Se conformer ou non à une mesure de poli que publique de sécurité rou ère : Comment expliquer ou prédire Tome 1 : Note de synthèse HABILITATION A DIRIGER DES RECHERCHES Men on : Psychologie Sociale
Résumé
The aim of this work is to idenƟfy key factors and propose a model to explain or
predict compliance with a public road safety policy. It is not so much a quesƟon of
knowing whether a measure is "acceptable", "acceptability" which we define in this
work by the fact of "being or not being favourable to" but which will have to be
defined more precisely in future work, but of knowing the reasons which will lead
users to comply or not with a given measure. Such an approach therefore takes into
account both incenƟves and coercive measures (penalƟes, etc.)..
Beyond that, it is a quesƟon of idenƟfying the points of vigilance to be taken
into account when studying compliance with a public road safety policy measure
(PRSP). This type of approach has been widely developed in the field of intelligent
transport systems (ITS) and the quesƟon arises as to whether it is relevant for
assessing compliance with a PRSP measure. The aim of this work is to provide
scienƟfic knowledge on a subject that has not yet been studied to any great extent,
and to respond to a strong demand from public authoriƟes, by ulƟmately offering a
decision-making tool.
Unlike the acceptability of STIs, where research is based on several
models, the acceptability of PRSPs remains liƩle studied and no theoreƟcal model
has yet been proposed.
In current research on the acceptability of PRSPs, researchers focus on a
parƟcular public policy and try to idenƟfy the explanatory and predicƟve factors,
parƟcularly in terms of reliability, equity, effecƟveness, social norms and
infringement of freedom. However, they do not aƩempt to model the interacƟons
that may exist between these factors.
Based on the exisƟng literature on the acceptability of STIs, the aim of our
work is first to define the variables that need to be taken into account, to consider
their relevance to the subject under study and to adapt them where necessary. Our
work began with the Svrai project. Its aim was to fit vehicles with data recorders
and thus gather geolocalised informaƟon on driver behaviour in terms of speed,
acceleraƟon and braking, with a view to determining the most dangerous areas of
the infrastructure. Our objecƟve was to idenƟfy the factors that explain and predict
the driving of vehicles fiƩed with these devices, both before and aŌer the users
have used them. Based on this research, we aƩempted to idenƟfy factors from the
literature on the acceptability of new technologies that would be relevant to the
study of compliance with PRSP measures, those that should be discarded and those
that should be added (Chapter 2). The reducƟon in speed on two-way roads
without a centre divider enabled us to test the factors idenƟfied in the previous
stage. Different survey waves were carried out on a representaƟve sample of the
French populaƟon at four different Ɵmes (one survey wave before and three survey
1
waves aŌer the introducƟon of the measure). The aim of this work was to make an
aƩempt at modelling, but also to make a disƟncƟon between "legiƟmacy" (Varet et
al., 2021), which is defined as "the properƟes that individuals associate with a given
traffic rule, which promotes acceptance of its implementaƟon and applicaƟon and
encourages individuals to comply with the resulƟng prescripƟons" (p.14), and
"acceptability" (chapter 3).
Various theories used in social psychology provide a more reliable model.
For example, some studies have shown a strong link between social representaƟons
and the acceptability of speed governors and speed limiters (L'Heureux, 2009),
while others have used stress models to study coping or avoidance strategies
(Hamelin and EyssarƟer, 2014).
Another objecƟve of our work has been to consider the contribuƟon of other
theories from social psychology, such as the theory of social representaƟons
(Chapter 4) or stress models (Chapter 5), in order to see to what extent they can
improve the predicƟvity of the models and above all to understand the phenomena
involved.
The PRSP measure on which we have based our thinking is automaƟc speed
radars and their impact on police officers, who are in charge of implemenƟng them,
and professional road users, who are the intended beneficiaries of the measure. As
far as professional road users are concerned, thanks to driver assistance tools, the
new driving pracƟces are unusual but consistent with their social representaƟon.
We assume that this is a gradual transformaƟon (Flament, 2001; Moliner, 2001) of
the social representaƟon of the profession. On the other hand, as far as the police
are concerned, the pracƟces associated with the ETED run counter to the social
representaƟon of the profession on several points, such as contact with users,
maintaining public order and crime prevenƟon. However, no change in
representaƟon was observed. There are several possible explanaƟons for this.
Firstly, the pracƟces generated by the tool are offset by other behaviours consistent
with the profession (such as the use of 'binocular' cameras) which may
counterbalance and prevent a change in the social representaƟon of the profession.
Another interpretaƟon could be linked to the perceived reversibility of the
situaƟon. For example, the automaƟc camera is imposed by a hierarchical
organisaƟon that has the power to free up police officers by entrusƟng its use to
others, such as agents from the Ministry of Ecology. However, this is not the
message the officers received from their superiors, who confirmed to us that this
opƟon had been ruled out.
The stress models used provide a relevant framework for understanding the
impact of speed cameras on the work tasks and movements of those responsible
for installing and implemenƟng them, as well as those for whom they are intended.
They help to understand the constraints weighing on the professional acƟvity and
2
also the skills acquired or the adaptaƟon strategies put in place to reduce them.
The results are counter-intuiƟve. It is the professionals responsible for the
operaƟon of the system and therefore, in France, the representaƟves of the public
authoriƟes, who seem to be having the most difficulty with the deployment of
automated systems. The explanaƟons given are the amplifying role of the
constraints generated by the interviewees' social environment, the threat to their
professional idenƟty and the limited room for manoeuvre they have to adapt to the
system. On the other hand, mobility professionals, those whose work requires
them to travel by road and who have a high probability of losing driving licence
points and receiving fines, appear to be much less affected by the system. Among
the explanaƟons put forward are the considerable autonomy these professionals
have in planning and carrying out their journeys, a social environment with few
constraints and, above all, the development of adaptaƟon strategies that are
parƟcularly effecƟve in reducing the stress generated by speed cameras. It is the
on-board technological equipment with which the interviewees have equipped
their cars that enables them to adapt effecƟvely to automaƟc speed cameras.
In the literature on PRSPs, the quesƟon researchers try to answer is: is this
measure acceptable or not? However, the ulƟmate aim of any
acceptability/acceptance study is to actually change pracƟces in the expected
direcƟon. What's more, the modificaƟon of pracƟces is not governed by the law of
all or nothing ("I SYSTEMATICALLY or NEVER comply with the measure") (chapter 6).
Indeed, while the ulƟmate aim of any public policy is to change pracƟces, the
theory of condiƟonality (Gaymard, 2007) and even work on perverse norms (Perez
et al, 2002) show that, for road users, certain driving situaƟons legiƟmise noncompliance with the legal rules of the road (e.g. not complying with speed limits in
a straight line). Thus, the fact that road users support a public policy does not
necessarily mean that they will systemaƟcally comply with it. Depending on the
driving context (weather condiƟons, whether or not there are passengers in the
vehicle, whether or not they are in a hurry, etc.), users may or may not comply with
a PRSP measure. In addiƟon, the introducƟon of a measure will have an impact on
the driving behaviour associated with it. For example, we have shown in the
experiment aimed at regulaƟng traffic between lanes that this measure affected the
control of rear-view mirrors or the wheel posiƟon of light vehicle (LV) users, or led
to new posiƟve or negaƟve behaviours in other areas, for example the reducƟon in
speed on two-way roads without a central separator led to an increase in vandalism
of automaƟc speed cameras.
In the same spirit, a public policy is generally implemented independently
but also evaluated independently of other public policies. We assume that the
introducƟon of a new public policy is likely to have an impact on the acceptability of
exisƟng PRSPs. For example, the introducƟon of the V80 in rural areas may have
3
had a collateral impact on the acceptability of speed in urban areas. On a different
note, research currently being conducted by M-A Granié is based on the hypothesis
that measures to curb the COVID 19 pandemic (a public health measure) have had
an impact on compliance with road safety rules (and therefore probably on their
acceptability). However, the quesƟon remains even for exisƟng PRSPs: PRSPs are
evaluated independently of each other. We assume that there is an interacƟon
effect between PRSPs, and this is precisely what we propose to study in order to fill
the theoreƟcal void on this subject.
As menƟoned earlier, PRSPs are studied in isolaƟon as if no other element
could have an impact on them. SomeƟmes, however, they receive so much media
coverage and provoke such strong reacƟons that we wonder about the effect
observed on pracƟces; is it really caused by the PRSP studied? It is indeed possible
that other factors may have had an impact on the results observed (for example,
the yellow waistcoat movement in the V80 study). Here again, the literature is nonexistent.
With regard to the model we propose (Chapter 7), in line with the work of
Varet et al. (2021), we believe that the variables relaƟng to the impact of the
measure on user behaviour are mediaƟng variables in the relaƟonship between
legiƟmacy, intra-individual variables, variables unrelated to the targeted behaviour,
variables related to the targeted behaviour and behavioural intenƟon. However, we
have some reservaƟons about the link between legiƟmacy and self-reported
behaviour, which could be mediated solely by affecƟve aƫtude and not by all the
variables referring to the direct link between the individual and the death.
Behavioural intenƟon will itself have an effect on self-reported behaviours.
FacilitaƟng condiƟons will have an impact on self-reported behaviour and not on
behavioural intenƟon. The percepƟon of the sancƟon will have a negaƟve effect on
internalisaƟon but may have an effect on behavioural intenƟon, which in turn will
have an effect on behaviour. Indeed, it is possible to comply with a rule without
internalising it, due to perceived external pressures (Deci and Ryan, 2008). The
social context and demographic variables will play a moderaƟng role. We also
assume that intra-individual variables will have an effect both on psychological
variables related to the job and on those related to the measure.
With regard to the model for professionals (Chapter 7), we believe that the
variables relaƟng to the impact of the measure on the respondent's behaviour
mediate the relaƟonship between legiƟmacy, intra-individual variables, workrelated variables and behavioural intenƟon. Behavioural intenƟon will itself have an
effect on self-reported behaviour. FacilitaƟng condiƟons will have an impact on selfreported behaviour and not on behavioural intenƟon. As with the user model, we
believe that the link between self-reported behaviour and legiƟmacy could be
mediated by affecƟve aƫtude or by the set of variables relaƟng to the impact of the
measure on the respondent's behaviour. However, this point will need to be tested
4
in future research. We also assume that the intra-individual variables will have an
effect both on the psychological variables related to the job and on those related to
the measure.
The aim of this work is to propose a model of compliance with PRSP
measures essenƟally for road users, in other words the final recipients of the
measure, the work for which the work is most accomplished, but also a model for
those in charge of its implementaƟon.
The final chapter (Chapter 8) deals with future research. The first objecƟve
is to develop a model of the acceptability-acceptance of public policies. The model
we presented in the previous chapter is a first draŌ that needs to be tested and
most certainly improved. The aim would therefore be to test this model on different
road safety policy measures in order to see the interacƟons between the different
factors and then to think about a possible improvement in its effecƟveness in
explaining compliance with a PRSP measure. The other point in this area will be to
study the evoluƟon of a policy over a relaƟvely long period of Ɵme, i.e. several
years. Generally speaking, studies focus on an evaluaƟon, at best, post and ante,
i.e. in 2 phases. However, the idea would be to study this development over a
longer period and the possible interacƟons with other PRSP measures.
It will also be a quesƟon of conƟnuing this work aimed at establishing the
link between the acceptability of ITS and the acceptability of RH public policy
measures. Indeed, the introducƟon of an RSPP measure by agents may seem like
the introducƟon of a new tool, but if this is not the case, it may at least have an
impact on the professional tasks to be carried out. In addiƟon, we believe that
further reading of the literature on work and organisaƟonal psychology is needed
to gain a beƩer understanding of the factors to be taken into account in the
implementaƟon of this measure.
Origine : Fichiers produits par l'(les) auteur(s)