ROAD TRAFFIC ACCIDENTS, RISK ATTITUDES AND BEHAVIOUR

Road traffic accidents, risk attitudes and behaviour among adolescents
Assoc. Professor Stig. H. Jørgensen
Dept. of Geography, Norwegian University of Science and Technology,
Trondheim, Norway
The paper aims to investigate aspects of risk-taking behaviour and motor vehicle accidents
among adolescents in a geographical setting. This project is funded by ‘The Risk and Safety
in Transport Programme (RISIT) in the Norwegian Research Council.
Recent trends in the accident pattern for motorised people aged 16 – 24 year are explored in
relation to other age groups in Norway. The total number of injured or killed motorists (16-24
years) in the period 1998 – 2002 was about 15200 based on the casualty’s place of residence.
The analyses are based on nationwide police reported road traffic accident data. Some data
accuracy issues in the data set are discussed. There are higher rates for killed and injured
motorized road-users aged 16-24 years living in rural and peri-urban areas than in urban areas
in the period: For killed and seriously injured road-users, this rural urban gradient is even
more pronounced. These profiles are more distinct for males than for females. Risk taking
behaviour as driving without protection (seat belt, helmet) is relatively more widespread
among young males living in rural and peri-urban areas.
A questionnaire study including self-reported accidents and ‘near accident’ situations from a
sample, in all about 600 youths (19 – 24 year) in 2 rural, 3 peri-urban and 2 urban
municipalities, is incorporated to achieve a broader risk-taking behaviour perspective. Various
self-reported accident experiences (incl. ‘near misses’) are slightly higher for young people
living in rural areas (not significant for all types of accidents). Rural adolescents score
significantly higher on some indicators expressing risk attitude and risk behaviour.
Road safety policy consequences can apply an ‘engineering’ approach versus a ‘behavioural’
approach. These approaches offer different means to reduce the accident levels. Engineering
countermeasures consist of technical and physical solutions involving the vehicles (‘alcohollock’) and the road system (road quality improvements. use of crash barriers separating
opposite carriageways). Behavioural road safety efforts imply a change in individual attitudes
and behaviour regarding drinking and driving, non-use of seatbelt, speeding, and so on.
Furthermore efforts involve influencing collective behaviour which might be embedded in the
existence of different safety cultures and group values influencing careless driving. The paper
deals with some geographical differences both in terms of ‘engineering’ and ‘behavioural’
safety strategies.
Several municipalities are operationalising and launching ‘Vision Zero’ plans with different
targets such as emphasising accident prevention (local safety campaigns) versus reducing the
consequences of crashes (physical countermeasures, posted speed limits). In general,
adolescents (19 – 24 year) are quite positive regarding various road safety countermeasures
and strategies inherent in the ‘Vision’. For countermeasures involving reduced speed and
speed limits, the adolescent’s degree of acceptance is lower, with rural adolescents expressing
the strongest negative views. Risk attitudes and trade-off between safety, freedom, boldness
and sensation seeking, are discussed. The complexity of accident patterns may call for ‘tailormade’ road safety policies.
© Association for European Transport 2004
Road traffic accidents, risk
attitudes and behaviour among
adolescents
Associate Professor Stig H. Jorgensen
Department of Geography,
Norwegian University of Science and
Technology (NTNU)
Presentation overview
1. Discuss some geographical perspectives on risk factors and
adolescent culture in motor vehicle accidents.
2. Present patterns of accidents, risk behaviour and attitudes;
focus on adolescents in a rural-urban context in Norway.
3. Give some statements on consequences for road safety policy.
Risk approaches in a geographical perspective
• The risk concept
• Driving as: routine everyday behaviour versus risk-taking activity
• Vehicle crashes due to lapses, driving errors and violent behaviour
• Risk minimization versus optimization
• Balancing losses (costs) and gains (benefits) from risk taking
• Risk-taking as individual and collective phenomena
• Risk-taking adolescent cultures and geographical variations
• Environmental risk / system risk
• Interaction between road-users and the physical and social environment
Risk-taking behaviour and contextual
explanations in geography
• Is there a geography of risk taking behaviour?
• Investigating the social and spatial patterning of risk
– Compositional effects (by age, gender, exposure level)
– Industrial structure, settlement pattern, employment and leisure
opportunities
– Institutional structures, local authorities’ engagement in road safety
– Contemporary social processes of non-uniform development might have
an impact on the reproduction of local risk cultures
• Urban-rural differences: physical and social environments
which have bearings on risk-related behaviour
Some methodological problems
Presentation:
area where the accident occurred or area of residence.
The risk of traffic accidents is influenced both by:
- the road environment at the scene
- driving behaviour ( socio-economic background, cultural norms)
influenced by place of residence.
Insider versus outsider
- separating local versus non-local casualties in a study area
Level of exposure:
- the population size is a crude surrogate measure
Traffic accident data
• Police-recorded road traffic accidents 1998-2002.
(Norwegian Public Roads Administration)
• Selection: motorized road users: 16-24 years moped, MC, automobile
Norway:
N=15 208 casualties
Selected study area (7 municipalities): N=633 casualties
• Population-based approach: the casualty’s place of residence
• Degree of coverage for residential municipality
- Norway: 76.1%
- study area: 88.5% (estimated)
• Missing percentages for:
– adequate protection: 39%
• Possible systematic bias: stronger underreporting in rural municipalities
Grouping municipalities into area types
•
Classification of 434 residential municipalities for casualties
in Norway and study area by
• Total population size of the biggest town in the municipality
Rural < 5000 population
Peri-urban 5000 – 15 000 population
Urban > 15 000 population
• The proportion living in densely populated areas
Rural
No of
municip
NORWAY
STUDY AREA
Peri-urban
Total
16 -24 years
population
No of
municip
Total
16 -24 years
population
Urban
No of
municip
Total
16 -24 years
population
322
152 404
67
110 256
35
229 021
2
924
3
3 496
2
20 957
Figure 1. Killed and all degrees of injuries in a 5-year period for males and females
motorized road-users per 1000 people, by type of area. 1998-2002. Norway.
Population-based rates. Source: Police-recorded data.
30
Rural
20
Peri-urban
Urban
10
Males (N = 20 767)
Females (N = 13 890)
75+
65-74
55-64
45-54
35-44
25-34
16-24
75+
65-74
55-64
45-54
35-44
25-34
0
16-24
Rate per 1000 for 5-year period
40
Figure 2. Killed and seriously injured in a 5-year period for males and females
motorized road-users per 1000 people, by type of area. 1998-2002. Norway.
Population-based rates. Source: Police-recorded data.
7
5
4
Rural
Peri-urban
3
Urban
2
1
Males (N = 2 687)
Females (N = 1 197)
75+
65-74
55-64
45-54
35-44
25-34
16-24
75+
65-74
55-64
45-54
35-44
25-34
0
16-24
Rate per 1000 for 5-year period
6
Figure 3. Killed and all degrees of injuries in a 5 years period for males and females
16-24 years, motorized road users per 1000 people by type of area. 1998-2002.
Norway and study area. Population-based rates. Source: Police-recorded data..
50
Norway Rural
40
Study area Rural
30
Norway Peri-urban
Study area Peri-urban
20
Norway Urban
0
Study area Urban
N = 15
10
N = 20
Rate per 1000 for 5-year period
60
Rural Peri-urban Urban
Rural Peri-urban Urban
Males
Females
16
14
12
Rural
10
Peri-urban
8
Urban
6
4
2
Males (N = 2 104)
Females (N = 404)
75+
65-74
55-64
45-54
35-44
25-34
16-24
75+
65-74
55-64
45-54
35-44
25-34
0
16-24
Involving suspection of alcohol (%)
Figure 4. Percentage males and females motorized road-users in accidents
(all degrees of injuries) involving suspicion of alcohol by type of area. 1998-2002.
Norway. Population-based. Source: Police-recorded data..
Figure 5. Percentage males and females motorized road-users in accidents
(all degrees of injuries) with no protection (seat belt, helmet) by type of area.
1998-2002. Norway. Population-based. Source: Police-recorded data.
.
18
N = 37
14
12
10
Rural
4
Peri-urban
N=6
6
N=6
N=6
8
Urban
2
Males (N = 1 995)
Females (N = 882)
75+
65-74
55-64
45-54
35-44
25-34
16-24
75+
65-74
55-64
45-54
35-44
25-34
0
16-24
No protection (%)
16
Figure 6. Percentage motorized road-users 16-24 years in accidents (all degrees of injuries
with no protection (seat belt, helmet) by type of area. 1998-2002.
Norway and study area. Population-based. Source: Police-recorded data.
16
No protection (%)
14
N = 587
N = 352
12
N=4
10
N = 11
N = 420
N = 38
8
6
4
2
0
Norway
Study area
Rural
Norway
Study area
Peri-urban
Norway
Study area
Urban
Study area
Survey data on accident experiences, risk attitudes and behaviour
• Random sample 19-24 years in 7 municipalities:
Rural (2) - Peri-urban (3) - Urban (2)
• Response rates:
40% in rural areas, 48% in peri-urban areas, 45% in urban areas.
Selection:
Adolescents traced back to where they lived in the age span 16-18 years
(important socialisation period) to avoid exposure effects of youths moving to
other types of environment (N= 607)
Data quality:
• Missing values for the variables vary from 0 to 3 %.
• Chance of recall bias, under- or overstating attitudes, behaviour
Figure 7. Type of self-reported accident experiences involving motor vehicles
for males and females 19-24 years by type of area. 2004. Study area. Percentages.
Source: Survey data 2004
60
Motor vehicle
accident with
injuries
50
40
= 2 "near miss"
N=7
N = 14
10
N=3
N = 16
20
N = 10
30
N = 10
Type of personal accident experience (%)
70
0
Rural
Peri-urban
Urban
Males (N tot = 272)
Rural
Peri-urban
Urban
Females (N tot = 314)
> 2 accidents
sum: injuries,
material or "near
miss"
Table 1. Mean score on Likert scales (5 to 7 points) for risk attitudes and
risk behaviour for 19–24 years by type of area. 2004. Study area.
Risk attitude:
Type of
area 1
Rural
Peri-urban
Urban
Total
Risk behaviour:
Accept of violation of
Occurrence of speed violation.
traffic rules and risk taking 1-5, where 1 is never
and 5 is very often
1-7, where 7 is totally agree
4.1
3.8
3.5
3.8
1
(71) *
(324) *
(195) *
(590)
2.9
2.6
2.5
2.6
(60) *
(289) *
(154) *
(503)
Based on the municipalities where the adolescents were living at the age of 16-18 years.
* significant (0.05)
Source: Survey data.
Risk behaviour:
Non-use of seat belt
1–5, where 1 is never
and 5 is very
often
2.0
1.5
1.2
1.5
(60) *
(290) *
(155) *
(505)
Figure 8. Mean scores on Likert scales (7 points) for acceptance of various safety measures *
for males and females 19-24 years by type of area. Study area. 2004.
Source: Survey data. * Related to Vision Zero
Very
7
negative
6
5
Rural
males
"Neutral" 4
Urban
males
Rural
females
2
Urban
females
1
Im
pr
ov
ed
ro
Cr
ad
as
ST
s
h
AT
b
a
SP
IO
rr
ie
EE
N.
rs
SP
D
BU
EE
M
D
PS
CA
M
M
ER
AX
Al
AS
co
SP
- lo
EE
ck
D
M
LO ers
or
C
e
KE
v
i
M
s
RS
i
or
bl
¤P
e
e
¤L
en
po
po
O
al
l
lic
i
W
ce
ty
e
E
po
co
R
SP
nt
in
ro
ts
EE
ls
on
D
dr
LI
Ad
M
iv
HI
ol
IT
er
G
es
S
´
H
s
ce
ER
l
i
ce
nt
FI
ns
a
NE
¤
cc
e
Ex
id
L
EV
en
te
nd
tc
EL
ed
am
S
dr
pa
iv
ig
in
ns
g
tra
in
in
g
Very
positive
3
Some limitations of the study
•
Rates by adolescents in the area of residence are critical (students)
•
Split by sex, age groups, give small numbers (non-significance)
•
Population-based approaches underscore physical geographical factors,
do not account for outsiders/through fare traffic in areas.
•
Lack of appropriate exposure data (police-reported accidents)
•
Differences in underreporting of traffic violations by areas might influence
proportions of risk-taking behaviour.
•
Omitted variables:
- information on posted speed limits and actual speed
- indicators on reckless driving
•
Elements of ‘natural variation’ exist:
composition of population, area size, non-built up areas, topography, road conditions.
•
Internal variations in crash patterns by type of municipalities
• Prevention strategies
• Geographical-physical perspectives and measures
versus
• Behavioural/socio-cultural perspectives and measures
• General risk-taking culture and risk behaviour do not predict
specific accident patterns.
• Legislation
• Enforcement
• Education/behavioural change
• Vision Zero for traffic accidents in Scandinavia (0 killed in year 2030)
Vision Zero Programme
• Focusing on serious and fatal accidents
(30% reduction within 2012 in Norway).
• Improvement of road safety environment (system risk).
- accident prevention.
- reduce consequences of initial driver error.
• Stronger acknowledgement of the importance of speed levels.
• The scope for technical vehicle control systems.
• The possibilities for reducing risk-taking behaviour.
• Need for stronger enforcement.
Towards implementing Vision Zero strategies
• Specific target groups and geographical areas have not been in focus
Road system improvement
• More difficult to reduce system risk in rural areas.
• Cost-efficiency and effect of engineering countermeasures.
Control and enforcement
• Policing and control are resource demanding in vast rural areas.
• Technical controls and surveillance: speed cameras, alco-lockers, maximum speed
control of vehicles, are tested.
Behavioural changes
• The need for reducing risk attitudes, violent behaviour and local risk cultures.
• Driver’s training, safety cultures, local campaigns, involvement and enthusiasm.
• Trade-off between the acceptance of accident risk mitigation and
risk related to freedom and need for mobility embedded in society.
Preliminary conclusions
• A rural–urban gradient exists for risk-taking behaviour for motorized
adolescents.
• Higher accident rates and self-reported risk attitudes and behaviour among
the adolescents living in rural areas.
These patterns do not disappear when controlling for sex, finer age groups
and exposure level.
• Males in rural areas experience a higher risk level than females.
• Risk-taking behaviour interplay with social and physical environments
and generate a complex “double ecology”.
• The social environment is expected to influence behaviour strongest.
The accident risk is also influenced by conditions embedded
in the physical environment where they are exposed.
These patterns of variation should be further disentangled.
Preliminary conclusions (cont.)
• The Vision Zero offers a good starting point for reducing the toll on the road.
The Vision and the countermeasures have relatively moderate support among
adolescents (variations in the rural and urban youths’ acceptance).
• Safety measures seem to be more difficult to implement in rural areas
due to lower efficiency of controls, stronger incentives for speeding, a greater
variety in causal factors for accidents (“unexplainable / random accidents”)
• There are some clues for prevention strategies focusing on strengthening
bottom-up road safety campaigns in rural areas.
• Place-specific efforts have to challenge and change rooted cultures based on
the automobile as a symbol of freedom, the need for mobility and adolescent
bravado and status linked to risk-taking in rural areas.