Safety Science 51 (2013) 165177

Contents lists available at SciVerse ScienceDirect

Safety Science
journal homepage: www.elsevier.com/locate/ssci

Selection of a relevant indicator Road casualty risk based on nal outcomes

Dragoslav Kukic a,, Krsto Lipovac b, Dalibor Peic b, Milan Vujanic b
a
b

Road Trafc Safety Agency of the Republic of Serbia, Belgrade, Serbia

University of Belgrade, Faculty of Transport and Trafc Engineering, Department of Trafc Safety, Belgrade, Serbia

a r t i c l e

i n f o

Article history:
Received 17 January 2012
Received in revised form 7 May 2012
Accepted 12 June 2012
Available online 27 July 2012
Keywords:
Road safety
Rates
Risks
Risk mapping
Municipalities
Linear correlation

a b s t r a c t
There are no dilemmas among the academics and experts whether it is important and necessary to
analyze the road casualty risk. The road casualty risk analysis is a very efcient way of ltering the most
dangerous sections, roads or specic territories. In previous analyses of road safety in Serbia, a value and
type of a specic risk according to the size of the observed area (state, region, district, municipality),
section length or the importance of a road category, were not explicitly determined. Differences in values
of the analyzed parameters could be expressed to such an extent that the acquired values of differences,
among some of the units that are being observed, represent range divided into risk bands. These differences are primarily the result of the severity of injuries and types of accidents used for calculating individual risk categories. In this paper, a model for selection of an acceptable risk in selected
municipalities in Serbia is presented. Here presented model will be used for future researches and nal
assessments of the state of road safety, i.e. for the reliable risk mapping of the Serbian municipalities. The
practical contribution of the risk analysis is in dening a reliable way of choosing acceptable nal outcomes rates for a dened unit of observation.
2012 Elsevier Ltd. All rights reserved.

1. Introduction
According to the data on the number and consequences of road
accidents, population, registered motor vehicles, road network
length, AADT value (Average Annual Daily Trafc) or kilometrage,
the values of so-called public, trafc, collective and dynamic
risks can be calculated. These are the most prominent relative indicators or nal outcomes rates of road safety in the scientic area of
Road safety. They are most often used for risk mapping and also as
the most important elements for describing the level of road safety
on the particular territories, roads or road sections.
Researches in which road safety levels and also safety risks
among countries have been compared (Koornstra et al., 2002;
Wegman et al., 2005 and Wegman et al., 2008), used several nal
outcomes, as well as several safety performance indicators (SPIs).
Namely, the following nal outcomes have been used:
 Distribution of fatalities per road transport mode (passenger
cars, commercial vehicles, motorcycles, bicycles, pedestrians,
etc).
 Fatalities per road users age groups.
 Fatalities per different road categories (highways, main urban
streets, rural roads, etc.).
Corresponding author. Address: Mihajlo Pupin Boulevard 2, 11000 Belgrade,
Serbia. Tel.: +381 113117928; cell: +381 648428028; fax.: +381 113117298.
E-mail address: kukicdragoslav@gmail.com (D. Kukic).
0925-7535/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ssci.2012.06.016

Wegman and Oppe (2010) state that comparing road safety

among countries is often conducted by using indicators rates that
take into account the number of fatalities, i.e. fatalities per population, and these indicators represent the so called public risk. Also
Wegman and Oppe (2010) are of opinion that dening a risk which
is based on the public risk has a disadvantage, as the degree of
motorization has not been taken into account, and suggest that
indicator obtained by fatalities and vehicle kilometers should be
often used as the indicator that gives better results of road safety
levels and risk assessments. This indicator is the so called dynamic
trafc risk and it takes into account the mobility of the population.
However, most countries do not register data about traveled kilometers. Therefore, indicator that is taking into account fatalities
and the number of registered vehicles is used instead, and indicator
is so called trafc risk.
Hermans et al. (2009) state that it is necessary to analyze available data in order to make a review of road safety. For the purposes
of ranking and comparing road safety among countries, data relating to accidents and consequences (fatalities, serious injuries and
slight injuries) per population could be used.
In road safety comparison of two countries (China and USA),
Zhang et al. (2010) used the trafc risk (the number of fatalities
relative to the number of motorized vehicles and relative to the
number of passenger cars), public risk (number of fatalities relative to the population) and the relation of the number of fatalities to
the gross national product (hereinafter referred to as GNP). Also for
the purposes of dening and comparing road safety levels between

166

D. Kukic et al. / Safety Science 51 (2013) 165177

countries, Zhang et al. (2010) have also determined and compared

the trends of certain analyzed indicators.
In his doctoral dissertation, Al-Haji (2007) has thoroughly analyzed the exposure (degree of motorization), trafc and public
risk (rates of nal outcomes such as number of fatalities divided by
the number of vehicles or divided by the number of citizens), as
well as trends in the degree of motorization, trafc and public risks
and the ability of assessing the state of road safety.
Applying absolute gures of fatalities or other road safety
outcomes and dividing them by a chosen exposure is the simplest
and most commonly used method for comparing road safety
performance of countries (Ekler, 2010). Exler produced risk maps
for Belgian municipalities based on Bayes relative risks. Focus is on
the nal outcomes fatalities per trafc and fatalities per trafc
trend.
The Road Assessment Programme RAP (EuroRAP, usRAP and
AusRAP) produces risk maps based on accident rates that combined effects of behavior, road and vehicle. RAP protocol focuses
on fatal and serious accidents. RAP models are generally used in
national targets and those that can have life-changing consequences. Consideration of fatal accidents alone would severely restrict the average accident frequency per site and make results
more variable (Hill, 2010). RAP models are concretely intended
for risk mapping of roads and road sections as observation units.
Nam and Song (2008) used Bayesian spatial modeling to estimate and map accident risk. This model analyses output indicators
(the number of fatalities, serious injuries and slight injuries) and
nal outcomes -rates. The Model is based on account for spatial

dependence in modeling and corresponding statistical inference.

The model is using a Global spatial autocorrelation. This is a global
measurement of spatial autocorrelation over the entire observations over an area of interest, used for testing spatial autocorrelation to detect departures from spatial randomness. Global spatial
autocorrelation is a measure of the overall clustering of the data.
One of the statistics used to evaluate global spatial autocorrelation
is Morans I (Anselin, 2005).
The reviews of the latest researches gave rise to the need for
detailed analyses of each and every indicator that is included in
a comprehensive road safety assessment. It also became necessary to include more and more indicators, according to the qualitative road safety assessment. In almost all recent researches,
nal outcomes have been used for the road safety assessment
at the researched territory. Final outcomes of the road safety
in Serbia are most often represented by a road casualty risk
where the calculated risks are classied in ve standardized
classes that are labeled with marks ranging from 1 to 5. The
lowest mark (1) is associated with the highest levels of risk
and reects the most unfavorable value of the observed indicator, while the highest mark (5) indicates the lowest risk and
the most favorable value of the same indicator. The classication
of the levels of road casualty risks has been made according to
the EuroRAP model.
It is very important to select the indicator for the risk mapping in
the observed area (municipalities) because of the signicant number of various sorts of dened risks (public risk, trafc risk, etc.).
In order to assess the road safety in the observed area, in the best

Diagram 1. A model for selection of a relevant indicator road casualty risk.

D. Kukic et al. / Safety Science 51 (2013) 165177

possible and most realistic way, it will be necessary to select the

road safety indicator risk that will assess the road safety in the
observed area in the best possible way.

The Eq. (1) shows the calculation of the Public risk based on the
weighted number of casualties PR WNC. This risk is based on the
type of consequences of road accidents weighted by corresponding
coefcients, depending on the level of injuries.

2. Methodology

Li  P1 Si  P2 F  P3
 10; 000
Population

PRWNC
In this paper, a statistical analysis of the indicators road casualty risks, has been made by calculating the value of linear
correlation.
The rst step in implementing a model for selecting a relevant
road casualty risk is to dene an observation unit for the so called
entity (municipality, region or else). After having dened the
observation units, output indicators of road safety are being singled out:
- Number of road accidents with casualties (includes the number
of road accidents with injured and the number of road accidents
with fatalities).
- Number of accidents with fatalities.
- Number of fatalities.
- Number of people with serious bodily injuries.
- Number of people with slight bodily injuries.

where Li number of slight injuries; Si number of serious injuries;

F number of fatalities in road accidents. P1 = 1; P2 = 10; and
P3 = 85. The values of coefcients that are associated with the certain levels of injury, were taken from the Report of the Road World
Association PIARC, and they have been published in the document
Road Safety Manual, Recomendations from the Road World Association
(PIARC, 2008)1. The values of coefcients P1, P2 and P3 were taken
from PIARC, because there is still no calculation of the accident costs
in Serbia.
The second risk is the Public risk obtained on a basis of number
of road accidents with casualties PR(RAc) Eq. (2), and it is calculated as the ratio of the number of road accidents with casualties
(number of road accidents with casualties Num.RAc) and the
number of inhabitants in a dened observation unit:

PRRAc
In order to get nal outcomes rates, adequate parameters or
exposure data are singled out and then compared to the output
indicators of road safety (values of road accidents and their consequences). The parameters are chosen depending on the select risk
for future researches. Thus, in order to calculate the value of public
risk, it is necessary to get the information on the population; for a
collective risk, we need data on the road network length, the length
of a road or a road section; for calculating the dynamic trafc risk,
the data on AADT (Average Annual Daily Trafc) and network
length will be necessary; and in order to calculate the trafc risk,
information on the number of registered vehicles is needed. The
number of vehicle kilometers the kilometrage can be calculated
in the following way: AADT  L  365 (veh. km/year) where L is the
length of the network (km). The number of registered motor vehicles within the territory of municipality can be used as a substitute
Per and Al-Haji (2005), Rumar (1999), Smeed (1972), Sorensen
(2002), liupas (2009). Transit trafc should also be added to this
value, i.e. by using fuel consumption data or AADT data divided
into domicile and transit trafc, etc.
The dynamic trafc risk is still unavailable for the most municipalities in Serbia, and therefore trafc risk that is calculated on the
basis of the number of registered motor vehicles within the territory of municipality can be used as a substitute. For those countries
in which the motor vehicle kilometers are not available, the fatality
rate dened as the number of fatalities per motor vehicle will
be used instead (Wegman and Oppe, 2010).
The isolated nal outcomes rates are not the only ones, but are
the most common ones. A model for selection of a relevant risk represents the calculation of ve types of output indicators of road
safety for each of the observed risks.
Input data for the observation of the public risk include: the
number of inhabitants in a dened observation unit, the number
of road accidents and their consequences. All risks are calculated
depending on the type of road accidents and the severity of injuries. Those indicators are obtained on the basis of the weighted
number of casualties (1); on the basis of the absolute data on
number of road accidents with casualties (2); number of fatalities
(3); number of accidents with fatalities (4); and the number of
killed and seriously injured (5). A model for selection of a relevant
risk as an acceptable indicator of the state of road safety is shown
in Diagram 1.
Equations for the calculation of all risks which belong to the
category of public risk, are presented in the text below.

167

Num:RAc
 10; 000
Population

The third risk is the public risk obtained on a basis of number of

fatalities in municipalities PRf Eq. (3). This risk is obtained as
the ratio of the total number of fatalities (number of fatalities
Num.F) and the number of inhabitants (population) in a dened
observation unit:

PRf

Num:F
 100; 000
Population

The fourth risk in the category of public risks is the Public risk
obtained on a basis of number of road accidents with fatalities
PR(RAf) Eq. (4). The number of road accidents with fatalities
and the number of fatalities are not the same number. These numbers can vary by more than 10% (according to the number of road
accidents and consequences in the Republic of Serbia from 2001 to
2010 source: Ministry of Interior of the Republic of Serbia). This is
why it is important to monitor both of these indicators the number of road accidents with fatalities and the number of fatalities.

PRRAf

Num:RAf
 100; 000
Population

where Num.RAf the number of road accidents with fatalities; Population the number of inhabitants in a dened observation unit.
The fth risk in the category of public risks is the Public risk obtained according to the number of fatalities and seriously injured
PRf + s Eq. (5), obtained as the ratio of the sum of the number of
fatalities in road accidents (Num.F) and the number of serious injuries in road accidents (Num.S), and the number of inhabitants (Population) in a dened observation unit. The researches of the
European Road Assessment Programme EuroRAP2 give a special
importance to this indicator, while the model itself is based on the
analysis of road accidents with fatalities and road accidents with
seriously injuried (Hill, 2010).

PRf s

Num:F Num:S
 100; 000
Population

1
PIARC 2008.ROAD SAFETY MANUAL, RECOMENDATIONS FROM THE ROAD WORLD
ASSOCIATION, Chapter 7 (Priority ranking).
2
Hill, 2010. EuroRAP202: Risk Mapping Manual, European Road Assessment
Programme.

D. Kukic et al. / Safety Science 51 (2013) 165177

168

Table 1
Values of correlations of the Public risk in all municipalities in the Republic of Serbia (risk within the territory of the whole municipality).
Means

Std. dev.

PR WNC

PRf

PRf + s

PR (RAf)

Correlations (Spreadsheet1) marked correlations are signicant at p < .05000 N = 161 (casewise deletion of missing data)
PR WNC
117.5824
50.59794
1.0000
0.9506
0.8920
0.9247
PRf
13.0582
7.24601
0.9506
1.0000
0.7278
0.9539
PRf + s
78.5961
31.11583
0.8920
0.7278
1.0000
0.7287
PR (RAf)
11.8421
6.21716
0.9247
0.9539
0.7287
1.0000
PR (RAc)
196.4544
71.89993
0.7528
0.5445
0.8203
0.5712
Mean value of the Public risk
83.5066
30.85363
0.9407
0.7978
0.9402
0.8016

In order to get indicator that represents casualty risk in the best

way, the linear correlation between observed risk and Mean value
of all calculated risks in the respective categories has been tested.
The main idea is to nd indicator that has high linear correlation
and because of that, the indicator could describe and assess road
safety in a more realistic way. Also, idea is to develop approach
for selection of a relevant indicator road casualty risk that will
be simple for use by others (decision makers, practitioners, etc.)
that, by default, do not have special knowledge about statistic
methodology and calculation, multicriteria analysis, etc. Therefore,
the last step of the presented model includes obtaining of just one
risk within the population of risks in an observation unit. It is the
risk with the highest value of linear correlation in comparison to
the mean value of the observed risks.
The linear correlation is most often used for random variable
analysis. Despite the progress in establishing an efcient road
safety system and reducing the number of road accidents like random events, it will be still impossible to estimate how many people
will be killed or injured in the road accidents. Proceeding from this
fact, assessment of the relation between variables, which are in
this case the risks obtained based on the number and consequences of road accidents, can be valid by using the coefcient of
linear correlation.
It is important to say that this research and the correlation
study include all the data on the number of road accidents and
their consequences, from all municipalities in Serbia, i.e. the data
on the total number of road accidents and their consequences from
161 municipalities in Serbia, in the period 2006 to 2008. In this
period, there was a total of 202,529 road accidents, with 2783 persons killed, 15,293 seriously injured and 47,624 slightly injured
source: Ministry of Interior of the Republic of Serbia.
The advantage of the presented way of selecting just one risk for
a particular unit of observation is the possibility of making a reliable risk mapping on the observed territory by using available data
(which are often insufcient). This model also makes possible to
single out the reliable absolute indicator of road safety.

PR (RAc)

Mean value of the Public risks

0.7528
0.5445
0.8203
0.5712
1.0000
0.9270

0.9407
0.7978
0.9402
0.8016
0.9270
1.0000

all the calculated risks for the municipalities in Serbia are given in
the Appendix A. All the assessed values of the calculated correlations are statistically signicant (which was actually expected, considering the origin of the variables whose correlations were
analyzed), and they range between 0.54 and 0.95.
The greatest correspondence with all other risks in the population of the public risks was observed on PR WNC risk. The value of
the linear correlation between this and all other calculated risks
ranges between 0.7528 in relation to the risk PR(RAc) and 0.9506
in relation to the risk PRf. A minimum strength of correlation with
regard to all other researched risks was recorded with the PR(RAc)
risk, which is obtained on the basis of the number of road accidents
with casualties. The value of the strength of the linear correlation
for this risk ranges between 0.5445 in relation to the risk PRf and
0.8203 in relation to the risk PR f + s.
The highest value of the linear correlation compared to the
Mean value of the public risks for all municipalities in Serbia,
was obtained in the Public risk based on the weighted number
of casualties PR WNC (0.9407) and the Public risk obtained
according to the number of fatalities and seriously injured PR
f + s (0.9402). The difference in the strength of the linear correlation for these two parameters is only in the fourth decimal place,
so if this detail is considered separately, it does not have a greater
impact on the decision for selection of an indicator for the risk
mapping. This especially makes sense if we accepted the logic

3D Surface Plot (Public risks analysis 2006-2008.sta 6v*161c)

Mean value of all public risks = Distance Weighted Least Squares
250
200
150
100
50
0

3. Results
Depending on the risk categories, the value of particular indicators describes the road safety situation in a different way. Large
dispersion of results of the analyzed risks is important reason to
choice just one risk for future analysis and nal assessment of
the trafc safety situation. Therefore, the best way to extract one
or two risks that will be appropriate indicators of road safety, is
to perform the selection of the parameters that are correlated the
most with the Mean value of all calculated risks in the respective
risk categories, and also mutually with all of the researched risks.
The values of the linear correlation between all the public risks,
as well as between their mean values for all observation units
municipalities in Serbia are given in Table 1. The risk which
correlates the most with the mean value of the public risks in population is marked in blue, as the best describing risk based on
incoming input for every municipality. Detailed information about

Fig. 1. A 3-D presentation of the risk values PR WNC, PR f + s and the Mean value of
all public risks.

D. Kukic et al. / Safety Science 51 (2013) 165177

169

Fig. 2. Distribution of the nal outcomes risks with highest value of linear correlation, PR WNC and TR WNC for Serbian municipalities (20062008).

Table 2
Values of correlations of the Trafc risk in all municipalities in the Republic of Serbia (risk within the territory of the whole municipality).
Means

Std. dev.

TR WNC

TRf

TRf + s

TR (RAf)

TR (RAc)

Mean value of the Trafc risks

0.7731
0.6003
0.8011
0.6247
1.0000
0.9291

0.9481
0.8292
0.9225
0.8272
0.9291
1.0000

Correlations (Spreadsheet1) marked correlations are signicant at p < .05000 N = 161 (casewise deletion of missing data)
TR WNC
TRf
TRf + s
TR (RAf)
TR (RAc)
Mean value of the trafc risks

57.1803
6.3702
38.0209
5.7775
95.4091
40.5516

25.7265
3.7628
14.9781
3.2574
34.5146
15.2181

1.0000
0.9592
0.8749
0.9261
0.7731
0.9481

of a reliable choice of risk in relation to the collection and processing of road accidents data, because the number of killed
and seriously injured in comparison to the number of slightly injured represents more reliable data, which will also be needed for
the calculation of the risk PR WNC. However, the mutual comparison of the severity of the linear correlation between each risk,
and signicantly better results are on the side of the risk PR
WNC. The values of the linear correlation between the risk PR
f + s and other researched risks in the population of public risks
range between 0,7278 in relation to the risk PRf and 0,8920 in
relation to the risk PR WNC. Taking the above into account, the
risk which coincides best with all the researched risks and also
with the average value of all other risks in the population of
public risks, is the risk PR WNC.
Based on a three-dimensional presentation of the observed
measures (Mean value of public risks, PR f + s and PR WNC), the surface of all obtained values has been designed (Fig. 1). On the y-axis,
the results of the Mean value of public risks in all municipalities in

0.9592
1.0000
0.7240
0.9481
0.6003
0.8292

0.8749
0.7240
1.0000
0.7164
0.8011
0.9225

0.9261
0.9481
0.7164
1.0000
0.6247
0.8272

the Republic of Serbia are shown; the values of PR WNC risk are
shown on the z-axis, while PR f + s risk is shown on the x-axes.
(See Fig. 2).
By comparison with the farthest angles of the obtained values,
the evenness of the inuence of the observed measures on the
Mean value of public risks could be seen (Fig. 1).
The values of the linear correlation between all the trafc
risks, as well as among their mean values for all observation units
municipalities in Serbia are given in Table 2. The highest
value of the linear correlation compared to the Mean value of
the trafc risks for all was obtained in the Trafc risk based on
the weighted number of casualties TR WNC (0.9481). This risk
coincides best with all the researched risks and also with the
average value of all other risks in the population of trafc risks.
The second risk with highest value of the linear correlation compared to the Mean value of the trafc risks is Trafc risk obtained
on a basis of number of road accidents with casualties TR(RAc)
(0.9291).

D. Kukic et al. / Safety Science 51 (2013) 165177

170

Scatterplot (Public risks analysis 2006-2008.sta 6v*161c)

PR WNC = -11,2398+1,5427*x; 0,95 Conf.Int.
Mean value of the Public risks: PR WNC: r = 0,9407; p = 00,0000
350

300

Cajetina

Savski Venac

Velika Plana
Alibunar

250

PR WNC

Kanjia
Sopot

200

Ub
150

100

50

0
0

20

40

60

80

100

120

140

160

180

200

220

240

Mean value of the Public risks

Fig. 3. Scatter plot diagram of differences of the PR WNC risk and the Mean value of public risks for the territory of the whole municipality.

Scatterplot (Public risks analysis 2006-2008.sta 6v*161c)

Mean value of the Public risks:PRf; r = 0,7978; p = 00,0000
PRf = -2,5875+0,1874*x; 0,95 Conf.Int.
45
40

Velika Plana Cajetina

Alibunar

35

Kanjia
Sopot Pecinci
Batocina

30

Savski Venac

PRf

25
20

Ub
15

Novi Sad
Backi Petrovac

10
5

Vracar

0
-5
0

20

40

60

80

100

120

140

160

180

200

220

240

Mean value of the Public risks

Fig. 4. Scatter plot diagram of differences PRf risk and the Mean value of public risks within the territory of the whole municipality.

4. Discussion
In order to determine the differences between the values of researched risks in the best possible way and consequently the reason for selecting an indicator for the risk mapping, a comparison
between the risk with the highest value of the linear correlation

and the risk with the lowest value of the linear correlation (PR
WNC and PRf) has been conducted. In order to best detect this difference, a scatter diagram has been used because the type and level
of their connection can be observed in the best possible way
through a graphical display of points determined by the values of
entities (observation units), at two variables simultaneously.

D. Kukic et al. / Safety Science 51 (2013) 165177

Dispersion of the values obtained for the strength of the linear

correlation is presented in Table 1 and indicates the possible differences that occur in calculating certain types of risk. It could be expressed to the extent in which the obtained value of the analyzed
risk for certain municipalities differs from the smallest to the biggest dened risk band.
In other words, if we observe a municipality which, according to
the value of the public casualty risk, based on the overall number
of road accidents with casualties PR (RAc), belongs to a municipality with a very low level of risk, the second indicator, such as
the Public risk of the weighted number of casualties PR WNC,
classies this municipality into the class of municipalities with a
high level of risk, compared to other municipalities in Serbia (for
example, the municipality of Vracar in the city of Belgrade).
In this case, the values of the observation unit are the Public risk
of the weighted number of casualties PR WNC (the highest value
of the linear correlation) and the Public fatality risk obtained on a
basis of the number of fatalities in municipalities PRf (the lowest
value of the linear correlation).
The examination of the scatter plot diagram shows that the general trend of the points is highly linear, i.e. that it forms an elongated ellipse (Fig. 3). The strength of the correlation for the
observed entities municipalities amounts to 0.9407 for the
PRWNC and the Mean value of all public risks.
The value r = 0.9407 represents the value of coefcient of the
linear correlation (the third row in the diagram header), while
the value p represents the probability of getting as big or a bigger
coefcient of the linear correlation on the sample of a given size, if
there is no linear connection in the population.3

P jtj P tcalculate jH0correct

In this unique way, we can single out those municipalities

where one variable in the observed entity differs greatly from
the increase or decrease of the value or other variable.
In other words, if we notice a considerable increase of the Public
risk of weighted number of casualties PR WNC, in relation to the
Mean value of public risks in some municipalities, these data could
make us direct the research and analysis of the state of road safety
to the municipality with distinct nonlinearity of observed variables.
Fig. 4 shows the scatter plot for the risk with the lowest value of
linear correlation with regard to the Mean value of all public risks,
it is the Public risk obtained according to the number of fatalities of
the whole municipality PRf. The coefcient value of the linear
correlation in this example amounts to r = 0.7978. We can determine that dispersion of points of the observed variables is much
higher compared to the distribution of points for Public risk obtained on the weighted number of casualties PR WNC.
On Fig. 4, likewise on Fig. 3, those municipalities with the biggest deviation from the direction of correlation line have been singled out. Apart from the fact that in some cases the deviation is far
smaller (municipality of Crna Trava), it is much bigger for the
majority of municipalities which are singled out, compared to
the case we had with the risk PR WNC (Fig. 3).
With the risk PRf (Public fatality risk), there is a signicant number of other municipalities with much bigger deviation from the
direction of correlation line (which is expected if we consider the
smaller value of coefcient of linear correlation).
If we consider the fact that when choosing the PRf risk, a greater
number of municipalities deviate from the direction of correlation
line, and that this risk has the lowest value of linear correlation
mutually with all other risks, and also with the Mean value of public risks, we can conclude that the PRf risk is not the best solution
for the risk mapping of the observed unit.
3

Tenjovic, 2002. Statistics in Psychology Manual; 2nd edition; Centre for applied
psychology, the association of psychologist of Serbia, Belgrade.

171

5. Conclusion
In previous researches, the values and names of certain risks
were not explicitly assessed according to the calculated parameters, which would be acceptable for all further researches and risk
mapping on the observed territory.4 In the most prominent papers
in the eld of road safety, the values of risk rates from estimations of
the most serious consequences of road accidents (number of fatalities) in relation to the population (i.e. number of registered motor
vehicles, number of vehicle-kilometers, AADT value, etc.), to taking
into consideration of all the consequences of a road accident,
weighted by adequate coefcients, depending on their severity. A
great number of different risks whose values could be calculated
for every observation unit has been dened in this way.
The main difference between Model for selection of a road casualty risk, which is presented in this paper and the RAP model
(EuroRAP, usRAP and AusRAP) is special analysis of the risk type
indicators (nal outcomes) for risk mapping. Also, the RAP model
is intended for risk mapping on the roads and road sections, not
for the risk mapping for the territories of municipalities, countries,
regions, etc. The RAP model does not analyze risk type indicators,
but it analyzes (or more precisely calculates) two rates (1) crash
risk per kilometer traveled and (2) crash density, obtained from
the data of road accidents with seriously injured and fatalities.
A Model-based risk map for roadway trafc crashes (Nam and
Song, 2008) is using Global spatial autocorrelation to describe
the risk on the observed territory. This model is analyzing all consequences of road accidents and in this sense it is similar to the
Model for selection of a relevant risk that is presented in this paper.
However, this model does not use statistical tools to compare the
risks. The model is using statistical tools for the modeling a spatial
distribution of road accidents.
The Model for selection of a relevant risk is in its basis a model for
monitoring the state of road safety based on nal output which is
presented in rates. It is a fact that every single risk of the analyzed
risks describes the state of road safety by a certain quality level.
However, due to the great dispersion of results of the analyzed
parameters, we have great differences in the observed risks.
The best contribution of this Model reects in the separation of
only one risk which will be used for future researches and assessments of the state of road safety and the risk mapping. This Model
also eliminates other risks which are not the best solution for a
reliable choice of relevant indicator for estimation of the real danger. The use of this Model has shown that the most reliable nal
output for a dened unit of observation municipalities in Serbia,
is the Public risk obtained on the basis of the weighted number of
casualties PR WNC. This risk could be used for future researches
and assessments of the state of road safety in a dened unit of
observation.
By implementing the Model for selection of relevant risk, other
regularities of the observed entities have been noticed, that might
be subjects to further researches. This primarily refers to the inuences of some occurrence or road users behavior on the level of
road safety Elvik (2004). It is indisputable that there are great differences between the municipalities in terms of importance, frequency and quality of the state and local roads, road users
behavior, social attitudes with regard to the dangers of road trafc
etc. In further researches, for the road safety assessment, some
SPIs, suggested for i.e. by ETSC (2001), Hakkert and Gitelman
(2007), Hermans et al. (2009), Gitelman et al. (2010), Wegman
and Oppe (2010) could be included Hakkert et al. (2007). SPIs that
can be used for the future researches are:
4
The so-called risks, are not real risks, they are only rates, risk type indicators.
Their value can be different from those between 0 and 1, and the exposure is only an
estimation of the real danger.

D. Kukic et al. / Safety Science 51 (2013) 165177

172

% of drivers under its inuence,

% of seatbelt using,
% of drivers who are speeding, etc.

In addition to the further researches we can do detailed analysis

and classication of road network in Serbia, road safety assessments, eet analysis, work of emergency services on the territory
of municipalities, etc.

Appendix A
All calculated Public risks for the municipalities in the Republic of Serbia
Num.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

Municipality

PR WNC

PRf

PR f + s

PR (RAf)

Savski Venac
Ub
Novi Sad
Lapovo
Lazarevac
Lajkovac
Ljig
ajetina
C
Indija
Zrenjanin
Backi Petrovac
Mionica
Smederevo
Temerin
Valjevo
Becej
Obrenovac
acak
C
Stara Pazova
Kraljevo
Ruma
Doljevac
Knic
Zabalj
abac
Bogatic
Vrnjacka Banja
Irig
Prijepolje
Kovin
Paracin
Cicevac
Pancevo
Cuprija
Sremska Mitrovica
Velika Plana
Svilajnac
Backa Palanka
Vladimirci
Pecinci
Odzaci
Stari Grad
Topola
Kikinda
Pozega
Senta
Veliko Gradite
Kovacica
Bela Crkva
Jagodina
Barajevo
Petrovac
Raca

282.63
155.12
153.43
244.69
231.64
248.11
185.48
284.75
191.97
139.90
131.24
159.27
167.96
174.83
125.36
122.15
123.14
146.92
146.75
131.87
187.48
150.98
173.81
219.05
156.40
136.00
117.39
176.28
124.77
176.26
135.73
159.00
139.01
138.53
133.33
253.73
141.77
104.92
157.73
205.52
124.31
84.62
149.72
98.11
138.32
131.94
144.57
168.88
114.73
105.46
173.02
98.52
135.30

25.10
15.57
13.03
32.41
26.21
29.30
15.95
36.26
23.52
13.13
9.08
14.13
18.52
18.86
11.71
11.39
13.62
15.09
16.77
12.60
23.33
17.04
20.64
29.08
15.73
13.14
10.07
21.63
12.07
21.74
12.58
18.60
14.68
14.90
13.97
35.98
16.99
10.39
18.00
29.45
11.24
6.00
15.82
8.95
15.48
14.34
16.14
21.51
11.46
8.93
20.29
7.73
12.86

215.66
101.75
111.15
125.59
152.68
160.20
164.06
174.90
105.49
87.09
122.61
133.23
105.33
124.96
86.47
78.89
69.51
101.08
90.27
96.95
102.21
85.20
111.47
121.15
125.58
106.09
93.11
108.15
83.40
105.07
115.49
108.48
93.32
96.32
90.03
135.67
83.62
63.97
116.17
99.20
96.49
74.42
104.12
71.14
87.74
97.78
98.42
95.61
80.20
89.81
109.57
86.93
110.61

25.10
15.57
11.14
28.36
24.50
25.40
15.95
36.26
18.81
11.86
9.08
14.13
16.39
16.50
11.02
11.39
12.21
13.67
15.29
12.87
20.55
13.63
18.58
24.23
14.10
13.14
13.84
16.22
6.58
18.11
10.86
12.40
13.11
11.92
11.25
29.98
16.99
9.84
16.36
26.35
11.24
6.00
15.82
6.96
13.42
14.34
16.14
21.51
11.46
7.05
16.23
8.69
12.86

PR (RAc)

Mean value of Pablic risks

579.54
440.24
361.29
352.46
343.53
341.89
328.12
319.94
313.79
308.72
304.25
288.66
283.83
282.94
277.32
276.51
275.68
275.33
273.77
272.79
272.75
272.65
268.35
266.54
265.00
259.67
255.42
254.14
249.09
247.27
245.28
244.85
244.57
241.31
240.97
240.61
240.42
238.93
235.61
235.59
234.20
233.45
231.96
230.34
229.15
228.15
227.50
227.08
225.86
225.69
224.56
224.08
223.78

225.60
145.65
130.01
156.70
155.71
160.98
141.91
170.42
130.72
112.14
115.25
121.88
118.41
123.62
102.38
100.06
98.83
110.42
108.57
105.42
121.27
107.90
118.57
132.01
115.36
105.61
97.97
115.28
95.18
113.69
103.99
108.66
100.94
100.59
97.91
139.19
99.96
85.61
108.77
119.22
95.50
80.90
103.48
83.10
96.82
97.31
100.55
106.92
88.74
87.39
108.73
85.19
99.08

D. Kukic et al. / Safety Science 51 (2013) 165177

173

Appendix A (continued)
Num.

Municipality

PR WNC

PRf

PR f + s

PR (RAf)

PR (RAc)

Mean value of Pablic risks

54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113

Apatin
id
Srbobran
Mladenovac
Sombor
Brus
Alibunar
Batocina
Sremski Karlovci
Kruevac
Bajina Bata
Backa Topola
Subotica
Vracar
Kula
Kosjeric
Zajecar
Gornji Milanovac
Nova Varo
Razanj
Ni
Bela Palanka
Kurumlija
Pozarevac
Despotovac
Novi Pazar
Palilula
Loznica
Kragujevac
Negotin
Beocin
Arilje
Koceljeva
Smederevska P.
Sokobanja
Surcin
Novi Becej
Bac
Aleksinac
Vozdovac
Leskovac
Ada
Opovo
Vrac
Meroina
Kanjiza
Aleksandrovac
Zitite
ukarica
C
Bojnik
Secanj
Zitorada

101.48
124.87
118.36
135.90
146.89
103.74
241.21
181.94
154.99
108.47
135.50
160.46
107.14
60.52
122.43
131.90
72.16
161.98
141.13
121.38
76.99
176.16
98.11
118.78
120.65
104.00
117.55
90.07
85.68 7.39
88.83
127.65
97.22
125.35
102.24
78.26
127.32
144.48
58.40
112.50
101.01
94.80
93.19
103.79
107.29
101.49
219.68
100.49
157.69
126.60
76.99
109.30
66.82
126.14
112.30
132.32
90.52
81.15
75.84
105.16
104.98

10.16
14.54
11.20
15.88
18.16
8.88
34.85
27.28
18.86
8.88
14.87
20.05
9.66
1.71
12.41
16.67
4.55
20.29
16.68
14.66
6.39
25.50
10.80
12.02
13.02
11.63
12.61
8.49
66.55
8.45
16.58
11.79
12.79
11.31
5.38
13.78
21.05
4.10
14.43
10.54
10.67
10.53
12.10
12.26
11.25
31.50
10.21
22.88
15.83
7.62
14.25
5.49
15.47
14.46
18.53
7.87
8.97
6.73
12.46
12.35

66.03
74.41
97.08
79.38
92.19
85.27
127.79
73.65
98.05
95.66
97.20
106.33
90.52
66.80
98.58
76.19
61.14
98.65
86.74
67.43
54.95
71.85
53.99
98.80
80.69
67.06
84.88
67.89
7.02
66.02
74.60
52.23
110.86
70.22
68.21
95.62
63.14
43.03
51.95
72.48
58.45
64.93
66.57
72.96
60.76
127.23
78.26
75.17
77.74
50.82
54.96
43.94
71.72
67.48
69.18
76.09
50.22
58.62
64.79
69.17

10.16
14.54
9.33
14.61
15.76
8.88
26.14
24.55
11.31
8.63
14.87
15.69
8.31
1.71
11.03
16.67
4.55
16.09
11.68
14.66
5.59
23.18
10.80
10.68
11.71
11.24
11.33
7.33
186.57
7.68
16.58
11.79
12.79
9.52
5.38
13.78
12.38
4.10
13.85
9.66
10.24
10.53
12.10
10.42
9.00
26.66
6.81
19.61
13.25
7.62
12.21
5.49
14.06
14.46
18.53
7.87
8.97
6.73
12.46
12.35

222.47
222.38
222.16
219.72
218.31
216.73
216.37
215.49
214.96
214.16
211.54
210.05
208.00
206.67
199.92
197.60
197.57
197.31
196.84
196.44
195.73
194.70
194.37
194.03
193.93
192.26
192.00
190.94
70.64
183.49
182.35
181.97
181.21
180.32
179.49
179.18
177.04
176.22
172.01
171.09
170.24
170.23
169.45
169.21
168.78
168.42
166.73
166.67
165.18
165.17
164.87
164.77
164.54
163.87
163.08
162.67
162.31
161.33
159.48
158.09

82.06
90.15
91.63
93.10
98.26
84.70
129.27
104.58
99.63
87.16
94.79
102.51
84.73
67.48
88.87
87.80
67.99
98.86
90.61
82.92
67.93
98.28
73.61
86.86
84.00
77.24
83.68
72.94

Vladicin Han
oka
C
Raka
Nova Crnja
Knjazevac
Novi Beograd
Plandite
Mali Ido

70.89
83.55
71.00
88.60
74.72
67.35
85.94
83.62
57.17
72.95
72.96
68.88
69.88
72.80
74.43
70.26
114.70
72.50
88.40
79.72
61.65
71.12
57.30
78.39
74.51
80.33
69.00
62.32
61.85
70.87
71.39
(continued on next page)

D. Kukic et al. / Safety Science 51 (2013) 165177

174

Appendix A (continued)
Num.

Municipality

PR WNC

PRf

PR f + s

PR (RAf)

114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161

Pirot
Boljevac
Vrbas
Vranje
Trstenik
Arandelovac
Uzice
Bujanovac
Rekovac
Gadzin Han
Grocka
Malo Crnice
Zvezdara
Prokuplje
Vlasotince
Blace
Svrljig
Ivanjica
Rakovica
Zemun
LucaniGuca
Mali Zvornik
Varvarin
Kladovo
Golubac
Lebane
Novi Knezevac
Zabari
Preevo
Zagubica
Bor
Sopot
Kucevo
Krupanj
Titel
Osecina
Surdulica
Ljubovija
Majdanpek
Medveda
Crna Trava
Sjenica
Bosilegrad
Dimitrovgrad
Priboj
Tutin
Babunica
Trgovite

79.32
125.77
142.20
76.53
112.21
99.59
77.37
84.52
122.01
50.65
125.44
140.76
65.25
62.68
57.34
60.08
99.32
95.17
77.88
72.30
66.22
82.17
75.37
70.44
62.21
54.71
85.81
60.61
81.18
71.74
58.70
194.87
56.18
45.07
53.57
49.55
72.71
91.88
66.94
25.71
39.02
56.49
34.24
41.71
25.41
44.70
26.91
37.66

8.36
16.83
20.36
6.87
13.59
11.08
7.63
10.01
14.76
3.19
15.46
16.84
5.78
6.19
4.00
4.85
13.50
11.29
7.41
6.44
5.42
9.47
8.28
7.06
6.73
4.01
10.28
5.11
10.51
6.75
5.37
29.43
5.32
3.30
5.87
4.40
10.52
13.68
7.03
0.00
0.00
7.15
3.36
2.84
1.62
4.44
2.12
5.23

50.16
67.30
76.33
57.66
73.40
74.11
52.60
45.42
88.55
41.41
81.71
105.87
52.53
38.49
49.03
48.45
52.07
62.07
61.28
61.22
50.11
56.83
53.01
55.05
40.35
46.82
56.52
58.82
38.20
58.47
43.00
107.90
47.85
41.27
33.24
37.44
27.04
43.01
53.44
34.08
65.03
34.56
23.50
42.56
17.80
36.60
23.30
20.92

5.75
10.52
13.09
6.49
12.91
9.70
6.83
9.24
12.30
15.93
13.25
14.44
5.78
6.19
4.00
4.85
13.50
9.40
6.40
6.44
5.42
11.84
8.28
7.06
10.09
4.01
7.71
7.67
8.60
6.75
4.18
26.16
5.32
3.30
5.87
4.40
10.52
11.73
7.03
0.00
0.00
5.96
3.36
2.84
1.62
4.44
0.00
5.23

PR (RAc)
156.24
155.64
154.85
154.28
153.61
152.37
151.77
150.11
150.05
149.72
149.29
149.19
143.52
142.95
141.09
140.51
138.86
137.30
136.36
134.97
134.07
132.61
132.52
131.28
131.14
131.10
131.02
127.87
127.02
123.68
123.02
120.97
115.20
113.91
113.39
107.92
106.65
97.74
97.03
96.03
91.04
84.61
80.56
79.45
78.50
76.53
65.68
41.85

Mean value of Pablic risks

59.97
75.21
81.36
60.37
73.15
69.37
59.24
59.86
77.53
52.18
77.03
85.42
54.57
51.30
51.09
51.75
63.45
63.05
57.87
56.27
52.25
58.59
55.49
54.18
50.10
48.13
58.27
52.02
53.10
53.48
46.86
95.86
45.97
41.37
42.39
40.74
45.49
51.61
46.30
31.16
39.02
37.75
29.00
33.88
24.99
33.34
23.60
22.18

Appendix B. Appendix
All calculated Trafc risks for the municipalities in the Republic of Serbia
Num.

Municipality

TR WNC

TRf

TRf + s

TR (RAf)

TR (RAc)

Mean value of Trafc risks

1
2
3
4

Bela Palanka
Lapovo
ajetina
C
Lajkovac

147.11
141.62
139.62
135.47

21.37
18.67
17.72
16.07

59.78
73.27
86.43
86.95

19.36
16.36
17.72
13.87

161.23
205.31
156.66
185.71

81.77
91.04
83.63
87.61

D. Kukic et al. / Safety Science 51 (2013) 165177

175

Appendix B (continued)
Num.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Municipality

TR WNC

TRf

TRf + s

TR (RAf)

TR (RAc)

Mean value of Trafc risks

Velika Plana
Doljevac
Nova Varo
Alibunar
Cicevac
Ljig
Pecinci
Knic
Temerin
Plandite
Vladimirci
Malo Crnice
Sopot
Razanj
Mionica
Smederevo
Stara Pazova
Vladicin Han
Ub
Lazarevac
Koceljeva
Novi Becej
Indija
Boljevac
Zabalj
Irig
Batocina
Gornji Milanovac
Barajevo
Ruma
Rekovac
Bajina Bata
Bogatic
Bojnik
Topola
Kovin
Kovacica
Zitite
Aleksinac
Preevo
Kula
Pozega
Kanjiza
Backa Topola
Raca
Svrljig
Vrbas
Prijepolje
Novi Pazar
Bujanovac
abac
Kurumlija
Kosjeric
Meroina
Mladenovac
Ljubovija
Veliko Gradite
Odzaci
Raka
Paracin
Brus

128.57
127.62
104.02
103.61
99.77
96.16
95.94
94.85
93.69
93.32
93.31
93.13
91.88
90.35
89.09
87.87
83.18
82.93
82.60
82.21
81.92
80.81
80.61
80.28
78.79
76.59
75.53
75.42
75.23
73.34
73.01
72.47
72.16
71.13
70.77
70.33
69.78
68.69
67.96
66.98
66.94
66.12
65.90
65.77
65.64
65.49
65.28
65.14
65.00
64.59
64.09
63.75
62.60
62.51
62.19
61.25
61.00
60.92
60.68
60.35
59.90

18.26
13.69
12.25
15.01
11.70
8.26
14.09
11.19
10.30
10.76
10.70
11.95
13.89
10.91
7.89
9.68
9.72
10.18
8.27
9.29
8.43
12.06
9.90
10.87
10.18
9.77
11.33
9.45
8.86
9.25
8.79
8.00
6.97
7.08
7.48
8.68
8.93
9.79
8.66
8.68
6.81
7.41
9.42
8.52
6.13
8.87
9.33
6.29
7.22
7.64
6.46
7.01
7.92
6.92
7.27
9.08
6.72
5.46
8.48
5.60
5.13

68.58
76.28
64.09
55.03
67.82
84.90
44.53
61.44
66.36
58.25
68.44
64.83
50.75
50.26
74.64
55.11
50.39
47.15
54.32
54.35
72.05
33.84
44.18
42.03
44.82
44.78
30.56
45.92
47.53
39.41
53.35
51.75
56.33
46.53
49.22
41.65
39.32
33.68
31.69
31.44
53.97
41.94
37.89
41.63
54.26
34.42
34.99
43.65
42.12
34.77
51.39
35.11
36.09
37.45
36.32
28.87
41.89
48.21
31.81
51.28
49.35

15.19
12.34
8.64
11.02
7.77
8.26
12.42
10.08
9.04
10.76
9.70
9.96
12.32
10.91
7.89
8.57
8.77
9.22
8.27
8.71
8.43
6.68
7.93
6.69
8.63
7.32
10.18
7.48
7.06
8.15
7.34
8.00
6.97
7.08
7.48
7.29
8.93
8.63
8.31
7.10
5.98
6.44
8.08
6.51
6.13
8.87
5.98
3.44
6.99
7.05
5.80
7.01
7.92
5.55
6.69
7.79
6.72
5.46
8.48
4.84
5.13

121.67
240.40
146.23
91.71
152.90
170.97
103.96
147.24
146.01
141.99
138.62
91.47
56.82
146.22
161.96
148.73
150.54
107.99
235.53
121.71
117.20
91.65
131.15
97.77
101.84
105.89
89.29
91.64
97.12
104.59
90.14
112.43
137.75
152.18
109.63
101.52
92.04
76.80
104.84
104.49
108.05
109.13
52.58
81.34
109.96
92.29
71.10
130.16
121.31
114.76
108.48
126.54
93.36
104.01
100.47
65.96
97.52
115.32
75.10
108.94
124.99

70.45
94.07
67.05
55.28
67.99
73.71
54.19
64.96
65.08
63.02
64.15
54.27
45.13
61.73
68.29
61.99
60.52
51.50
77.80
55.25
57.60
45.01
54.76
47.53
48.85
48.87
43.38
45.98
47.16
46.95
46.53
50.53
56.03
56.80
48.92
45.89
43.80
39.52
44.29
43.74
48.35
46.21
34.77
40.76
48.43
41.99
37.34
49.74
48.53
45.76
47.24
47.88
41.58
43.29
42.59
34.59
42.77
47.07
36.91
46.20
48.90
(continued on next page)

D. Kukic et al. / Safety Science 51 (2013) 165177

176

Appendix B (continued)
Num.

Municipality

66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126

Sombor
Kraljevo
Bela Crkva
acak
C
Secanj
Savski Venac
Sjenica
Obrenovac
Sremski Karlovci
Smederevska P.
Cuprija
Zrenjanin
Svilajnac
Valjevo
id
Pancevo
Grocka
Leskovac
oka
C
Trstenik
Sremska Mitrovica
Mali Ido
Zitorada
Mali Zvornik
Zagubica
Beocin
Kruevac
Ivanjica
Jagodina
Kikinda
Becej
Pozarevac
Loznica
Despotovac
Aleksandrovac
Srbobran
ukarica
C
Majdanpek
Sokobanja
Vrnjacka Banja
Knjazevac
Senta
Backi Petrovac
Arilje
Novi Sad
Opovo
Apatin
Backa Palanka
Arandelovac
Blace
Palilula
Nova Crnja
Petrovac
Surdulica
Pirot
Tutin
Vranje
Negotin
Novi Knezevac
Lebane
LucaniGuca

TR WNC
58.96
58.43
58.41
58.21
58.07
57.23
56.05
55.27
55.18
54.99
54.11
53.42
52.95
51.90
51.81
51.31
51.03
50.96
50.95
50.91
50.61
50.32
50.21
50.15
49.80
49.69
48.99
47.99
47.50
46.89
46.74
46.53
46.25
46.14
46.11
45.78
45.19
45.06
45.03
44.76
44.34
43.75
43.58
43.25
43.25
43.20
42.98
42.47
42.44
42.08
41.95
41.59
41.52
40.25
39.84
37.91
37.84
37.78
37.51
37.47
37.17

TRf
7.24
5.59
6.05
5.97
7.92
5.08
7.12
6.10
6.54
6.09
5.81
5.04
6.36
4.87
6.15
5.43
6.31
5.74
6.31
6.18
5.43
5.58
4.15
5.79
4.64
6.36
4.00
5.66
4.03
4.30
4.39
4.70
4.34
4.98
4.66
4.39
5.64
4.74
3.17
3.86
4.90
4.66
2.89
5.29
3.67
4.92
4.33
4.17
4.72
3.39
4.51
2.72
3.27
5.82
4.20
3.78
3.40
3.58
4.74
2.77
3.00

TRf + s

TR (RAf)

37.38
42.96
39.23
40.17
28.14
43.72
34.25
31.25
35.89
37.73
37.64
33.15
31.05
35.72
30.13
34.07
33.15
31.38
31.72
33.16
33.52
35.45
32.99
34.64
40.73
29.65
43.24
31.49
40.40
33.58
30.12
38.76
34.93
30.83
36.06
37.22
27.82
36.00
38.91
35.33
27.44
32.81
41.27
23.02
31.30
28.84
27.75
26.08
31.57
33.98
30.29
41.72
36.52
14.98
25.16
31.08
28.46
28.11
23.11
31.83
28.28

6.31
5.71
6.05
5.40
6.39
5.08
5.92
5.47
3.92
5.14
4.65
4.52
6.36
4.58
6.15
4.92
5.39
5.50
6.31
5.88
4.29
5.58
4.15
7.25
4.64
6.36
3.89
4.74
3.18
3.24
4.39
4.18
3.75
4.49
3.11
3.53
4.73
4.74
3.17
5.28
4.90
4.66
2.89
5.29
3.14
4.92
4.33
3.98
4.14
3.39
4.05
2.72
3.66
5.82
2.87
3.78
3.22
3.26
3.76
2.77
3.00

TR (RAc)

Mean value of Trafc risks

88.94
120.76
113.66
109.01
80.21
117.19
83.25
124.13
79.81
96.93
94.30
116.90
90.01
114.30
91.04
90.85
60.37
91.48
80.14
69.66
89.86
79.96
123.73
80.79
86.64
72.17
96.87
69.41
101.56
109.75
104.68
76.01
98.63
74.28
76.78
84.39
58.99
65.02
101.90
97.33
88.71
78.06
101.41
80.34
101.98
70.48
93.81
97.49
64.98
98.23
68.48
84.97
94.47
59.10
78.68
64.24
76.33
78.25
55.51
89.37
75.83

39.77
46.69
44.68
43.75
36.15
45.66
37.31
44.44
36.27
40.18
39.30
42.61
37.35
42.28
37.06
37.32
31.25
37.01
35.09
33.16
36.74
35.38
43.04
35.72
37.29
32.85
39.40
31.86
39.34
39.55
38.06
34.04
37.58
32.14
33.34
35.06
28.47
31.11
38.44
37.31
34.06
32.79
38.41
31.44
36.67
30.47
34.64
34.84
29.57
36.21
29.86
34.75
35.89
25.19
30.15
28.16
29.85
30.20
24.93
32.84
29.46

D. Kukic et al. / Safety Science 51 (2013) 165177

177

Appendix B (continued)
Num.

Municipality

TR WNC

127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161

Zajecar
Subotica
Prokuplje
Varvarin
Vrac
Kragujevac
Osecina
Vozdovac
Trgovite
Bosilegrad
Crna Trava
Uzice
Ada
Golubac
Kladovo
Vlasotince
Krupanj
Gadzin Han
Rakovica
Ni
Zabari
Dimitrovgrad
Titel
Medveda

36.91
36.91
36.85
36.85
36.29
35.17
34.66
34.44
33.70
33.51
33.04
33.02
32.84
32.68
31.84
31.71
31.70
31.23
29.72
29.62
29.04
28.48
28.31
27.42
27.39
26.70
24.16
23.09
22.52
22.02
21.69
20.29
18.91
16.38
0.00

Bor
Zemun
Kucevo
Priboj
Babunica
Zvezdara
Bac
Novi Beograd
Stari Grad
Vracar
Surcin (excluded)

TRf
2.34
3.30
3.61
4.01
4.12
3.04
3.19
3.61
4.70
3.38
0.00
3.26
3.65
3.57
3.19
2.21
2.34
1.95
2.83
2.46
2.45
1.93
3.03
0.00
2.50
2.37
2.30
1.47
1.82
1.96
1.31
1.79
1.35
0.46
0.00

TRf + s

TR (RAf)

31.20
31.23
22.76
26.18
24.75
27.32
25.67
24.67
18.64
22.41
54.64
22.42
23.09
21.03
24.91
27.06
29.06
25.55
23.37
21.14
28.17
28.98
18.25
36.44
20.16
22.64
20.51
16.18
19.18
17.72
16.91
15.73
16.60
18.10
0.00

2.34
2.87
3.61
4.01
3.52
2.88
3.19
3.30
4.70
3.38
0.00
2.92
3.65
5.41
3.19
2.21
2.34
9.92
2.44
2.15
3.68
1.93
3.03
0.00
1.95
2.38
2.30
1.47
0.00
1.95
1.31
1.79
1.35
0.46
0.00

References
Al-Haji, G., 2007. Road Safety Development Index (RSDI) Theory, Philosophy and
Practice. Dissertation No. 1100, Department of Science and Technology,
Linkping University, Sweden.
Anselin, L., 2005. Exploring spatial data with GeoDATM: a workbook. Spatial
Analysis Laboratory, 138.
Ekler, V., 2010. Measuring and understanding road safety performance at local
territorial level. Safety Science 48, 11971202.
Elvik, R., 2004. Dimensions of Road Safety Problems and Their Measurement.
Institute of Transport Economics, Oslo, Norway.
European Transport Safety Council, 2001. Transport Safety Performance Indicators,
Brussels.
Gitelman, V., Doveh, E., Hakkert, S., 2010. Designing a composite indicator for road
safety. Safety Science.
Hakkert, A.S., Gitelman, V., 2007. Road Safety Performance Indicators: Manual.
Deliverable D3.8 of the EU FP6 Project SafetyNet.
Hakkert, A.S, Gitelman, V., Vis, M.A., 2007. Road Safety Performance Indicators:
Theory. Deliverable D3.6 of the EU FP6 project SafetyNet, 10.
Hermans, E., Brijs, T., Wets, G., Vanroof, K., 2009. Benchmarking road safety: lessons
to learn from a data envelopment analysis. Accident Analysis and Prevention 41,
174182.
Hill, J., 2010. EuroRAP202: Risk Mapping Manual. European Road Assessment
Programme, Brussels.
Koornstra, M., Lynam, D., Nilsson, G., Noordzij, P., Pettersson, H.-E., Wegman, F.,
Wouters, P., 2002. SUNower: A Comparative Study of the Development of Road
Safety in Sweden, the United Kingdom, and Netherlands. SWOV Institute for
Road Safety Research, Leidschendam, The Netherlands.

TR (RAc)

Mean value of Trafc risks

100.79
72.33
84.48
65.27
57.96
76.48
73.17
58.12
36.95
77.31
78.20
64.85
62.32
67.91
59.48
78.10
79.26
92.69
52.15
75.18
61.47
54.36
61.26
101.78
57.35
49.86
49.34
71.35
54.25
48.29
71.82
43.35
52.09
56.01
0.00

34.72
29.33
30.26
27.26
25.33
28.98
27.98
24.83
19.74
28.00
33.18
25.29
25.11
26.12
24.52
28.26
28.94
32.27
22.10
26.11
24.96
23.14
22.78
33.13
21.87
20.79
19.72
22.71
19.55
18.39
22.61
16.59
18.06
18.28
0.00

Nam, C., Song, J., 2008. A Model Based risk Map for Roadway Trafc Crashes.
University of Arkansas.
Per, L., Al-Haji, G., 2005. Road Safety in Southeast Asia Factors Affecting Motorcycle
Safety. Dept of Science and Technology, Linkoping University, Sweden.
PIARC, 2008. Road Safety Manual, Recomendations from the Road World
Association (Priority Ranking) (Chapter 7).
Rumar, K., 1999. Transport Safety Visions, Targets and Strategies: Beyond 2000. First
European Safety Lecture, European Trafc Safety Council, Brussels.
liupas, T., 2009. The impact of road parameters and the surrounding area on trafc
accidents. Transport 24 (1), 4247.
Smeed, R.J., 1972. The usefulness of formula in trafc engineering and road safety.
Accident Analysis and Preview 4, 3312.
Sorensen, M., 2002. Best Practice Guidelines on Black Spot Management and Safety
Analysis of Road Networks. Brussels.
Tenjovic, L., 2002. Statistics in Psychology Manual, second ed. Centre for Applied
Psychology. The Association of Psychologist of Serbia, Belgrade.
Wegman, F., Oppe, S., 2010. Benchmarking road safety performances of countries.
Safety Science 48 (9), 12031211.
Wegman, F., Eksler, V., Hayes, S., Lynam, D., Morsink, P., Oppe, S. 2005.
SUNower+6: A Comparative Study of the Development of the Road Safety in
the SUNower+6 Countries: Final Report. SWOV Institute for Road Safety
Research, Leidschendam, The Netherlands.
Wegman, F., Commandeur, J., Doveh, E., Eksler, V., Gitelman, V., Hakkert, S., Lynam,
D., Oppe, S., 2008. SUNowerNext: Towards a Composite Road Safety
Performance Index. SWOV Institute for Road Safety Research, Leidschendam,
The Netherlands.
Zhang, W., Tsimoni, O., Sivak, M., Flannagan, M.J., 2010. Road safety in China:
analysis of current challenges. Journal of Safety Research 41 (1), 2530.