents and providing them with strategies to protect their children, without unnecessarily hindering them, is a productive perspective for road user safety interventions.
• Peers play an important role in developing risky road user behaviour, and it requires courage and social skills to resist peer pressure. The issue is how young adolescents can be taught how to protect themselves against the negative consequences of peer pressure.
• Even more so than in the case of adults, behavioural intentions of young adolescents have limited value in predicting risky behaviour. Therefore, evaluations measuring the effect of intervention programmes of adolescents against the changes in behavioural intentions offer little certainty about the ultimate effects on behaviour.

In comparison with other age groups, young adolescents, and boys in particular, die of unnatural causes relatively often. To a large extent this comprises road crashes. It probably plays a role that young adolescents 'deliberately' exhibit more risky behaviour. Because of their physical and mental development, they are more drawn to risky challenges, more susceptible to peer pressure, and they have less self-control and overview than older adolescents for example. Besides, in this stage in life, young people travel increasingly large distances in road traffic and they also get access to new modes of transport, such as mopeds. On the other hand, young people naturally learn by experience by more fully participating in traffic. To begin with, they develop their skills in operating a vehicle; essential higher order skills, such as risk perception, develop far more slowly. These backgrounds to risky behaviour of young adolescents naturally have implications for intervention programmes focussed on limiting the consequences of this road user behaviour. 

Bokhorst, C.L., Sumter, S.R. & Westenberg, P.M. (2010). Social support from parents, friends, classmates, and teachers in children and adolescents aged 9 to 18 years: Who is perceived as most supportive? In: Social Development, vol. 19, p. 417-426.

Crone, E. (2008). Het puberende brein; Over de ontwikkeling van de hersenen in de unieke periode van de adolescentie. Bert Bakker, Amsterdam.

Dijksterhuis, A.P. (2008). Het slimme onbewuste; denken met gevoel. Bert Bakker, Amsterdam.

Dorsselaer, S. van, Zeijl, E., Eeckhout, S. van den, Bogt, T. ter, et al. (2007). HBSC 2005 Gezondheid en welzijn van jongeren in Nederland. Trimbos-instituut, Utrecht.

Feenstra, H., Ruiter, R. & Kok, G. (2010). Social-cognitive correlates of risky adolescent cycling behavior. In: BMC Public Health, vol. 10, nr. 1, p. 408.

Giedd, J.N. (2008). The Teen Brain: Insights from neuroimaging. In: Journal of Adolescent Health, vol. 42, nr. 4, p. 335-343.

Jessor, R. (1992). Risk behavior in adolescence: A psychosocial framework for understanding and action. In: Developmental Review, vol. 12, nr. 4, p. 374-390.

Junger, M., Mesman, J. & Meeus, W. (2003). Psychosociale problemen bij adolescenten, prevalentie, risicofactoren en preventie. Zicht op jeugd. Van Gorcum, Assen.

Junger, M., Terlouw, G.-J. & Heijden, P.G.M. van der (1994). Crime and accident involvement in young road users. In: Behavioural research in road safety V. Proceedings of a seminar 6-7 September 1994, Nottingham University.

Kahneman, D. (2003). Maps of bounded rationality: A perspective on intuitive judgment and choice. In: Frangsmyr, T. (red.), Les Prix Nobel: The Nobel Prizes 2002 The Nobel Foundation, Stockholm, p. 449-489.

Keating, D.P. (2007). Understanding adolescent development: Implications for driving safety. In: Journal of Safety Research, vol. 38, nr. 2, p. 147-157.

Monshouwer, K., Dorsselaer, S. van, Gorter, A.F., Verdurmen, J.E.E., et al. (2004). Jeugd en riskant gedrag; kerngegevens uit het peilstationsonderzoek 2003. Trimbos-instituut, Utrecht.

Nelis, H. & Sark, Y. van (2010). Puberbrein binnenste buiten: wat beweegt jongeren van 10 tot 25 jaar? Kosmos Uitgevers B.V., Utrecht/Antwerpen.

Nieuwenhuijzen, M., Junger, M., Klein Velderman, M., Wiefferink, K., et al. (2009). Clustering of health-compromising behavior and delinquency in adolescents and adults in the Dutch population. In: Preventive Medicine vol. 48, p. 572-578.

Reyna, V.F. & Farley, F. (2006). Risk and rationality in adolescent decision making: implications for theory, practice and public policy. In: Psychological science in the public interest, A Supplement to Psychological Science vol. 7, nr. 1, p. 1-44.

Steinberg, L. (2008). A social neuroscience perspective on adolescent risk-taking. In: Developmental Review, vol. 28, nr. 1, p. 78-106.

Twisk, D., Bos, N., Shope, J.T. & Kok, G. (2013a). Changing mobility patterns and road mortality among pre-license teens in a late licensing country: an epidemiological study. In: BMC Public Health, vol. 13, nr. 333.

Twisk, D.A.M. & Vlakveld, W.P. (2010). Recent findings on risky acts in adolescence: implications for understanding European drivers. In: Dorn, L. (ed.), Driver behaviour and Training, Volume IV; Human factors in road and rail transport. Ashgate, Aldershot UK, p. 7-21.

Twisk, D., Vlakveld, W., Commandeur, J., Shope, J.T., et al. (forthcoming). Towards the development of a survey method: Predictors of risky behaviour and crashes among young adolescent cyclists. In. 

Twisk, D., Vlakveld, W., Mesken, J. & Shope, J.T., et al. (2013b). Inexperience and risky decisions of young adolescents in interactions with lorries, and the effects of competency versus awareness education. In: Accident Analysis & Prevention, vol. 55, p. 219-225.

Vlakveld, W.P. (2011). Hazard anticipation of young novice drivers; Assessing and enhancing the capabilities of young novice drivers to anticipate latent hazards in road and traffic situations. Thesis University of Groningen. SWOV Dissertation Series. SWOV Institute for Road Safety Research, Leidschendam.

Webb, T.L. & Sheeran, P. (2006). Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. In: Psychological Bulletin, vol. 132, nr. 2, p. 249-268.

Wurst, T. (2002). “Hardcore” problem groups among adolescents: their magnitude and nature, and the implications for road safety policies. R-2002-25. SWOV Institute for Road Safety Research, Leidschendam.