PUBLIC ROAD
SAFETY POLICY
ASSESSMENT
Public thematic report
June 2021
2
Abstract
If we observe the statistics on the scale of half a century, the effectiveness of the road
safety policy is indisputable and the evolution of its results spectacular: from more than 18,000
killed on the roads of France in 1972, that number fell to less than 3,500 in 2018 and 2019. It
is therefore an indisputable success of a public policy pursued with perseverance over several
decades.
However, there are signs that this virtuous dynamic may have reached a plateau. On the
one hand, since 2013, the number of deaths has stopped decreasing. While the number of
deaths per million inhabitants rose in France from 69.4 to 52.2 between 2008 and 2013, it
remained broadly stable at that level until 2019. However, that is the main indicator of the
success of this policy. This plateau phenomenon is common to most European countries,
suggesting the idea, if not of an impenetrable floor, at least of a diminishing return to the policy.
However, France's relative position worsened over the period: while the number of
deaths per million inhabitants remained stable in France, the European average fell by 7 points
to level out at 51.5. The European countries whose results were closest to France in 2008
surpassed it (Spain and Luxembourg), and others like Italy and Belgium improved more than
France did over the period. From seventh place in the Europe of 28 in 2008, France fell to
fourteenth in 2019.
Evolution of the number of deaths per million inhabitants in Europe and France
(2008-2019)
Source: Court of Accounts based on Europe Transport data
3
At the same time, the change in the number of people injured as a result of traffic
accidents, as well as the situation of certain vulnerable categories of road users such as
cyclists or seniors, suggest that beyond the number of deaths, a more detailed measurement
of the performance of the road safety policy would lead to a more nuanced assessment over
the recent period.
These data do not call into question the long-term success of the policy pursued.
However, they justify an in-depth analysis of its motivations and its performance, in order to
identify areas for progress and adjustments that would make it possible to better take
advantage of them. This is what the Court set about when it undertook to carry out a
comprehensive and detailed assessment of road safety policy for the first time.
For this assessment, it chose the 2008-2019 period, in order to have a sufficiently broad
scope around the pivotal year that 2013 appears to have been. This enabled it to analyse the
instruments of long-term success reflected in the drop in the number of deaths up to that date,
while questioning the causes of the apparent interruption of this dynamic in the period that
follows, by seeking if it is possible to re-engage with it and by what means.
The methodology used is based on a global approach to road safety policy, based on
assessment questions. Conversely, the Court did not study specific measures such as the
points-based licence, a subject for which, moreover, an assessment was initiated by the road
safety delegation, or preventing risks in companies.
The assessment questions were defined for this purpose in consultation with the public
authorities in charge of policy as well as road safety users and stakeholders. They were divided
into three major questions, relating to the relevance of the priorities and levers of this policy,
its effectiveness in view of the results obtained, and its consistency with other public policies
(education, health and mobility in particular) with which it is linked at the national and local
level.
The Court chose not to make a cost/benefit analysis of the policy, due to insufficient available
data.
The road safety policy is implemented by many ministries, in its dimensions of prevention,
education, control and sanction, but also by local authorities and the health insurance system. Despite
methodological imperfections, State expenditures are estimated at € 4
billion, mainly borne by the
Ministry of the Interior (69% of expenditures) and the Ministry for the Ecological and Inclusive
Transition (24% of expenditure). The actions of local authorities are only known for road maintenance
and renovation expenditures
(over € 7.2 billion), not all of which have road safety as their priority. This
approach is nevertheless reductive, with the communities actively contributing to the financing of
preventive actions which are not quantified. The cost of the actions of nursing staff in the service of
risk prevention (especially addictions) is also unknown.
In total, the estimated budget of over € 11 billion devoted annually to road safety policy, which
appears stable over the period studied, only provides an imperfect order of magnitude.
At the end of its assessment work, the Court arrived at qualified conclusions on these
three points: two studies conducted on this occasion confirm the principles of the policy
implemented, particularly the priority given to the behaviours of road users. This essential lever
appears relevant in retrospect with regard to accident data, the objectives pursued, and the
most efficient levers of action available to the State.
4
However, in view of the recent evolution of the results, the major national measures
targeting behaviours will have to be increasingly supplemented by the use of a diversified
range of other means of action, relating to vehicles, signage, and infrastructure. Such
measures should be situated within a conceptual framework that has been rethought in view
of the changes in mobility and technologies. This diversification of levers must correspond to
a better linking of road safety with the public policies which contribute to it and which fall under
other ministerial departments than the Ministry of the Interior, thanks to a strengthening of its
interministerial dimension.
Finally, in its implementation, the road safety policy must be more attentive to its
acceptability, its application at the local level, as well as the prioritizing and clarity of its content.
On these bases, the road safety policy would benefit from being part of a renewed
conceptual framework, which could be inspired by the so-
called “safe system” approach
adopted by the European countries which have had the best performance in this area, a
comprehensive approach which consists in the long term of using additional levers, covering
behaviour, vehicles and infrastructure all at once, in a broader concept of road safety.
A policy defined in the 2000s, focused on user behaviour
The fundamental orientation chosen in the 2000s, and which continues to inspire the
approach of the public authorities, legitimately favours actions targeting the behaviour of road
users.
Since the 1970s, knowledge of the determining factors of road accidents has been the
subject of numerous scientific studies, which classify the causes of accidents into three main
families: relating to behaviour (H), environment and infrastructure (E), the last being that
relating to the vehicle (V). This work consistently highlights a decisive presence of factors
belonging to the H family. As part of this assessment, the study conducted by the Court on
fatal accident frequency for 2015 in France, in cooperation with the Centre for Studies and
Expertise on Risks, the Environment, Mobility and Planning (CEREMA), confirms this
conclusion.
The central option of one action on behaviour is further supported by two factors; on the
one hand, due to the existence of a powerful and efficient tool to serve this choice at a limited
cost: automatic radar control; on the other hand, due to the fact that the State had only a limited
capacity for taking action on the two other families of factors, particularly road infrastructure,
which has now essentially been passed on to the administrative departments.
The priority that the public authorities have thus given to the behaviour of road users was
in fact facilitated by the appearance automatic traffic control in 2003. From that date, its
deployment, supported by an efficient chain for processing offences entrusted to a public
agency, the ANTAI, has made it possible to obtain undeniable results at a limited cost.
The assessment confirmed this, by establishing that the decommissioning of the radars
during the yellow vests crisis had led to a significant relaxation of the behaviour of motorists
and an increase in the average speed on the road segments concerned, thus validating the
relevance and contribution to user safety of automated control. The decrease in the average
speed observed on the roads affected by the reduction of the speed limit to 80 km/h goes in
the same direction.
5
With regard to infrastructure, the dynamic of decentralizing road management has
resulted in network segmentation according to the division of jurisdiction between public
authorities, without sufficient feedback for defining and implementing a comprehensive
infrastructure modernization strategy. In reality, the State hardly manages to act beyond the
network for which it is responsible.
In addition, advances in the design and equipment of vehicles that make them safer
mainly depend on the innovative capacity of the car manufacturers themselves. The State can
only act indirectly in that area, by engaging in public-private cooperation, by supporting
certification and listing efforts at the European level for driver assistance systems and by
adapting regulations, to facilitate the development of the driving assistance and safety devices
it includes, pending automated and connected driving. It is nevertheless a promising area for
progress which road safety policy must fully take into account.
Fundamentally relevant, the choice of a priority action on behaviour was also a default
choice, which reflected the State's difficulties in acting on other factors, particularly road
infrastructure.
Number of measures by priority action type (2006-2019)
Source: Court of Accounts according to the IRB (1) Have
been grouped under the heading of “other offences” such as
crossing the white line, or failure to respect emergency lanes, which were the subject of a measure over the period. (2)
One-off measures that cannot be grouped together with others have been grou
ped together under the heading, “other”.
6
A model that must evolve in three directions: clarification, differentiation,
inclusiveness
If the Court's assessment thus confirms the central focus of the road safety policy, it
nevertheless leads to questioning concerning the significance of the plateau phase of its
results, observable since 2013. There was, of course, a drop of 196 (- 5.3%) in the number of
fatalities in 2018 compared to 2017, but this trend did not continue in 2019, a stable year
compared to the previous one with 3,239 people dying on the roads. There were 3,268 killed
in 2013. With a drop of only 30 people killed per year over seven years, it would be premature
to see 2018 as another pivotal year that would see the downward trend resume that was
interrupted in 2013. The health crisis and the corresponding drop in road traffic will also make
it impossible to interpret the sharply declining data for 2020 and 2021.
In view of such data, it is legitimate to retain the hypothesis of a plateau in the results,
especially as they remain clearly below the government objectives, which were set in 2012 at
reducing fatalities to 2,000 deaths by 2018. Thus, the effectiveness of the policy is slowing. It
was therefore necessary to examine in depth the reference model of road safety policy to
identify avenues likely to improve its results which, in any event, remain far from the objectives
and leave France significantly less well placed than it had been, compared to its most
successful European partners.
This examination led the Court to retain three main findings.
The objectives of the road safety policy remain too focused on the number of fatalities
and should evolve to reflect two series of developments: the number of serious injuries, which
is increasing, a phenomenon whose consideration runs up against methodological difficulties
which other countries such as Germany have nevertheless been able to overcome and which
the majority of European countries retain today; the number of victims among vulnerable
populations, the volume of which is increasing due to changes in mobility (pedestrians and
cyclists) and changes in demographics (elderly). The somewhat hesitant emphasis placed on
these categories by the road safety policy should be more constant and more resolute.
The priority given to automatic traffic controls must be accompanied by a clarification of
the volume, methods and objectives of such tools and an increased consultation with local
authorities on their implementation. It must also be accompanied by a continued presence,
active or dissuasive, of law enforcement officers at the roadside and in traffic flows, particularly
on the part of the national police, whose direct involvement in the road safety policy has
decreased significantly thanks to the development of automatic traffic controls.
Finally, road safety policy must better integrate the differentiation, which is increasing,
between urbanized areas where the problem of road safety is changing thanks to changes in
mobility, and rural and outlying are
as, where the classic model of “road safety” on roads and
highways outside the urban environment survives. Thanks to this development, cities are now
inventing innovative approaches, which would benefit from being better taken into account by
the national road safety policy. More generally, the issues linked to mobility and ecological
transition policies, which are currently converging, too often leave aside the road safety policy,
which should be made more consistent with those policies.
A common factor in these findings is the issue of citizens' adherence to road safety
policy. The lack of readability of the doctrine for the use of automated sanction control fuels
public opinion with criticisms, which are often unjustified, but to which appropriate responses
are not always provided. The emblematic measures on which the debate is focused would
7
benefit from being better linked to an overall medium-term strategy which would lend itself
more to establishing a consensus.
In this regard, the governance of road safety policy could be significantly improved,
particularly in its interministerial dimension, which is currently insufficient, while the road safety
delegation is a department of the Ministry of the Interior. The territories are involved to a certain
extent in the implementation of the policy, but hardly at all in its conception. Finally, consultation
with private, industrial, and stakeholder actors could be improved.
However, the support of the citizens is an essential condition for success. The qualitative
study initiated by the Court through “discussion groups” led, according to the “focus group”
method, by the French Institute of Public Opinion (IFOP) as part of the assessment, shows
that, beyond adherence in principle, the policy led gives rise to strong reservations and
ambiguities. It is desirable, under these conditions, that the tools of communication,
association, or engagement likely to reinforce the public’s adherence, be more mobilized.
The opportunity for a conceptual and practical renewal
Carrying out a conceptual and practical renewal of the French road safety policy appears
possible and necessary.
A comparison of the French case with that of other European countries highlights the
advantages of comprehensive strategies, which tackle road safety systemically and seek to
act simultaneously on behaviour, vehicles, and infrastructure, according to an approach
sharing the conception and the implementation of policies with the largest number of actors
possible.
The “safe system”, promoted by the Inter
national Transport Forum (ITF) and which has
its origins in the approaches implemented in Sweden and the Netherlands in the 1990s,
represents the most successful comprehensive approach. This integrates all the dimensions
of the policy to be conducted: establishing long-term objectives and intermediate stages, as
well as assessment indicators. The “safe system” aims to ensure that, in the “behaviour,
vehicle, infrastructure” triad, the failure of one element can be mitigated by the other two.
Anticipating the possible failure of the driver does not reflect a greater tolerance for deviant
behaviour, but requires infrastructure to be designed in such a way as to minimize the material
and bodily consequences of accidents.
Whether they are copying the “safe system” model or adapting it to local characteristics,
most of the countries that have the best road safety results in Europe rely on long-term plans,
generally over ten years, which include in-depth mid-term assessments. These plans are
multidimensional and supported by a plurality of political, economic, and social forces. Such a
conception of public action could serve as a useful reference for French road safety policy.
The French approach would also benefit from relying more on initiatives and successes
in the field. Responsible for road safety policy, the State must ensure the prevalence of national
rules which govern the behaviour of road users, guarantee the quality of the infrastructure
made available to them, and make vehicles safer. However, the identification, promotion and
dissemination of effective local practices from one territory to another are powerful levers of
action, which could be further mobilized. The link between national and departmental plans,
which is currently limited, should be strengthened.
8
As part of the assessment, the Court compared three pairs of administrative departments
with contrasting results in terms of road safety, despite very similar socio-geographic
characteristics; from this it drew five key success factors: “The existen
ce of clear, explicit local
policies”; “The quality of exchanges and cooperation between the network of associations, the
prefecture, and the departmental council”; “The continuity of preventive actions carried out by
local actors”; “The commitment of national education”; “The availability and sharing of quality
information on accidents in the department”. These conclusions join those drawn from the
observation of foreign best practices, which highlight a decisive role played by the quality of
exchanges between actors and the consensus reached around the major choices of road
safety policy.
The overall management of road safety policy must also give a full place to
experimentation and assessment. To achieve this more easily, it is necessary to improve the
tools for measuring and qualifying accidents. The current accident analysis method should be
overhauled and modernized, by generalizing the Traxy system, developed under the project
management of the National Interministerial Road Safety Observatory (ONISR). As the new
road safety measures by nature lend themselves to experimentation, it should be used more
widely. For measures for which the support of citizens is decisive, the implementation of Article
37-1 of the Constitution could be considered, which allows, in the law and the regulations,
providing for experimental provisions for a limited duration and purpose. Finally, road safety
action plans should be systematically associated with monitoring and impact measurement
instruments, starting from their conception.
Knowledge of the costs incurred in supporting the road safety policy can also be
significantly improved. Originally designed to give the public assurance that the proceeds of
fines from automatic traffic controls would be allocated to the road safe
ty policy, the “road
traffic and parking control” special allocation account (CAS) has reached a degree of
complexity which prevents it from achieving its objective, which led the Court to recommend
its abolition. It proposes, instead of the CAS, to make the transversal policy document (DPT)
“road safety” more exhaustive and readable in order to improve knowledge of the costs of the
policy, in any case much higher than the proceeds from fines contained in the CAS, and thus
be more accountable to the public on the content and means of the policy.
Finally, to reorient French policy towards a more coherent and better shared approach,
it is essential to adapt its governance, by increasing its interministerial dimension and ensuring,
in particular, that the ministry responsible for transport plays an increased role in it, alongside
the Ministry of the Interior. The formalized involvement of Parliament, local authorities and their
federative organizations is also desirable, in order to strengthen the consensus around the
road safety policy. A clarified multi-year strategy with a fixed interval, comprising a greatly
reduced number of measures, associated with measurable objectives, could be submitted to
Parliament for approval, and the duration aligned with the five-year departmental scheduling
currently in force.
All in all, a conceptual and practical renewal of the road safety policy appears to the
Court to be necessary and possible. This should make it possible to extend the historic success
recorded over the past decades to take a new step, which would bring France to the level of
results of its most effective European partners
9
Summary of recommendations
Adopt strategic planning
15.
Develop in 2021, according to a process involving the National Road Safety Council
(CNSR) and the Parliament, a national action plan for the 2022-2030 period, revisable at
mid-term, accompanied by diversified, quantified objectives consistent with those of the
European Union and inspired by the “safe system” approach
(SGG)
.
3.
Establish, in the next road safety plan, diversified, quantified objectives for improving the
results of the policy, including the indicator of the number of “seriously injured”, in
accordance with the objectives set by the European Union in the 2020-2030 action plan
(DSR).
7.
Establish in the road safety plans an assessment plan integrating the implementation of
measures and, whenever possible, monitoring their impact
(DSR).
16.
Mandate the departmental prefects to draw up, according to a process involving the
territorial road safety committee (CTSR) and representatives of local authorities, a
departmental action plan for the 2022-2030 period, revisable at mid-term, consistent with
the national action plan, with quantified objectives and including a section devoted to
prevention
(DSR).
2.
In the measures and resources of the road safety plans, maintain the support provided by
the public authorities for the development of automatic driving assistance
(DSR, DGE,
SGPI).
Seek greater adherence
11.
For certain measures for which the support of the citizens is decisive, conduct experiments
based on Article 37-1 of the Constitution
(SGG)
.
8.
Develop communication actions encouraging their recipients to engage in actions seeking
to change behaviour
(DSR).
Optimize behavioural control
5.
Establish, in a multi-year perspective, the objectives and the doctrine for using automatic
traffic control
(DSR).
6.
Clarify the objectives of setting up radars, systematically involve local players in the choices
made, and adapt communication on the decisions taken and their justification
(DSR).
9.
Complete the control plans carried out by law enforcement officers with a section describing
their participation in preventive actions
(DGPN, DGGN)
.
4.
As part of the reform of the civil reserve of the national police, integrate road safety into the
missions of reservists from civil society and provide the necessary training
(DGPN).
10
Improve driving instruments
1.
Further engage the local road management authorities in establishing and reporting
statistics relating to the road infrastructure for which they are responsible by adopting the
application texts provided for by Article L. 1614-7 CGCT and Ordinance No. 2016-1018 of
27 July 2016
(DGCT, DGITM).
10.
Complete the roll-out of Traxy and provide appropriate access to the resulting data, as
broad as possible, to various categories of audiences: national and regional public
decision-makers, association managers, researchers, citizens
(DSR).
17.
Eliminate the “road traffic and parking control” special allocation account (CAS) and regroup
all road safety expenditures in a single budget annex, making it possible to make a
connection with road safety plans while still identifying the use of traffic fine proceeds
(DSR,
DB).
Improve administrative organization
12.
Place the road safety delegation under the joint authority of the ministers responsible for
the interior and transport
(SGG).
13.
Create, within the National Road Safety Council (CNSR) a State-Territory Commission,
bringing together the competent ministers and associations representing regions,
departments and the municipal block, and submit for its opinion the orientations of the road
safety policy
(SGG).
14.
In the administrative department
s, replace the road users’ consultative commission and the
departmental road safety commission with a territorial road safety council (CTSR)
exercising broader consultative powers and involving all the stakeholders in the policy, like
the National Road Safety Council (
DSR)
.
11
Appendix 4: main road safety measures in European Union
countries
Speed limits
Country
Maximum speed limit on
motorways (or
“expressways”)
Maximum speed limit on
“non
-
urban roads”
Germany
No speed limit on most portions
(130 km/h recommended)
100 km/h
Austria
130 km/h
100 km/h
Belgium
120 km/h
70 km/h (Flanders)
90 km/h (Wallonia)
Bulgaria
120-140 km/h
90 km/h
Cyprus
100 km/h
80 km/h
Croatia
130 km/h
90 km/h
Denmark
130 km/h
80 km/h
Spain
120 km/h
90 km/h
Estonia
90 km/h
(110 in summer on two-lane
expressways)
90 km/h
Finland
Speed limit always indicated by
signs:
80/100/120 km/h
80 km/h
France
110-130 km/h
80 km/h (on two-way roads
with one lane on each side,
without a centre divider)
90 km/h
Greece
110-130 km/h
90 km/h
Hungary
110-130 km/h
90 km/h
Ireland
120 km/h
80 km/h (on secondary
roads)
100 km/h (on national
roads)
Italy
130 km/h
90 km/h
Latvia
No motorways
80-90 km/h
Lithuania
110 km/h from November to
March
130 km/h from April to October
90 km/h (asphalt or concrete
roads)
70 km/h (other roads)
Luxembourg
130 km/h
(110 in case of rain)
90 km/h
Malta
No motorways
80 km/h
Netherlands
100-130 km/h
80 km/h
Poland
100-140 km/h
90-100 km/h
Portugal
100-120 km/h
90 km/h
Czech
Republic
110-130 km/h
90 km/h
Romania
130 km/h
90-100 km/h
United
Kingdom
112 km/h
96 km/h
Slovakia
130 km/h
90 km/h
Slovenia
110-130 km/h
90 km/h
Sweden
110 km/h
70 km/h
12
The blood alcohol level permitted while driving
The European Commission has recommended that all countries adopt a normal
authorized level not exceeding 0.5 g/l. All the Member States respect this threshold, except
Malta, which authorizes a maximum rate of 0.8 g/l.
While the maximum authorized level is usually 0.5 grams (g) per litre of blood - i.e., 0.25
mg of alcohol per litre of exhaled air - as is the case in France, it varies between 0 and 0.8 g/l
in EU countries.
The presence of alcohol in the blood of drivers is prohibited in four countries: the Czech
Republic, Hungary, Slovakia, and Romania. In three other countries (Estonia, Sweden,
Poland), the rate should not exceed 0.2 g/l.
Some countries have also introduced a lower blood alcohol level for drivers who have
little experience on the roads (usually less than five years). Among them, Spain limits the blood
alcohol level to 0.3 g/l for young drivers. Greece, Ireland, Latvia, Luxembourg, the Netherlands
and France have set this threshold at 0.2 g/l, and Austria at 0.1 g/l.
In Croatia, Germany, Italy, Slovenia and Lithuania, although alcohol is usually tolerated
up to 0.5 g/l (0.4 g/l in Lithuania), it is strictly prohibited for young drivers.
The use of psychoactive substances
The prohibition on psychoactive substances while driving exists in many countries
(Portugal, Sweden, France). In France, the use of these substances is strictly prohibited,
regardless of the quantity absorbed, and punishable (main penalties of two years'
imprisonment, a fine of € 4,500 and withdrawal of six points on the driving license, and
additional penalties, particularly a license suspension or cancellation, etc.).
Other countries (Spain, Ireland, the Netherlands, etc.) penalize the fact that the ability to
drive is impaired if it is due to the consumption of drugs or medication. Ireland further specifies
that “
the major consideration is fitness to drive or safety, rather than the legal status of the
product
”.
Finally, a few (e.g., Belgium, Finland) sanction the use of such substances in different
ways by combining the two approaches.
Wearing of seat belts and helmets
It is compulsory in all Member States:
for the seat belt, both in the front and in the rear of the vehicle;
for helmets, both for the driver of the motorcycle and for the passenger.
The use of mobile phones
The use of hand-held mobile phones while driving is prohibited in all EU states.
Overall, all countries tolerate the use of “hands
-
free” kits but some strongly advise
against it, such as Greece.
13
The points license
Most countries use the points license system: this involves the withdrawal of points or
the addition of penalty points (Germany, Greece, Slovenia, Cyprus, Ireland, Hungary and
Denmark) in the event of violations. The initial points level or the number of penalty points
varies from country to country. Most countries start with an initial points level between 12 and
20 points, with the exception of Bulgaria, which provides for 40. Regarding penalty points, the
range is from 18 (Germany) to 3 (Denmark). Since June 1, 2016, Portugal has adopted the
points-based license system (in the form of points withdrawals).
Other countries should adopt it soon, such as Sweden. Contrary to what was expected,
the question of the points license in Belgium has been postponed to 2021.