ReviewLiteratureTrafficCalming en

National Collaborating Centre
for Healthy Public Policy

www.ncchpp.ca
URBAN TRAFFIC CALMING

AND HEALTH
LITERATURE REVIEW | NOVEMBER 2011
National Collaborating Centre
for Healthy Public Policy
www.ncchpp.ca
URBAN TRAFFIC CALMING

AND HEALTH
LITERATURE REVIEW | NOVEMBER 2011
AUTHORS
Olivier Bellefleur
National Collaborating Centre for Healthy Public Policy
François Gagnon
LAYOUT
Isabelle Hémon
Production of this document has been made possible through a financial contribution from the
Public Health Agency of Canada through funding for the National Collaborating Centre for
Healthy Public Policy (NCCHPP).
The National Collaborating Centre for Healthy Public Policy is hosted by the Institut national de
santé publique du Québec (INSPQ), a leading centre in public health in Canada.
The views expressed herein do not necessarily represent the views of the Public Health Agency
of Canada.
All Images in this document have been reproduced with permission or in accordance with
licences authorizing their reproduction. Should you discover any errors or omissions, please
advise us at ncchpp@inspq.qc.ca.
This document is available in its entirety in electronic format (PDF) on the Institut national de santé publique du
Québec website at: www.inspq.qc.ca/english and on the National Collaborating Centre for Healthy Public Policy
website at: www.ncchpp.ca.
La version française est disponible sur les sites Web du Centre de collaboration nationale sur les politiques
publiques et la santé (CCNPPS) au www.ccnpps.ca et de l’Institut national de santé publique du Québec
au www.inspq.qc.ca.
Reproductions for private study or research purposes are authorized by virtue of Article 29 of the Copyright Act.
Any other use must be authorized by the Government of Québec, which holds the exclusive intellectual property
rights for this document. Authorization may be obtained by submitting a request to the central clearing house of
the Service de la gestion des droits d’auteur of Les Publications du Québec, using the online form
at: www.droitauteur.gouv.qc.ca/en/autorisation.php or by sending an e-mail to droit.auteur@cspq.gouv.qc.ca.
Information contained in the document may be cited provided that the source is mentioned.
e
LEGAL DEPOSIT – 2 QUARTER 2012
BIBLIOTHÈQUE ET ARCHIVES NATIONALES DU QUÉBEC
LIBRARY AND ARCHIVES CANADA
ISBN: 978-2-550-64358-6 (FRENCH PRINTED VERSION)
ISBN: 978-2-550-64359-3 (FRENCH PDF)
ISBN: 978-2-550-64360-9 (PRINTED VERSION)
ISBN: 978-2-550-64359-3 (PDF)
© Gouvernement du Québec (2012)
Urban Traffic Calming and Health
ACKNOWLEDGMENTS
The authors would like to thank the following for their precious comments and suggestions
on earlier drafts of this document:
• Erna van Balen, National Collaborating Centre for Environmental Health;
• Lucie Lapierre, Québec en forme;
• Patrick Morency, Direction de santé publique of the Agence de la santé et des services
sociaux de Montréal;
• Florence Morestin, National Collaborating Centre for Healthy Public Policy;
• Stéphane Perron, Direction de santé publique of the Agence de la santé et des services
sociaux de Montréal;
• Diane Sergerie, Institut national de santé publique du Québec.
The authors would also like to thank Florence Morestin for her support and guidance in the
area of knowledge translation.
National Collaborating Centre for Healthy Public Policy I

Institut national de santé publique du Québec
ABOUT THE NATIONAL COLLABORATING

CENTRE FOR HEALTHY PUBLIC POLICY
The National Collaborating Centre for Healthy Public Policy (NCCHPP) seeks to increase the
expertise of public health actors across Canada in healthy public policy through the
development, sharing and use of knowledge. The NCCHPP is one of six Centres financed by
the Public Health Agency of Canada. The six Centres form a network across Canada, each
hosted by a different institution and each focusing on a specific topic linked to public health.
In addition to the Centres’ individual contributions, the network of Collaborating Centres
provides focal points for the exchange and common production of knowledge relating to
these topics.
National Collaborating Centre for Healthy Public Policy III

TABLE OF CONTENTS
LIST OF TABLES ............................................................................................................... VII
LIST OF FIGURES............................................................................................................... IX
LIST OF BOXES .................................................................................................................. XI
1 INTRODUCTION.............................................................................................................1
1.1 Definition of the object............................................................................................1
1.2 Intervention logic ....................................................................................................5
1.3 Working proposal and organization of the literature review ....................................7
2 METHOD ........................................................................................................................9
2.1 Documentary sources and research methods ........................................................9
2.2 Inclusion and exclusion criteria ............................................................................10
2.3 Selected documents.............................................................................................12
2.4 Perusal and extraction of evaluative data .............................................................12
2.5 Classification of evaluative studies .......................................................................12
3 SYNTHESIS OF DATA DRAWN FROM THE LITERATURE ........................................ 15
3.1 Number and severity of collisions .........................................................................15
3.1.1 Mechanisms of action ..............................................................................15
3.1.2 Effects of isolated traffic-calming measures............................................. 18
3.1.3 Effects of area-wide calming schemes ....................................................24
3.1.4 Conclusion of the section on the number and seriousness of
collisions .................................................................................................34
3.2 Air quality .............................................................................................................36
3.2.4 Conclusion of the section on air quality ...................................................50
3.3 Environmental noise.............................................................................................52
3.3.4 Conclusion of the section on noise pollution ............................................ 62
3.4 Active transportation ............................................................................................63
3.4.4 Conclusion of the section on active transportation ................................... 76
4 CONCLUSION OF THE LITERATURE REVIEW ..........................................................79
4.1 Revisiting the working proposal ............................................................................79
4.2 Contributions and limitations of the literature review............................................. 80
4.3 Research gaps .....................................................................................................81
National Collaborating Centre for Healthy Public Policy V

BIBLIOGRAPHY ................................................................................................................. 85
APPENDIX 1 INTERVENTION LOGIC OF THE TWO APPROACHES TO TRAFFIC
CALMING ................................................................................................... 95
APPENDIX 2 GLOSSARY OF TRAFFIC-CALMING MEASURES ................................... 99
APPENDIX 3 SUMMARY TABLES OF EVALUATIVE STUDIES ................................... 111
VI National Collaborating Centre for Healthy Public Policy

LIST OF TABLES
Table 1 Principal characteristics of the two approaches to traffic calming .....................5
Table 2 Effectiveness of roundabouts.........................................................................20
Table 3 Effectiveness of road diets .............................................................................21
Table 4 Comparison of the effectiveness of three types of calming measures ............ 22
Table 5 Effectiveness of area-wide strategies according to collision site .................... 26
Table 6 Effectiveness of area-wide strategies ............................................................26
Table 7 Estimated effectiveness based on the conflicts technique of a mini-
roundabout scheme .......................................................................................29
Table 8 Effectiveness of 20-mph (32-km/h) zones at reducing personal injury
collisions occurring within them......................................................................30
Table 9 Effects of the number and distribution of calming measures on
inequalities in road injuries among child pedestrians ..................................... 32
Table 10 Effects of various calming measures on air emissions ................................... 39
Table 11 Effects of calming measures on air emissions according to type of
vehicle ...........................................................................................................40
Table 12 Effects of speed humps on air emissions .......................................................41
Table 13 Effects of calming measures on air emissions ............................................... 42
Table 14 Effects of various area-wide strategies on air emissions ................................ 44
Table 15 Effects of the Leigh Park traffic-calming scheme on air emissions ................. 46
Table 16 Effects on air emissions of a scheme involving installation of 21 mini-
roundabouts ..................................................................................................47
Table 17 Effects of calming schemes on air quality as perceived by residents ............. 49
Table 18 Effects of calming measures on the maximum noise level of vehicles
moving at constant speeds ............................................................................56
Table 19 Effects of calming measures on vehicle speeds and environmental noise ..... 59
Table 20 Effects of calming schemes on environmental noise as perceived by
residents ........................................................................................................60
Table 21 Overall increase in the number of pedestrians following calming of a
road ...............................................................................................................67
Table 22 Self-reported changes in the travel behaviour of residents of a calmed
road ...............................................................................................................68
Table 23 Residents’ perception of safety following calming of their street .................... 69
Table 24 Modal share of residents’ trips, by destination, before (n=151) and after
(n=150) implementation of an area-wide calming strategy ............................. 72
Table 25 Effects of an area-wide strategy on the safety of pedestrians and cyclists
as perceived by residents ..............................................................................75
National Collaborating Centre for Healthy Public Policy VII

LIST OF FIGURES
Figure 1 An example of the black-spots approach .........................................................3
Figure 2 An example of the area-wide approach............................................................4
Figure 3 Classification of principal traffic-calming measures ..........................................6
Figure 4 Intervention logic of traffic calming ...................................................................6
Figure 5 Narrowing of field of vision as speed increases ............................................. 16
Figure 6 Probability that a pedestrian will die as a result of a collision with a car in
relation to the speed at impact .......................................................................16
Figure 7 Number of points of conflict (black dots on the diagrams) at a traditional
intersection and in a roundabout ....................................................................17
Figure 8 A single-lane roundabout ...............................................................................20
Figure 9 Before (left) and after (right) a road diet .........................................................21
Figure 10 An example of collision dispersion in an urban setting ................................... 25
Figure 11 City of Vancouver West-End neighbourhood traffic-calming scheme ............. 27
Figure 12 Scheme involving installation of 21 mini-roundabouts in Växjö, Sweden........ 28
Figure 13 20-mph zones and adjacent areas in London, U.K......................................... 29
Figure 14 Speed table: a measure designed for faster lanes and heavy vehicles .......... 35
Figure 15 Effect of mean travelling speed on emission levels from passenger cars
with catalysts .................................................................................................37
Figure 16 Diagram of a road with calmed intersections .................................................41
Figure 17 Leigh Park area calming scheme ...................................................................45
Figure 18 A “green” mini-roundabout .............................................................................52
Figure 19 Increase in maximum vehicle noise level according to speed for three
types of vehicles ............................................................................................53
Figure 20 Textured crosswalk ........................................................................................54
Figure 21 Estimated variations in noise levels after the installation of traffic-calming
measures according to traffic composition .....................................................57
Figure 22 The vicious circle linking active transportation to perceived danger ............... 65
Figure 23 Diagram of a road network that is more permeable to active means of
transportation than to motor vehicles .............................................................66
Figure 24 Two road narrowings configured differently ................................................... 70
National Collaborating Centre for Healthy Public Policy IX

LIST OF BOXES
Box 1 Limitations of “before-after” studies evaluating the effects of isolated
interventions on road safety ...........................................................................18
Box 2 Theoretical advantages of area-wide strategies over isolated
interventions for improving road safety in urban settings ............................... 24
Box 3 Measurements of environmental noise and a few points of reference ............ 55
Box 4 Theoretical limitations of isolated interventions for encouraging active
transportation.................................................................................................67
Box 5 Particular importance of subjective studies to understanding the effects
of traffic calming on active transportation .......................................................68
Box 6 Theoretical advantages of area-wide interventions for encouraging
active transportation and methodological limitations ...................................... 71
National Collaborating Centre for Healthy Public Policy XI

1 INTRODUCTION
1.1 DEFINITION OF THE OBJECT
This literature review examines the effects of traffic calming in urban environments on four
health determinants, namely:
• The number and severity of road collisions, 1
• Air quality,
• Environmental noise, and
• Physical activity associated with active transportation.
Traffic calming is a manner of intervening in the built environment that appears to offer
significant potential as a way to improve population health, and the evaluative literature is
sufficiently abundant to require a literature review. This review will also facilitate comparison
of the two approaches to traffic calming: the black-spots approach and the area-wide
approach.
As specified in our document introducing the concept of traffic calming, 2 there exists no
consensus as to the precise contours of the concept in question, hence the importance of
making explicit and providing the rationale for our working definition.
The concept of traffic calming as used in this literature review refers to engineering
measures (speed humps, curb extensions, etc.) and to strategies structuring their
implementation (30-km/h zones, meeting zones, etc.) 3 whose purpose is to reduce vehicle
speeds 4 and/or traffic volumes on existing public roads to achieve various goals and
objectives. These strategies are based either on a black-spots approach (where isolated
measures are installed at specific points on the road network) or on an area-wide approach
(where measures are installed in a systematic and integrated manner within a geographic
area comprising several streets).
Thus, the concept of traffic calming as used here refers to engineering measures, strategies
and approaches that, concretely speaking, vary in terms of their goals and objectives.
Regarding these measures, there is a clear lack of consensus as to what precisely is meant
by the term “engineering measures.” While everyone agrees that devices such as speed
humps or curb extensions with pedestrian crosswalks are included under this designation,
views diverge significantly concerning what is excluded. For most authors, in fact, road signs
1
The term “collision” has been chosen instead of the term “accident,” because the latter refers not only to an
impact between two bodies, but also connotes an event of a random chance occurrence, i.e. one that is of an
accidental nature. In this respect, the term “collision” seems more neutral and thus directs discussion more
toward risk factors (Stewart & Lord, 2002).
2
For more information on the origins of this working definition, please see our document entitled “Traffic
Calming: An Equivocal Concept” at: http://www.ncchpp.ca/175/publications.ccnpps?id_article=648.
3
The glossary, in Appendix 2, describes and illustrates each traffic-calming measure mentioned in this literature
review. It also lists their most common names in English and in French.
4
In urban settings, traffic calming is usually aimed at reducing the speed of motor vehicles relative to the
50-km/h norm, established in the mid-20th century.
National Collaborating Centre for Healthy Public Policy 1

are excluded, except when they serve to indicate, for example, a speed limit of 30 or 40 km/h
at the entrance to areas where speed humps or other devices of this nature have been
installed. Some definitions diverge specifically with regard to the use of stop signs. For some
authors, stop signs cannot be considered traffic-calming measures, whereas others include
them when they are installed as part of an area-wide calming strategy. Thus, road signs were
included in many of the evaluative studies considered in the present literature review. The
results associated with road signs were therefore reported, not only because they are
included within the perimeter of study of some evaluations, but also because some engineers
developing intervention strategies themselves view these as belonging to the repertory of
traffic-calming measures.
With regard to strategies, these are based on one of two approaches: the black-spots
approach, which favours interventions at one or more discrete points considered in isolation
from the road network, and the area-wide approach, which underpins interventions applied
on a more or less large geographic scale. 5
In Canada, calming strategies based on the black-spots approach predominate. They rest on
the idea that the road network is, in general, designed to be safe and that it is well-adapted to
the various uses of public roads and their associated functions, but that it contains some
isolated design flaws, that require correcting, whether through traffic-calming measures or
otherwise. Figure 1 depicts a map produced by the City of Edmonton that illustrates this
approach. The map shows the location of “black spots,” that is of places considered to be
high risk–in this case, intersections where, during 2007, 40 or more collisions occurred. 6 The
city’s authorities make use of these black-spots analyses to guide their choice of
interventions aimed at improving road safety, including those that can be considered as
traffic-calming measures. Installation of such traffic-calming measures is usually aimed at
improving safety at specific points on the road network, particularly by reducing motor vehicle
speeds.
5
This distinction is also made using the expressions “targeted approach” and “systemic approach.”
6
There is no standard or single criterion for defining a black spot–practices vary from one city to another.
2 National Collaborating Centre for Healthy Public Policy

Figure 1 An example of the black-spots

approach
This map identifies the most high-risk sites on the city of
Edmonton’s road network (40 or more collisions in 2007). The
number of collisions thus mapped is significant (2724) even
though this represents only about 10% of all the collisions that
occurred during the same year in the city.
Source: City of Edmonton, 2008, p. 29.
As regards calming strategies based on the area-wide approach, these are still rare in this
country, although some have been implemented in a few cities, in particular, Vancouver,
North Vancouver, Toronto, Ottawa and, more recently, in the Plateau-Mont-Royal borough of
Montréal. 7 Area-wide strategies usually rest on the idea that the road network design in
certain areas of varying size is systematically biased in favour of motorized traffic, to the
detriment of other modes of travel (cycling, walking, etc.) and other uses of urban space
(homes, schools, meeting areas, etc.). Area-wide strategies are thus intended to help correct
this bias.
7
A Transport Canada (2005) document identifies many more Canadian cities as having implemented area-wide
strategies than we have included here. The difference between the authors’ working definition of an “area-wide
strategy” and ours probably, to a large degree, explains this discrepancy, but the document in question does
not allow for a detailed comparison of these definitions.

To this end, the installation of calming measures on a road network is systematically

planned, in general so as to reduce driving speeds, but often also to reduce the volume of
“through traffic” on local streets. 8 This “through” traffic makes use of streets, whose main
function is residential, to travel across a given sector instead of using roads specifically
intended for this purpose, namely arteries and highways. Figure 2 depicts a map of the
largest 30-km/h zone in France that illustrates the potential geographic scope of strategies
based on the area-wide approach.
Figure 2 An example of the area-wide

approach
The largest 30-km/h zone in France covers 87 km of
residential streets in the central boroughs of Lyon (transit
roads are not affected). Installation of these calming
measures began in 2003 and is still underway.
Source: adapted from Grand Lyon / Communauté urbaine de
Lyon/France, 2007.
8
The streets in an urban road network are usually classified hierarchically, based on function, into three
categories: local streets, collector streets, and arteries. Local streets are those whose main function is to allow
access to residences on the street. Collector streets, between local streets and arteries, are those whose
purpose is to allow access to residences and businesses as well as to collect through traffic and direct it to
arteries. As for arteries, their main function is to allow the passage of through traffic.

The present literature review is organized so as to allow for comparison of the effects, in an
urban setting, of strategies based on the two approaches to traffic calming. 9 Table 1
summarizes the principal characteristics of these approaches.
Table 1 Principal characteristics of the two approaches to traffic calming
Black-spots approach Area-wide approach
Systematic bias of roads and of the public road

Conception of the
Isolated design flaws on certain network in favour of motorized traffic, to the
problems to be
public roads. detriment of other modes of travel and other
addressed
uses of adjacent areas.
Specific points on the road
More or less extensive geographic areas
Scale of interventions network (intersection, street block
comprising several streets.
or, sometimes, an entire street).
Improvement of road safety and, more broadly,

Mainly the improvement of road
Goals of the living environment (noise, air quality,
safety.
active transportation, green space, etc.).
Integrated series of calming measures aimed
Calming measures principally
at reducing motor vehicle speeds and traffic
aimed at reducing motor vehicle
Means volumes (in the latter case, usually by
speeds and, more rarely, traffic
channeling traffic toward arterial networks or
volumes.
highways).
1.2 INTERVENTION LOGIC
The effects of traffic-calming measures and strategies on the four health determinants
considered in this review (number and severity of collisions, air quality, environmental noise,
and active transportation) are usually conceived of as being linked–directly or indirectly–to
the two main mechanisms of action associated with such measures, namely, the reduction of
driving speeds and of traffic volumes at intervention sites. Moreover, some authors (Ewing,
1999; Transportation Association of Canada [TAC] & Canadian Institute of Transportation
Engineers [CITE], 1998) use the two main mechanisms of action to classify traffic-calming
measures into two broad categories: those used to reduce driving speeds and those used to
reduce traffic volumes. These two categories are often further divided into sub-categories.
Calming measures used mainly to reduce speeds are classified into three sub-categories:
vertical deflections (e.g., speed humps) which, as the name indicates, vertically deflect
vehicles; horizontal deflections (e.g., chicanes) which laterally deflect vehicles; and road
narrowings (e.g., chokers). Calming measures used to reduce traffic volumes are divided into
two sub-categories: those that reduce road capacity (e.g., reduction in the number of lanes)
and those that alter street connectivity (e.g., diagonal diverters). Figure 3 summarizes this
classification scheme.
9
To learn more about the political contexts within which traffic-calming interventions are developed, refer to our
document entitled “Traffic Calming: Political Dimensions” at: http://www.ncchpp.ca/175/Publications.ccnpps?id
_article=670.

Traffic-
calming
measures
Reduction of Reduction of
vehicle traffic
speeds volumes
Vertical Horizontal Reduction of Alteration of

deflections (e.g., deflections (e.g., Road narrowings road capacity the road
speed humps, chicanes) (e.g., chokers) (e.g., network
speed cushions) reduction in connectivity
the number (e.g.,
of lanes) diagonal
diverter)
Figure 3 Classification of principal traffic-calming measures
Although the effectiveness of traffic-calming measures at reducing traffic speeds and

volumes may vary according to the context in which measures are implemented and the
strategies guiding their implementation, this effectiveness has been relatively well-
documented and tested (TAC & CITE, 1998; Ewing, 1999; Ewing & Brown, 2009). Therefore,
the present literature review foregoes an examination of this aspect to concentrate on the
effects of these measures on health determinants. Figure 4 graphically summarizes the
intervention logic in its simplest form. The model links the two approaches to traffic calming
to their effects on health determinants by means of their preferred mechanisms of action.
These mechanisms are presented in a manner that takes into account the fact that, by
reducing traffic speeds, one may also seek to reduce traffic volumes. The idea, then, is that
reducing speeds would make a travel route less attractive to motor vehicles than other routes
towards which one desires to channel them.
TRAFFIC-CALMING PRINCIPAL MECHANISMS EFFECTS ON HEALTH

POLICY OF ACTION DETERMINANTS
Reduction of the number and
Reduction of severity of collisions
Black-spots approach
speeds Improvement of air quality
Reduction of environmental noise

Area-wide approach Reduction of
volumes Increase in active transportation
Figure 4 Intervention logic of traffic calming

This simplified diagram presents the two main mechanisms of action as they are described in
the traffic-calming literature, thus excluding several mechanisms that more specifically target
one or several health determinants (e.g., improvement of pedestrian visibility→ reduction in
number and severity of collisions). It also excludes intermediary effects between the
mechanisms shown and the effects on health determinants (e.g., reduction of speed→
enhanced perception of safety→ increase in active transportation), as well as potential
interactions between health determinants (e.g., decrease in the number and severity of
collisions → enhanced perception of safety → increase in active t ransportation). Finally,
because the diagram represents the logic of the intervention and not its actual effects,
potentially negative effects on health determinants are not included (e.g., reduction of speeds
→ increased emission of pollutants per vehicle → deterioration of air quality). These
elements excluded from the diagram will be addressed in the sections specific to the four
health determinants. In addition, readers will find a more detailed version of this illustration in
Appendix 1.
1.3 WORKING PROPOSAL AND ORGANIZATION OF THE LITERATURE REVIEW
The literature review is organized so as to allow for comparison of the effects on four health
determinants of strategies based on the two approaches to traffic calming, implemented in an
urban setting. A priori, it seems likely that strategies based on the area-wide approach have
greater potential for improving population health than strategies based on the black-spots
approach. Indeed, strategies based on the area-wide approach favour interventions
implemented on a larger geographic scale (networks comprising more than one street), they
are often driven by broader objectives and goals (road safety and improvement of the living
environment) and they tend to involve both the main mechanisms of action associated with
calming measures (reduction of traffic speeds and volumes) rather than just one. The intent
of this literature review is to allow for an informed discussion of this theoretical advantage.
The literature review is divided into four main sections, each covering one of the
aforementioned health determinants. These sections are themselves divided into two sub-
sections, one presenting the results of studies allowing for evaluation of the effects of the
black-spots approach and the other presenting the results of studies allowing for analysis of
the effects of the area-wide approach. In addition, the results are presented so as to
distinguish between studies whose approach is objective and those whose approach is
subjective. 10 Each section begins by identifying the mechanisms of action associated with the
calming strategies specifically tied to the health determinant discussed in the section, and
concludes by summarizing the results and identifying research gaps.
10
By objective research, we mean research that develops and seeks to measure objects of study that are
considered to exist and be measurable whether or not they are perceived by particular persons. For example,
researchers have attempted to determine the influence of the built environment on modes of travel by
measuring indicators such as the number of pedestrians and cyclists, building density, the level of mixed
usage, the connectivity of road networks, the proximity of parks and other recreational areas, the number of
calming measures, etc. By subjective research, we mean research that develops and seeks to measure
objects of study whose existence is based on the perceptions of particular individuals under study. Returning,
partially, to the same example, other researchers have evaluated the relationship between the use of parks
and recreational areas and nearby residents’ perceptions of their user-friendliness (safety, cleanliness).

2 METHOD
The documentary search was carried out during May and June 2010, and scanning
continued until the end of October 2010.
2.1 DOCUMENTARY SOURCES AND RESEARCH METHODS
Given that the expression “traffic calming” has been the standard in the English literature
since 1999 (Ewing, 1999), databases were searched using this key expression. Tests carried
out in PubMed and MedLine using the expressions “traffic abatement,” “traffic mitigation,”
“traffic management,” “traffic control,” and “traffic moderation” confirmed that, for the period
covered, the articles concerning traffic calming could be located by relying on the expression
“traffic calming.” For the French literature, searches were carried out using the expressions
“apaisement de la circulation,” “modération de la circulation,” and “modération de la vitesse.”
Given the diversity of the potential effects of traffic calming on health, the searches were
carried out by simply entering the key expressions in the “title” and “keyword” fields.
The following databases were searched:

• PubMed;
• The Transportation Research Information Services (TRIS) database;
• Scopus;
• Scirus;
• IngentaConnect;
• Repère;
• Using the CSA Illumina platform (ERIC, Health and Safety Science Abstracts, PILOTS
Database, CSA Sociological Services Abstracts, CSA Sociological Abstracts, CSA
Worldwide Political Science Abstracts, Recent References Related to the Social
Sciences);
• Using the EBSCOhost platform (Medline, PsycINFO, SocIndex, E-Journals, CINAHL);
• Using the OvidSP platform (Ovid MedLine (R), EMBASE, EBM Reviews–Cochrane
Database of Systematic Reviews, EBM Reviews–ACP Journal Club, EBM Reviews–
Database of Abstracts of Reviews of Effects, EBM Reviews–Cochrane Central Register of
Controlled Trials, EBM Reviews–Cochrane Methodology Register, EBM Reviews–Health
Technology Assessment, EBM Reviews–NHS Economic Evaluation Database, Global
Health).
The Google and Google Scholar search engines were searched using the French and
English key expressions, to extend the scope of the search. Several websites were also
consulted, including in particular:
• Transport Canada (Canada), http://www.tc.gc.ca/eng/programs/environment-utsp-
trafficcalming-1172.htm;

• Transportation Demand Management Encyclopedia, Victoria Transport Policy Institute

(Canada), http://www.vtpi.org/tdm/tdm4.htm;
• Conseil régional de l’environnement de Montréal (Canada), http://www.cremtl.qc.ca/index.
php?id=648;
• The City of Ottawa (Canada), http://www.ottawa.ca/residents/onthemove/driving/traffic/at
m/appendices/appendix_c_en.html;
• The City of Toronto (Canada), http://www.toronto.ca/transportation/traffic/traffic_calming.ht
m;
• The City of North Vancouver (Canada), http://www.cnv.org//server.aspx?c=3&i=275;
• Institute of Transportation Engineers (United States), http://www.ite.org/traffic/;
• U.S. Department of Transportation, Federal Highway Administration (United
States), http://www.fhwa.dot.gov/environment/tcalm/;
• Pedestrian and Bicycle Information Center (United States), http://www.walkinginfo.org/ind
ex.cfm;
• Transport Research Laboratory (England), http://www.trl.co.uk/default.htm;
• Centre d’études sur les réseaux, les transports, l’urbanisme et les constructions publiques
(France), http://www.certu.fr/;
• SWOV Institute or Road Safety Research (Netherlands), http://www.swov.nl/index_uk.htm.
Documents were also located through “snowballing,” a method using bibliographies as the
means to find other pertinent resources.
2.2 INCLUSION AND EXCLUSION CRITERIA
Content of documents
To be selected for purposes of analysis, documents had to evaluate at least one effect on a
determinant of population health caused by: 1) at least one intervention aimed at calming
traffic on existing streets in an urban setting or; 2) traffic-calming measures installed on test
tracks for evaluation purposes. Interventions in rural settings were excluded, because they
most often involve installations aimed at slowing down drivers travelling through inhabited
zones on roads that, outside of these zones, are designed for high speeds–a situation that
applies imperfectly, if at all, to urban settings. Following a preliminary reading of the
documents, it became evident that the studies centred around four health determinants,
namely, the number and severity of collisions, air quality, environmental noise, and active
transportation. The documentary search then continued, with only documents focused on at
least one of these four determinants. Documents in which methodologies for evaluating an
effect were developed without that effect being measured and those focused on the
economic costs of traffic-calming strategies and measures were excluded.

Countries studied
In addition to studies examining Canadian experiences, studies focused on the United
States, European countries, and Australia were included, because these countries were
judged to be sufficiently similar to Canada.
Period considered
The period extending from January 2000 to May 2010 was considered during the
documentary search of databases. Scanning was subsequently carried out until the end of
October 2010. The search was broadened using “snowballing.” Some indispensable
documents on little-documented aspects were added as a result, even though they were not
published within the period initially considered.
Languages
The documents selected were written in English or French.
Methodological quality of studies

Studies using a simple before-and-after design to evaluate the effect of calming measures by
comparing the number and severity of collisions at specific points in time were not retained,
because these studies were shown to be marred by significant bias (see section 3.1.2.1,
Box 1). No other systematic exclusion based on methodological criteria was carried out.
Studies based on both objective and subjective epistemological approaches were included in
this literature review. Moreover, there were no exclusions on the basis of method of inquiry,
data collection technique or analytical procedures employed. 11
In the case where the methodology used in a particular study was likely to affect its validity
(conceptual, internal, external) or the reliability of its results, comments to this effect were
included where the study was presented. In this regard, several of the studies included in this
literature review do not indicate whether tests of statistical significance were carried out. This
absence of this statistical information does not imply that relevant statistical analyses were
not carried out, but neither does it allow us to conclude that they were. To draw the reader’s
attention to this fact, we have mentioned it in the text and in the summary tables where these
studies are presented (Appendix 3). Overall evaluations of the methodological strengths and
weaknesses of the studies selected and of their implications for the robustness of results are,
for their part, integrated into the various section summaries and discussed in the Conclusion.
11
A single before-after type study (Lindenmann, 2005) focused on the effectiveness of 30-km/h zones installed in
Switzerland at reducing the frequency and severity of collisions was excluded from the literature review for
reasons relating to conceptual clarity. The severity of collisions, for example, is expressed therein as a function
of the number of “accident victims,” but this term is not defined. Moreover, the results concerning collisions are
classified according to three collision occurrence patterns (case 1, case 2, case 3), for which no explanations
are given. This exclusion should not affect the results of this review, since the conclusions the author of the
study draws from his data are in line with those presented in this literature review (that is, a reduction in the
number and severity of collisions after areas are calmed).

2.3 SELECTED DOCUMENTS
Following verification of their conformity to the inclusion and exclusion criteria, the
documents selected for the purpose of analyzing the results of the evaluative research
included 19 drawn from the scientific literature (peer-reviewed articles published in scientific
journals) and 10 drawn from the grey literature (reports and conference presentations).
To determine the effects of the different health determinants on population health, the
mechanisms of action of the calming measures, research gaps, and methodological
problems, we broadened the documentary search and extended the inclusion and exclusion
criteria. Studies that were not directly focused on calming measures, but on their
mechanisms of action, for example, were thus included, along with studies focused on
methodologies or traditional literature reviews identifying research gaps. In this manner,
36 documents drawn from the scientific literature and 38 documents drawn from the grey
literature were added to the literature review.
To allow readers to distinguish between studies drawn from the scientific literature and those
drawn from the grey literature, the term “article” was reserved for studies drawn from the
scientific literature, whereas the terms “report” and “conference presentation” were
systematically reserved for the grey literature. The term “study” for its part, was used
generically to refer to either.
2.4 PERUSAL AND EXTRACTION OF EVALUATIVE DATA
All of the documents retained for the purpose of analyzing the results of the evaluative
research were obtained and read in their entirety by both authors of this review. They then
compiled the data in extraction tables, systematically considering the questions asked by the
studies' authors, the methodology used, the answers given to the questions, conceptual and
internal validity and the reliability of results. Differences in interpretation were resolved
through consensus. These tables are presented in Appendix 3.
2.5 CLASSIFICATION OF EVALUATIVE STUDIES
The results of the evaluative studies were classified according to the two approaches to
traffic calming and the four health determinants covered in this literature review. Separating
studies examining area-wide strategies from those examining black-spots strategies is an
exercise that presents challenges. In reality, although traffic-calming interventions are quite
diverse (types of calming measures, combinations, number, density, extent of geographic
area calmed, etc.), they are rarely precisely described in evaluative studies and the logic
having governed the installation of calming measures (black-spots approach, area-wide
approach) is still less often explained.
When the intervention logic is not made explicit, it is standard practice in the literature on
traffic calming in general, and in that evaluating collisions, injury and death in particular, to
separate evaluative studies on the basis of the geographic extent of the interventions
considered (at one spot, along a line, for example, a street block or an entire street, or within
a geographic area, that is, a network of more than one street). Studies evaluating the effects

of a series of calming measures installed in a geographic area comprising several streets are
used to provide information about the effects of calming strategies based on the area-wide
approach. As for studies evaluating the effects of isolated calming measures installed at
specific points on the road network, they are undertaken to evaluate either the effects of one
calming measure (e.g., those of speed humps), of one type of calming measure (e.g., those
of vertical deflections), or of the black-spots approach in general, since isolated interventions
are usually the product of strategies based on this approach.
The classification of interventions involving several calming measures installed on a street

block, or even on an entire street, is more controversial: they are sometimes included in the
first category (Bunn et al., 2003; 2009) and sometimes in the second (Elvik, 2001; Boulter &
Webster, 1997).
This classification of studies according to the geographic extent of the interventions

examined has its limitations within the context of this literature review, for it presupposes that
the geographic extent of the interventions evaluated is indicative of the logic that governed
the planning and installation of the calming measures evaluated. It is possible, for example,
for a calming measure’s effectiveness at reducing collisions, injuries and deaths to be
evaluated at a specific point on a road network, even though the measure is part of a
configuration of calming measures that was planned for and installed on a network
comprising several streets. However, despite its limitations, the geographic extent of the
interventions evaluated by a study remains the best criterion for classifying it, when the
intervention logic is not specified.
In this literature review, evaluative studies were thus classified according to the intervention
logic (black-spots/area-wide), when this was specified, and according to the geographic
extent of the interventions evaluated, when the logic was not made explicit. Since
interventions on a street block or an entire street are not generally designed to reduce traffic
volume by redirecting through traffic toward arteries and highways, they are thus most similar
to isolated interventions installed at specific points on the road network, intended mainly to
reduce driving speeds. Moreover, some discrete calming measures (chicanes, road diets,
etc.) can only be installed on street blocks or along entire streets. For these reasons, we
opted for the classification system adopted by Elvik (2001) and by Boulter & Webster (1997).
We thus classified these studies alongside those focused on isolated interventions at specific
points on the road network, so as to glean from them information about the effects of
strategies based on the black-spots approach. Studies focused on interventions installed in a
geographic area comprising several streets were, for their part, used to enhance
understanding of the effects of strategies based on the area-wide approach. To allow readers
to classify these studies differently, the tables summarizing the evaluative studies included in
this literature review are presented in Appendix 3. Interventions involving several calming
measures installed on a street block or on an entire street are there presented as a separate
category.

3 SYNTHESIS OF DATA DRAWN FROM THE LITERATURE

3.1 NUMBER AND SEVERITY OF COLLISIONS
Motorized traffic is responsible for a significant number of injuries and deaths occurring when
collisions take place or in their aftermath (World Health Organization [WHO], 2009; Canadian
Institute for Health Information [CIHI], 2010). In Canada, for 2006, the number of deaths 12
caused by road collisions was evaluated at 2889 and the number of injured at 199,337
(WHO, 2009). Collisions involving motor vehicles were also responsible for 42% of
hospitalizations due to serious injury in Canada for the 2007-2008 period (WHO, 2007; CIHI,
2010). Interventions aimed at improving the road safety record are usually classified into four
categories depending on whether their aim is to modify individual behaviour (e.g., campaigns
to raise awareness about the dangers of driving while fatigued), construction norms for
vehicles (e.g., air bags), the physical environment (e.g., traffic-calming measures) or the
social and political environment (e.g., criminalization of impaired driving) (Haddon, 1980).
Interventions targeting the physical environment, such as traffic-calming strategies, have the
advantage of not depending on the presence of the police force to be effective. In addition,
they have the potential to improve the safety of all road users. Finally, since the lowest-
income population segments and those with the lowest socioeconomic status (SES) are
often overrepresented in road fatalities and injuries in urban settings (WHO, 2004; Laflamme,
Hasselberg, & Burrows, 2010; Cubbin & Smith, 2002), these interventions can help to reduce
health inequalities (Jones, Lyons, John, & Palmer, 2005; Morency & Cloutier, 2005; Grundy,
Steinbach, Edwards, Wilkinson, & Green, 2008a). These are the main reasons that traffic-
calming strategies–in conjunction with other interventions–are being promoted.
3.1.1 Mechanisms of action
According to the studies consulted, traffic-calming strategies can influence road safety
results by means of six main mechanisms:
a) Reduction of vehicle speed
Research shows that the number and severity of collisions increases with speed (Sergerie et
al., 2005). A report even suggests that each decrease of 1.6 km/h in an urban setting results
in a 3 to 6% decrease in collisions, depending on how major a road is being considered
[Taylor, Lynam, & Baruya, 2000]). As Figure 5 illustrates, increasing speed decreases a
driver’s field of vision, which goes from slightly wider than 150 degrees at very slow speeds
to around 75 degrees at 100 km/h, thus reducing the likelihood that a dangerous situation will
be noticed in time. Moreover, increasing speed increases stopping distance, that is, the
distance travelled by the vehicle during the time it takes a driver to react plus the vehicle’s
braking time, which accordingly reduces the likelihood that the vehicle will stop in time to
avoid a collision or that it will have slowed down enough to avoid a serious collision. For
example, a car travelling at 30 km/h on dry pavement, and whose driver takes two seconds
to react, will stop after travelling a distance of a little over 20 metres. At 50 km/h, the same
car would have travelled twice that distance, a little over 40 metres (Bureau de prévention
12
The WHO defines a road fatality as a death caused by a road collision, occurring either at the time of collision
or in the subsequent thirty days (WHO, 2004).

des accidents, 2008a). The seriousness of collisions also increases with speed, especially
when vulnerable road users, such as pedestrians and cyclists, are involved. As Figure 6
shows, for every ten pedestrians struck by a car, nine will survive when the speed at impact
is 30 km/h, but only three will survive at 50 km/h, and only one will survive at 60 km/h
(Bureau de prévention des accidents, 2008b). By aiming to reduce driving speeds (often to
about 30 km/h), and particularly those of the fastest drivers (Transportation Demand
Management [TDM] Encyclopedia, 2010a), traffic-calming strategies should help reduce the
number and severity of collisions.
Figure 5 Narrowing of field of vision as

speed increases
Source: Société de l’assurance automobile du Québec, 2011.
Figure 6 Probability that a pedestrian will die

as a result of a collision with a car
in relation to the speed at impact
Source: Adapted from Bureau de prévention des accidents, 2008b.

b) Speed homogenization
The presence of speed differentials in the traffic flow increases the probability of collisions
(Ewing & Edwards, 2009; Ewing, 2000). Yet, traffic-calming strategies can cause speed
differentials, for example, when drivers slow down to go over a speed hump and accelerate
afterward. However, they have a tendency to more greatly reduce the speed of the fastest
drivers and, thus, to reduce speed differentials in the traffic flow. The effect of a calming
strategy on the number of collisions tied to speed differentials should therefore depend on
the configuration of traffic-calming measures.
c) Reduction of motorized traffic volume

In general, the number of road injuries in urban settings varies according to traffic volume
(Ewing, 2000). Calming strategies aimed at channelling traffic elsewhere therefore risk, all
risk factors being equal, displacing collisions from one location to another instead of reducing
their number. On the other hand, strategies that succeed in reducing traffic volume in
absolute terms (by increasing active transportation, for example) are likely to contribute to a
decrease in the number of road injuries.
d) Reduction of the number of points of conflict

Points of conflicts are locations “where the paths of two vehicles or the paths of a vehicle and
a cyclist or pedestrian cross or intersect” (Ministère des Transports du Québec, 2007). As
Figure 7 illustrates, some calming measures, such as roundabouts, reduce the number of
points of potential conflict between public road users, which can result in a reduction in the
number of collisions (Ewing, 1999).
Figure 7 Number of points of conflict (black dots on the

diagrams) at a traditional intersection and in a
roundabout
Sources: Stein, Kittelson, Newton, & Hottmann, 1992, p. 43 in Ewing, 1999, p. 111.
© 1992 and 1999 Institute of Transportation Engineers, 1627 Eye Street, NW, Suite 600,
Washington, DC 20006 USA, www.ite.org. Used by permission.

e) Improvement of visibility and reduction of exposure

Certain calming measures, such as curb extensions, can improve the visibility of users (e.g.,
pedestrians) or reduce their exposure to the risk of collision (e.g., by reducing the distance to
be travelled when crossing a street). Improving these two factors can contribute to a
reduction in the number of collisions in which they play a part (Ewing & Brown, 2009).
f) Increase driver alertness

By helping create more complex environments, calming strategies can also increase driver
alertness, potentially decreasing the number of collisions (Ewing, 2000).
The prevention of collisions, road injuries and deaths are among the main objectives of
traffic-calming strategies, which explains why a significant portion of the existing literature is
devoted to evaluating their effectiveness at doing so. In the following pages, we will first
present studies that have evaluated the effects of an isolated measure installed at a specific
point on the road network or of a series of isolated measures installed along a single street.
We will then summarize the studies evaluating area-wide strategies. 13
3.1.2 Effects of isolated traffic-calming measures
3.1.2.1 Objective studies
Most of the studies consulted that evaluated the effectiveness at improving road safety of
isolated measures installed at specific points on the road network or on a single street simply
recorded the number of collisions before and after an intervention was implemented. This
method tends to systematically overestimate the effectiveness of calming measures, for
which it has been severely criticized (see Box 1). These studies were systematically
excluded from the literature review.
BOX 1 LIMITATIONS OF “BEFORE-AFTER” STUDIES EVALUATING THE EFFECTS OF ISOLATED

INTERVENTIONS ON ROAD SAFETY
Studies with a “before-after” design evaluate the effect of interventions simply by

comparing the number of collisions before and after an intervention. These studies have
been criticized for not taking into account the following three confounding factors:
a) The underlying trend in the number of collisions
Studies that simply indicate the number of collisions before and after the installation of
calming measures at an intervention site cannot control for an underlying trend upwards or
downwards in the number of collisions. However, such trends can be caused by factors
unrelated to the intervention studied, such as changes in the method for compiling data on
collisions, in driving habits, in weather conditions, etc. (Persaud & Lyon, 2007).
13
Because the improvement of perceived safety has been identified as a mechanism of action that can influence
the decision of whether or not to opt for an active mode of transportation, the results of subjective studies
focused on safety are not presented in this section, but in the section on “active transportation.”

“Before-after” studies evaluating the effects of calming measures on road safety can thus
ascribe to calming measures a level of effectiveness that they do not possess (Elvik,
1997).
b) Regression to the mean

The installation of isolated calming measures at specific points on a road network is often a
response to a number or rate of collisions occurring at such spots that is judged to be
abnormally high. When this is the case, we say of the calming strategy implemented that it
is based on the black-spots approach, because the latter targets points on the road
network where collisions, injuries or deaths are concentrated. Evaluating the effectiveness
of such interventions simply by tallying the number of collisions before and after installation
of the measures proves problematic, because places that register an abnormally high or
low number or rate of collisions tend to regress, with time, toward an average number or
rate of collisions. Studies that do not control for this statistical phenomenon, known as
regression to the mean, can thus overestimate the effectiveness of calming measures
(Elvik, 1997). Indeed, Mountain & Fawaz (1992) refer to one of their articles (1991) in
which they recorded having observed a 62% decrease in collisions at 155 “black spots”
and a 14% increase at adjacent sites, simply by monitoring the distribution of collisions
over a two-year period, without any calming measure having been installed.
c) Collision migration
Following a targeted intervention, observers sometimes note what appears to be a
displacement of collisions, from whence the expression “collision migration” (Joly,
Bourbeau, Bergeron, & Messier, 1992). Several explanations have been proposed for this
phenomenon, including regression to the mean and diversion of traffic caused by calming
measures, which can, depending on the context, produce an increase in collisions
elsewhere on the road network (Mountain & Fawaz, 1992). If the effects of a targeted
intervention on adjacent sectors are not measured, a before-after tally of collisions at the
intervention site may interpret as an improvement in safety what, in reality, is simply a
migration of collisions (Elvik, 1997).
Having conducted a meta-analysis of 36 studies examining targeted measures, Elvik

(1997) discovered that the reported effectiveness of interventions at reducing the number
and severity of collisions was dependent on the number of confounding factors for which
studies controlled. In general, studies reporting reductions in the number of collisions in the
order of 50% to 90% did not control for any factors. The more factors a study controlled for,
the less significant the reported effectiveness, to the point where studies that
simultaneously took into account the underlying trend, regression to the mean, and
collision migration did not generally report statistically significant effects.
These limitations do not exclude the possibility that interventions targeting “black spots”
effectively reduce the number and severity of collisions at targeted sites. However, they
point to certain difficulties related to the evaluation of interventions based on this approach
and to the possible overestimation of effects reported by studies whose design is less than
robust.

On the other hand, recent studies have documented significant improvements in road safety
using the Empirical Bayes Method. This method makes it possible to take into account
changes in traffic volume, the underlying trend in the number of collisions and the effects of
regression to the mean (Persaud & Lyon, 2007). It uses data collected from control sites, the
number and severity of collisions at the intervention sites before installation of the calming
measures, and data concerning the evolution of traffic volumes to estimate, using regression
models, the number and severity of collisions that could reasonably have been expected to
occur without the installation of traffic-calming measures. The effectiveness of the
intervention is then determined by comparing the number and severity of collisions having
occurred since the intervention with those expected. Since we judge this method to be
sufficiently reliable, three articles having used it are presented in this section. Subsequently,
an article documenting an observational case-control study is also presented.
The first article (Retting, Bhagwant, Garder, & Lord, 2001) evaluated safety gains resulting
from the replacement of 24 intersections controlled by stop signs or traffic lights with as many
roundabouts, in the United States. Table 2 summarizes the results presented in this article.
Table 2 Effectiveness of roundabouts
Calming measure Personal injury Fatal and incapacitating

Collisions
collisions injury collisions
Roundabout (n=24) -38%* -76%* -89%*
Single-lane roundabout replacing
-61%* -77%* -
stop signs (n=9)
Multi-lane roundabout replacing stop
-5% - -
signs (n=6)
Single-lane roundabout replacing
-35%* -74%* -
traffic lights (n=4)
* Significant variation at, at least, p<0.05.
Source of the data: Retting et al., 2001.
Figure 8 A single-lane roundabout

Source: http://www.flickr.com. Photographer: WSDOT.

According to the authors, reduced speeds and fewer points of conflict at roundabouts can
explain the significant reduction in the number of collisions recorded and in the seriousness
of those that occur. It is interesting to note that replacement of an intersection controlled by
stop signs with a single-lane roundabout is particularly effective at reducing the number of
collisions and injuries. Since the majority of collisions occur at intersections, the article
concludes that replacing intersections controlled by traffic lights or stop signs with
roundabouts, where conditions permit, has considerable potential for reducing collisions, but
above all for reducing injuries and deaths.
An article (Stout, Pawlovich, Souleyrette, & Carriquiry, 2006) evaluates the impact of 15 road
diets in the United States, using the Empirical Bayes Method (23 years of data) and a
“before-after” design with control sites (10 years of data). This calming measure usually
consists of converting four-lane roads into three-lane roads, with one lane for traffic going in
each direction and a central lane reserved for left turns from either direction. As shown in
Figure 9, the public space recovered when road diets are implemented can be used to install
bike lanes and sidewalks or to green streets.
Figure 9 Before (left) and after (right)

a road diet
Source: Rosales, 2007, p. 2.
Table 3 summarizes the results presented in this article.
Table 3 Effectiveness of road diets
Collisions Collision rates Total

Fatal and
Calming Method of per kilometre (controlled for personal Minor injury
serious
measure evaluation (collision variations in injury collisions
collisions
density) volume) collisions
Bayes -25% -19% - - -
Road diet
“Before-after”
with control -29% -21% -34% -30% -11%
sites
Note: No statistical significance test mentioned.
Source of the data: Stout et al., 2006.

The article concludes that road diets can be used to reduce the number of collisions, injuries
and deaths, but the authors do not indicate whether these reductions are statistically
significant.
An article (Mountain, Hirst, & Maher, 2005) compares the effectiveness of 149 traffic-calming
interventions on roads with 48-km/h (30-mph) speed limits located throughout England.
These interventions were classified into three groups: the 79 interventions using speed
cameras, the 39 using vertical deflections (alone or combined with horizontal deflections or
road narrowings) and the 31 interventions using only horizontal deflections, narrowings,
speed activated signs (n=4) or markings indicating the speed limit (n=1). The final unusual
grouping is justified, according to the authors, by the similarity of the associated results.
Table 4 summarizes the key results presented in this article.
Table 4 Comparison of the effectiveness of three types of calming measures
Personal Number of personal

Average Fatal and serious
injury injury collisions avoided
Calming measures speed collisions
collisions (collision/km/year)
{CI 95%} {CI 95%}
{CI 95%} {CI 95%}
-13.5 km/h* -44%* -35%* -1.00*
Vertical deflections
{-16.6 to -10.5} {-54 to -34} {-54 to -18} {-1.4 to -0.6}
Horizontal -5.3 km/h* -29%* -14% -0.78*
deflections or
narrowings {-7.1 to -3.7} {-48 to -8} {-44 to +32} {-1.6 to -0.2}
-6.6 km/h* -22%* -11% -1.03*

Speed cameras
{-7.6 to -5.5} {-30 to -13} {-26 to +6} {-1.4 to -0.8}
* Significant variation at, at least, p <0.05.
Source of the data: Mountain et al., 2005.
Traditionally, the effectiveness of interventions is expressed in terms of a percentage

reduction in the various types of collisions. Presented thus, interventions making use of
vertical deflections are significantly more effective at reducing personal injury collisions than
those making use of speed cameras. This result seems to correspond to the significantly
more pronounced effect of vertical deflections on speed. However, if the effectiveness of
interventions is expressed in terms of the number of personal injury collisions avoided, the
significant effectiveness of the three types of calming measures is similar. The authors
attempt to explain the variation in the results associated with these two perspectives by
pointing to differences in implementation contexts. The speed cameras referred to in this
article were installed, for example, on streets with high traffic flow and at sites where more
personal injury collisions were recorded (almost double the collisions before the
interventions) than at sites where vertical deflections were installed, usually on local streets
with lower traffic flow. Thus, in addition to demonstrating the overall effectiveness of traffic-
calming measures at significantly reducing the number of collisions and personal injury
collisions, the article underlines the importance of taking into account implementation
contexts when comparing the effectiveness of different measures.

An article (Tester, Rutherford, Wald, & Rutherford, 2004) presenting an observational case-
control study evaluates the effectiveness of speed humps at reducing collisions causing
death and injury among child pedestrians under 15 years old in the city of Oakland, in the
United States. The article analyzes the admission data of hospital emergency departments
over a period of five years (1995-2000) to identify children living on local streets who were
admitted after being struck by a car while walking near their home (radius of 0.4 km). The
article excludes collisions caused by cars backing out of a driveway or a parking space. Each
of the one hundred children whose situation satisfied the selection criteria was matched with
two control children admitted on the same day to the emergency department for other
reasons, but matched in terms of gender and age and also living on a local street. The three
hundred children were classified into two categories: those living on a street block with at
least one speed hump and those living on a street block without any speed humps. Among
the control children, 23% (n=46) lived near a speed hump, whereas among case children, the
rate goes down to 14% (n=14). The matched-pairs analysis indicates that the probability of a
child living near a speed hump being injured by a car near home (radius of 0.4 km) and
admitted to hospital is significantly lower (OR14: 0.50, CI 95%: 0.27 to 0.89) than for a child
not living near this type of calming measure. When the analysis considers only those children
injured on the street block where they live (n=49), the protective effect of speed humps is
even greater (OR: 0.38, CI 95%: 0.15 to 0.90). Controlling for race and ethnicity, the
protection provided by proximity to a speed hump remains statistically significant and similar
for children injured within a radius of 0.4 km (OR: 0.47, CI 95%: 0.24 to 0.95); however,
although the point estimate still indicates a protective effect, this is no longer statistically
significant for children injured on the street block where they live (OR: 0.40, CI 95%: 0.15 to
1.06). On the basis of these results, the article concludes that speed humps make the
environment safer for child pedestrians.
3.1.2.2 Summary of the results for isolated traffic-calming measures
The documentary search led to the selection of four articles (Retting et al., 2001; Stout et al.,
2006; Mountain et al., 2005; Tester et al., 2004) belonging to the scientific literature having
evaluated the effectiveness of various calming measures (roundabouts, road diets, vertical
deflections, horizontal deflections or narrowings, and speed cameras) at reducing the
number of collisions, injuries and deaths. The authors of three articles used the Empirical
Bayes Method, whereas one article presents a “case-control” study. The results were not all
tested for statistical significance, but the majority of the results for which tests are known to
have been carried out were significant. The key points to recall are the following:
14
OR or odds ratio: An odds ratio is a statistic produced through logistic regression that expresses the
relationship between the probability that an event will occur and the probability that it will not occur. An odds
ratio can be used to determine if, all other things being equal, persons living in an environment defined by
certain characteristics are more or less likely to move around by bicycle, for example. An odds ratio of 1 would
indicate that the characteristic had no effect, whereas a ratio of less than 1 would indicate that the
characteristic in question lowers the likelihood that such persons would move around by bicycle and a ratio of
more than 1 would indicate an increase in the likelihood. The further the odds ratio is from 1, the greater the
effect on the probability of the event occurring.

a) Reduction in collisions, injuries and deaths

All of the articles report a reduction in collisions, injuries and deaths tied to the presence of
the calming measures evaluated.
b) Speeding, collisions, injuries and deaths

One article (Mountain et al., 2005) supports the assumption in the logic model of a
relationship between a reduction in vehicle speeds and a reduction in collisions, injuries and
deaths.
c) Promising calming measures

Among the calming measures evaluated, single-lane roundabouts (Retting et al., 2001) and
vertical deflections (Mountain et al., 2005), such as speed humps (Tester et al., 2004), were
shown to be particularly effective at reducing collisions, injuries and deaths.
3.1.3 Effects of area-wide calming schemes
BOX 2 THEORETICAL ADVANTAGES OF AREA-WIDE STRATEGIES OVER ISOLATED

INTERVENTIONS FOR IMPROVING ROAD SAFETY IN URBAN SETTINGS
With regard to reducing the number and severity of collisions in urban settings, it is
reasonable to assume that area-wide traffic-calming strategies have, in theory, three main
advantages over strategies that intervene at specific points or on specific segments of the
road network:
a) Collision dispersion
To begin with, as illustrated in Figure 10, collisions in urban settings are generally
dispersed throughout the road network occurring at a multitude of points each registering a
relatively small number of collisions. Given this context, a classic black-spots approach
does indeed make it possible to identify the places where a greater number of collisions or
injuries occur and to take action there, but an area-wide approach offers the potential
advantage of decreasing the risk of collision or injury at a greater number of sites, and thus
at a significant proportion of potential collision sites within an area or neighbourhood. For
example, the article by Morency & Cloutier (2006) identifies 22 “black spots” for
pedestrians on the island of Montréal, with “black spots” being arbitrarily defined as places
where at least eight pedestrians were injured during a five-year period. However, these
“black spots” represent only 4% of the pedestrians injured during this period, whereas
5082 pedestrians were injured at over 3500 different sites.

Figure 10 An example of collision dispersion in an urban

setting
Each point on this map of the island of Montréal indicates a spot where an ambulance
was sent during the 1999-2003 period for a pedestrian injured by a motor vehicle.
Source: Adapted from Morency & Cloutier, 2005, p. 31.
b) Collision migration
If drivers use adjacent residential streets to avoid points on the road network that have
been calmed, a targeted approach that intervenes at “black spots” is likely to displace
collisions elsewhere. To prevent this migration of collisions, the area-wide approach often
explicitly attempts to manage diverted traffic by channelling it toward roads that are better
adapted to higher speeds and volumes and by ensuring the safety of the latter, notably,
through the use of appropriate calming measures (e.g., curb extensions, roundabouts,
speed tables, etc.).
c) Volume reduction
Whereas targeted interventions installed at “black spots” generally seek to reduce speeds
at specific points on the road network to improve safety at those points, an area-wide
approach typically seeks to reduce traffic volumes, as well as speeds, throughout an area.
Such reductions in volume can result not only from efforts to channel a portion of traffic
(through traffic) onto roads that are better designed to receive it, but also from
interventions that encourage the adoption of alternate means of travel and, in particular, of
active and public transportation. In theory, the area-wide approach may thus have an
advantage over the black-spots approach, in that it seeks not only to reduce vehicle
speeds, but also to reduce the number of trips made in automobiles in a given area.

Elvik (2001) published a systematic review with meta-analysis focused on 33 studies

evaluating the effectiveness of area-wide strategies at reducing the number of collisions with
and without injuries. No studies were excluded on the basis of the methodology used.
Table 5 presents the significant results of the meta-analysis, when these include the data
reported by the 33 studies examined:
Table 5 Effectiveness of area-wide strategies according to collision site
Collisions without injury Personal injury collisions

Calming strategy Collision site
{CI 95%} {CI 95%}
Local streets -29%* {-25 to -22} -24%* {-29 to -18}
Area-wide strategies Main roads -11%* {-16 to -6} -8%* {-12 to -5}
Entire area -16%* {-19 to -13} -15%* {-17 to -12}

Source of the data: Elvik, 2001.
The article concludes that there were significant reductions in collisions with and without
injury on local streets and main roads, as well as throughout the area. The reductions
observed on local streets were significantly greater than those on main roads. However,
when retaining only the studies with the most robust designs (“before-after” with control sites)
for the purpose of evaluating the effect on personal injury collisions, only the reduction of
12% (CI 95%: -21 to -1) for the entire calmed area remains statistically significant. The
studies were grouped by decade of publication and by country to allow for verification of the
external validity of the results. These results were relatively consistent across decades and
countries, so the reported effectiveness does not seem to be the result of confounding
factors, as these would have had to influence results in a consistent and uniform manner
across the varying implementation and evaluation contexts.
Another systematic review with meta-analysis, first published in 2003, then updated in 2009,
was carried out by Bunn et al. (2003; 2009). The first version of this review synthesizes the
results of 12 studies examining 16 area-wide strategies, and the second version synthesizes
the results of 18 studies examining 22 area-wide strategies, implemented in high-income
countries. Having searched systematically for studies using “before-after” designs with
control sites or randomized trials, the authors found only “before-after” studies with control
sites. Table 6 presents the results of both versions of the meta-analysis.
Table 6 Effectiveness of area-wide strategies
Personal injury
Collisions involving
Collisions collisions (including Fatal collisions
Version a pedestrian
{CI 95%} death) {CI 95%}
{CI 95%}
{CI 95%}
2003 -5% {-19 to +11} 0% {-16 to +18} -11% {-20 to 0} -37% {-86 to +159}
2009 -11% {-24 to +5} +1% {-12 to +16} -15%* {-25 to -4} -21% {-77 to +168}
Sources of the data: Bunn et al., 2003; 2009.

Despite these generally positive results, only the reduction in personal injury collisions from
the 2009 analysis is statistically significant. This significant reduction (-15%) corresponds to
the significant reduction in personal injury collisions for calmed areas reported in the meta-
analysis by Elvik (2001). It should be noted that the diverse range of calming schemes
included in the literature review (going from a scheme using only stop signs to schemes
using a wide variety of engineering measures) may obscure the superior effectiveness of
some types of schemes.
Figure 11 City of Vancouver West-End

neighbourhood traffic-calming scheme
Source: Zein et al., 1997, Figure 1, p. 4. Copyright, National Academy of
Sciences, Washington, D.C., 1997. Reproduced with permission of the
Transportation Research Board.
An article (Zein, Geddes, Hemsing, & Johnson, 1997) evaluates four area-wide schemes
implemented in the Greater Vancouver region, in Canada, including one in the West-End
neighbourhood of downtown Vancouver, which is illustrated in Figure 11. As an indicator of
the severity of collisions, the value of insurance claims linked to collisions before and after
the interventions was compared. 15 The data used cover a period of one year prior to the
interventions, but no details are provided concerning the data collected after the
interventions. Neither the methodology nor the data used to arrive at the results are
presented in the document. Moreover, no statistical significance test is mentioned. The
results should therefore be interpreted with caution.
15
“Before-after” type studies evaluating the effects of area-wide interventions on collisions, injuries and deaths
are not as subject to bias as those evaluating black-spots interventions. That is why we have included the
former and excluded the latter. In specific terms, because area-wide interventions are not implemented at
specific points where there is a high risk of collisions, their evaluations are less prone to the effects of
regression to the mean. Moreover, since they typically aim at directing diverted traffic toward the appropriate
roads and at making these safe, they are less likely to result in collision migration–which the studies of Grundy
et al. appear to confirm (2008a; 2008b; 2009). However, “before-after” studies are just as likely to mistake a
general downward trend in the number of collisions, injuries and deaths for the effect of interventions,
regardless of whether these are area-wide or isolated.

The article concludes that an average reduction in collisions of 40% occurred, with
substantial variations between areas (-18% to -60%), and that the average reduction in the
value of claims was 38% (-10% to -57%).
In 1999, the Transport Research Laboratory published a report (Cloke et al., 1999)
evaluating the effects of an area-wide scheme, implemented in the Leigh Park area of
Havant, England, comprising various calming measures (raised intersections, pedestrian
refuges, curb extensions, raised pedestrian crossings, speed cushions, raised medians, mini-
roundabouts and gateways). The schematic plan of this intervention is included in
section 3.2.3.1. Among the effects evaluated was the impact on the number of collisions and
personal injury collisions, with occurrences before (3 years) and after (20 months) the
intervention being compared. Because the data collection periods were judged to be too
short, the authors presented the results without carrying out statistical significance tests.
While these results should be interpreted with caution, it is interesting to note that the
number of collisions per year in the area remained similar, along with the number of personal
injury collisions involving motorcyclists, cyclists or adult pedestrians, whereas the number of
personal injury collisions involving child pedestrians decreased by 50% per year.
Figure 12 Scheme involving installation

of 21 mini-roundabouts in
Växjö, Sweden
Source: Hyden & Várhelyi, 2000, p. 12.
The article by Hyden & Várhelyi (2000) evaluates the effect on road safety of a scheme
involving the installation of 21 mini-roundabouts in Växjö, Sweden (see Figure 12), at
intersections with daily traffic volumes of up to 23,500 vehicles. The scheme replaced one
intersection controlled by traffic lights and 20 intersections without signals with mini-
roundabouts. An indirect measurement of road safety was obtained using the conflicts
technique, which made it possible to estimate the number and type of collisions before and
after implementation of the scheme based on the conflicts observed between the various
road users. Thus, twelve intersections affected by the changes were observed for thirty hours
each (the article does not specify how these hours were divided between the periods before
and after the changes). Table 7 presents the estimates based on these observations.

Table 7 Estimated effectiveness based on the conflicts technique of a

mini-roundabout scheme
Personal injury Collisions with Collisions with Collisions with

Calming strategy
collisions injured driver injured pedestrian injured cyclist
Scheme involving
-44% +12% -80% -60%
21 mini-roundabouts
Source of the data: Hyden & Várhelyi, 2000.
According to the authors, the estimated increase in the number of drivers injured can be
partly explained by a poorly-designed mini-roundabout where the risk of personal injury
collision increased by approximately 200%. The authors conclude that, if this mini-
roundabout with a defective design is excluded, the mini-roundabout scheme can be seen to
have led to a “very significant” reduction in risk for vulnerable users (pedestrians and cyclists)
without either increasing or decreasing the risk for drivers and passengers, but no statistical
analysis is presented in support of this claim.
Figure 13 20-mph zones and adjacent

areas in London, U.K.
Source: Grundy et al., 2009, p. 2.
A report (Grundy, Steinbach, Edwards, Wilkinson, & Green, 2008b), as well as a more
synthetic article in a scientific journal (Grundy et al., 2009), examine the 399 zones with
20-mph (32-km/h) speed limits that have been installed in London, U.K., gradually since
1990 (see Figure 13). They evaluated the effects of these zones on collisions, injuries and
deaths occurring within the zones and on their periphery. Typically, the entrance to and exit
from the zones are marked with signs, and traffic within the zones is calmed, in particular,
through the use of vertical deflections (e.g., raised intersections) and horizontal deflections
(e.g., chicanes). The size of calmed zones ranges from a 0.07-km stretch of road to an area
covering 37 km of roads, with a median size of 3.6 km. The authors used collision data
collected by the police department over a period of twenty years (1986-2006) to calculate the
effect of the zones on the roads within them and to verify whether collisions had migrated to
adjacent roads (within 150 m). The data was also used to control for the phenomenon of

regression to the mean and for the underlying downward trend in the number of collisions
occurring in London. According to the results presented, the 20-mph zones are responsible
for a significant reduction in the number of collisions; specifically, a reduction of 37.5%
(CI 95%: -31.6 to -43.4), with no indication of collision migration. In fact, even after controlling
for the underlying downward trend, a significant reduction in collisions of 7.4% (CI 95%: -3.8
to -11.0) was observed on the periphery of the calmed zones. This reduction cannot,
however, be explained with reference to the article. The authors speculate that it may be due
to the proximity of the 20-mph zones or to other interventions on these roads (e.g., speed
cameras on the main roads bordering the zones). Table 8 presents the key results
concerning the number of road injuries by category of user.
Table 8 Effectiveness of 20-mph (32-km/h) zones at reducing personal injury

collisions occurring within them
Drivers or
Pedestrians Cyclists Motorcyclists Total
passengers
{CI 95%} {CI 95%} {CI 95%} {CI 95%}
{CI 95%}
Personal injury -32.4%* -16.9%* -52.5%* -32.6%* -41.9%*
collisions (PICs) {-37.7 to -27.1} {-29.0 to -4.8} {-62.4 to -42.5} {-43.4 to -21.7} {-47.8 to -36.0}
-46.2%* -27.7%* -48.5%*

0-15 years - -
{-55.5 to -36.8} {-49.1 to -6.3} {-55.0 to -41.9}
Collisions with
persons killed -34.8%* -37.6%* -61.8%* -39.1%* -46.3%*
or seriously {-47.5 to -22.18} {-60.9 to -14.4} {-71.7 to -52.0} {-59.1 to -19.0} {-54.1 to -38.6}
injured (KSI)
-43.9%* -50.2%*
0-15 years - - -
{-61.3 to -26.6} {-63.2 to -37.2}
Sources of the data: Grundy et al., 2008b; 2009.
These results show that the 20-mph zones significantly reduced the number of personal
injury collisions (PICs) and of collisions with persons killed or seriously injured (KSI) for the
various groups of public road users. The point estimates suggest that the 20-mph zones are
particularly effective at protecting children (0-15 years old), but in the absence of data
applying solely to adults, it is impossible to determine whether these differences are
statistically significant. A comparison of the measure’s effectiveness at reducing PICs and
KSI collisions in large zones (more than 3.6 km of road) and in small zones (3.6 km and less)
revealed no significant difference. On the basis of the most moderate risk reduction factor,
Grundy et al. (2009) estimated that the 399 20-mph zones in London prevent 203 persons,
including 51 pedestrians from being injured each year, and that, of these, 27 would have
been seriously injured or killed.
In another report, Grundy et al. (2008a) evaluate the effects of the same 399 20-mph zones
on inequalities in road injuries among five levels of socioeconomic status (SES) and four
racial categories. The locations where the collisions occurred were classified into five SES
quintiles based on the Index of Multiple Deprivation 2004, using 2001 census data and police

data from 1987 to 2006. Those injured in road collisions were grouped into four racial
categories (white, Asian, black and other) 16 using police data from 1996 to 2006.
Effects according to socioeconomic status

Based on the classification of areas into quintiles, the authors determined that in 2000 the
percentage of road kilometres covered by 20-mph zones was less than 2%, regardless of
SES quintile. By 2008, the situation had substantially changed in favour of the most deprived
persons. In fact, the more deprived an area, the higher this percentage rose. The percentage
of roads covered was 2.5% for Q1 (the least deprived quintile), 5.9% for Q2, 9.7% for Q3,
17.7% for Q4, and 27.5% for Q5 (the most deprived quintile). Between 1987 and 2006, the
effectiveness 17 of 20-mph zones at reducing annual amounts of personal injury collisions
(PICs) and collisions with person killed or seriously injured (KSI) on roads covered by the
zones and on adjacent roads (within 150 metres) did not vary significantly according to the
SES quintile where zones were implemented. However, there were two exceptions: the
reduction in PICs involving motorcyclists and the reduction in PICs involving car occupants
were significantly higher on roads adjacent to 20-mph zones implemented in the most
deprived areas than on roads adjacent to zones implemented in the least deprived areas.
Despite the fact that the majority of 20-mph zones were implemented in areas with a low
SES, annual aggregate reductions across London, between 1987 and 2006, were
significantly greater in the quintiles with the highest SES, as regards total PICs and collisions
involving children, pedestrians, cyclists, motorcyclists and car occupants.
Effects according to racial categories

Between 1996 and 2006, on roads adjacent to 20-mph zones, annual reductions in PICs and
KSI collisions did not vary significantly among people identified as white, black and Asian.
However, in 20-mph zones, annual reductions were significantly lower for black people than
for white or Asian people, as regards total KSI collisions and PICs involving pedestrians and
child pedestrians. The report does not provide an explanation for these differences, but the
authors speculate that they may be the result of a proportionately and progressively greater
exposure of black people to the risk of collision, in particular because they would have
walked more than white or Asian people. Between 1996 and 2006, annual reductions were
also significantly greater for white people than for Asian or black people, as regards total KSI
collisions, total PICs and PICs involving children, pedestrians, child pedestrians, cyclists,
motorcyclists and car occupants.
16
The researchers used the terms “Black,” “White,” “Asian,” and “Other” to group together “police-assigned
‘ethnicity’ code from the STATS19” (Grundy et al., 2008a, p. 10), the form used by London's police to report
collisions. While the authors of the present document have reservations about the use of these particular
terms, neither do we wish to ignore the evidence that inequalities exist along the line of long-standing
categories related to perceptions of race. With this proviso, we will repeat them as such.
17
As is standard practice in the literature, effectiveness is here expressed as a percentage reduction in personal
injury collisions and fatal or serious injury collisions, and not in terms of the number of injuries, serious injuries
and deaths avoided. This practice is not without consequence for the results, for there are often more
collisions, injuries and deaths in deprived areas than in the least deprived areas (Laflamme et al., 2010;
Cubbin & Smith, 2002; Morency & Cloutier, 2005). Thus, it is likely that interventions whose effectiveness,
expressed as a percentage, is the same for areas with different SES would prevent more collisions, injuries
and deaths if they were implemented in areas with a lower SES.

These results led the authors to conclude that inequalities in road injuries have increased in
London despite the more intensive implementation of 20-mph zones in the most deprived
areas. On the basis of a less moderate risk reduction factor than that used in 2009 (Grundy
et al., 2009), the authors conclude that by targeting deprived areas for the implementation of
20-mph zones, 1193 personal injuries were prevented each year, with almost half of these
being in the most deprived quintile. Based on the number of injuries avoided, 20-mph zones
can thus be seen to have reduced the widening of the gap between the least deprived and
the most deprived quintiles by about 15%. It should be noted that this result was not
accompanied by a statistical significance test.
Another article (Jones et al., 2005), focused on two similar cities in England, also evaluates
the potential of traffic-calming measures for reducing inequalities in road injuries, but in this
case the focus is limited to child pedestrians (4-16 years old) injured following a collision in
areas with varying levels of socioeconomic status (SES). The two cities were divided into
areas classified according to four levels of SES. An audit was carried out of the calming
measures installed in the two cities, but this audit was limited to speed humps, road
narrowings, and road closures. In city A, the areas in the most deprived fourth had
4.80 (CI 95%: 3.71 to 6.22) times more calming measures per 1000 residents than those in
the least deprived fourth; that is, a significantly greater ratio between these two fourths than
that of city B (1.88 [CI 95%: 1.46 to 2.42]). Rates of personal injury collisions per
1000 residents were compared for the cities and for areas. In the absence of precise
information about the dates when the calming measures were installed, the periods of 1992-
1994 and 1998-2000 were compared, municipal authorities in both cities having indicated
that installation of the measures occurred mainly from the mid 1990s onwards. The analyses
took into account the number of children in both cities, the number of families with a car and
the number of children walking to school. It should be noted that, although the evaluation
considers the effect of calming measures on an area comprising more than one road, the
article does not indicate whether the measures installed on these roads were planned and
implemented so as to function in a systemic manner (area-wide approach) or resulted from
an accumulation of individual measures installed at specific points over time (black-spots
approach). Table 9 summarizes the results presented in this article.
Table 9 Effects of the number and distribution of calming measures on

inequalities in road injuries among child pedestrians
Child pedestrian (4-16 years old) Gap between the rates (‰) of the
injury rate (‰) least and most deprived fourths
Difference 1992-1994 1998-2000
1992-1994 1998-2000
{CI 95%} {CI 95%} {CI 95%}
-2.14* 3.21* 2.01*
City A 6.98 4.84
{-2.81 to -1.48} {2.27 to 4.54} {1.45 to 2.87}
Most deprived -3.68*
9.53 5.85
areas {-5.28 to -2.13}
-0.80 4.27* 3.96*
City B 6.05 5.25
{-1.88 to 0.27} {2.51 to 7.28} {2.26 to 6.95}
Most deprived -1.32
8.99 7.67
areas {-3.62 to 0.97}
* Significant variation or difference of at least, p <0.05.
Source of the data: Jones et al., 2005.

These results led to the observation that in city A, where a greater number of calming
measures were installed (city A: 891/city B: 553), the child pedestrian injury rates decreased
significantly throughout the city and in the most deprived areas. As for city B, the point
estimates indicate reductions, but these are not statistically significant. Based on the point
estimates, the reduction in the gap between the injury rates of the least and most deprived
areas seems greater for city A, where a greater proportion of measures were installed in the
most deprived areas (city A: 4.80/city B: 1.88). However, neither of the gap reductions is
statistically significant, nor, moreover, is the gap between the rates of the two cities.
Nevertheless, the results of the article indicate that variations in the child pedestrian injury
rate are inversely correlated (r=-0.769, p=0.026) with the density of traffic-calming measures
in an area, that is, to the number of calming measures per kilometre of road. Thus, the article
supports the conclusion that the more numerous traffic-calming interventions in city A are
associated with a significant reduction in collisions involving injured child pedestrians, but it
does not support the conclusion that having targeted the most deprived areas led to a
significant reduction in inequalities. Ultimately, even though the authors of the article were
unable to measure such an effect, increasing the density of calming measures in deprived
areas would likely help reduce inequalities.
3.1.3.2 Summary of the results for area-wide calming schemes
The documentary search led to the identification of three systematic reviews with meta-
analyses (Elvik, 2001; Bunn et al., 2003; 2009) (including one revised update) and three
other articles (Zein et al., 1997; Grundy et al., 2009; Jones et al., 2005) belonging to the
scientific literature, as well as four studies (Cloke et al., 1999; Hyden & Várhelyi, 2000;
Grundy et al., 2008a; 2008b) drawn from the grey literature. These studies evaluate the
effectiveness of various area-wide schemes at reducing collisions, injuries and deaths, as
well as their unequal distribution within the population.
They use different methodological approaches and are not all equally rigorous, but they
generally arrive at results which converge and which agree with the main mechanisms of
action recognized. The key points to recall are the following:
a) Reduction of collisions, injuries and deaths
Two of the three meta-analyses report significant reductions of 15% in personal injury
collisions (Bunn et al., 2009; Elvik, 2001), while the third reports reductions, but not
significant ones (Bunn et al., 2003). With regard to the studies examining specific area-wide
interventions, these all report overall reductions in collisions, personal injury collisions, and
collisions causing serious injury or death. Of these latter studies, all those whose results are
accompanied by statistical significance tests report significant reductions (Grundy et al.,
2008b; 2009; Jones et al., 2005).
b) Protection of the most vulnerable road users

Neither of the meta-analyses that focused on the effects on pedestrians notes any effect on
this category of user (Bunn et al., 2003; 2009). However, the other studies that focus on
vulnerable users report significant reductions in collisions involving pedestrians, cyclists and
motorcyclists (Grundy et al., 2008b; 2009; Jones et al., 2005) or report reductions without
mention of statistical significance tests (Hyden & Várhelyi, 2000). Three studies also point

out that more marked reductions were observed among children walking or cycling than
among adults using these same modes of travel (Grundy et al., 2008b; 2009; Hyden &
Várhelyi, 2000).
c) Reduction of inequalities
One of the studies controlling for several confounding factors (Grundy et al., 2008a)
concludes that the effectiveness (expressed as a percentage) of 20-mph zones at reducing
personal injury collisions did not vary according to the socioeconomic status of the area in
which they were implemented, but that they tended to protect people identified as black less
than those identified as white or Asian. This study and an article that controlled for fewer
factors (Jones et al., 2005) evaluate whether having targeted deprived areas for the
implementation of area-wide schemes led to a reduction in inequalities. One (Jones et al.,
2005) reports a non-significant reduction in inequalities and the other (Grundy et al., 2008a)
reports a reduction in the growth of inequalities, but without accompanying this result with
statistical significance tests.
d) Collision migration
Notwithstanding the fact that area-wide strategies generally seek to redirect some of the
traffic using an area’s residential streets toward the main arteries of a road network, one
meta-analysis (Elvik, 2001) and one of the more recent articles (Grundy et al., 2009) indicate
that not only do collisions not migrate with the redirected traffic, but also that their number
decreases significantly on these arteries (Elvik, 2001; Grundy et al., 2009). However, the
articles do not explain these results.
e) Promising area-wide strategies

The studies consulted do not allow for a comparison of the effectiveness of the various area-
wide strategies evaluated. However, one article (Jones et al., 2005) indicates that the density
of calming measures (no. of measures/km) is significant, because it was inversely correlated
to the rate of child pedestrians injured. Another study (Grundy et al., 2008b) suggests that
the size of calmed areas does not affect the effectiveness of area-wide strategies at reducing
personal injury collisions and collisions causing serious injury or death.
3.1.4 Conclusion of the section on the number and seriousness of collisions
3.1.4.1 Section summary
Traffic-calming interventions are mainly promoted as a way of reducing the number of

collisions, injuries and deaths. Despite the range of traffic-calming measures and strategies
evaluated, the studies consulted lead to the conclusion that, in general, traffic calming
effectively helps improve road safety. This conclusion is based on several studies whose
methodologies and levels of rigorousness vary and whose results converge and, generally,
agree with the mechanisms of action identified. The studies show that interventions targeting
locations where the risk of collisions and injuries is high (black-spots approach) as well as
area-wide strategies can substantially reduce collisions, injuries and deaths. Thus, it is
difficult to draw conclusions about the advantages and disadvantages of individual
interventions relative to area-wide interventions, especially since the significance of gains in

both cases likely derives from the intensity, the extent and the context of interventions,
regardless of the approach used.
3.1.4.2 Research gaps
Comparison of approaches
It would be relevant for studies to seek to compare the effectiveness at reducing collisions,
injuries and deaths of calming strategies based on the black-spots approach and strategies
arising from the area-wide approach. A city such as Montréal, where the two types of
strategies are currently being developed in different boroughs, offers a very interesting
opportunity to compare the two approaches.
Traffic diversion
The area-wide approach rests partly on the hypothesis that the diversion of motorized traffic
away from local residential streets and onto the main road network (arteries, highways) will
not simply displace collisions from one place to another, but will rather decrease their
number or seriousness. However, the risk of collision and injury on arteries, in particular, is
generally high (Dumbaugh & Rae, 2009; Lovegrove & Sayed, 2006). While two articles
(Elvik, 2001; Grundy et al., 2009) report that the area-wide strategies evaluated did not
cause collisions to migrate from local streets to arteries, neither of these articles took into
account in their analyses the possible presence of measures aimed at adapting these roads
to accommodate an increase in volume or at enhancing their safety, for example, through the
use of calming measures adapted to this type of road (e.g., roundabouts, curb extensions,
speed tables, etc.). Research would be required to determine the conditions under which
diverting traffic through area-wide strategies would not cause collisions to migrate to the
main road network.
Figure 14 Speed table: a measure

designed for faster lanes
and heavy vehicles
Source: Boulter et al., 2001, p. 11.

Inequalities
In a similar vein, if it is possible that some interventions which redirect a certain amount of
traffic from local streets onto arteries increase road safety problems on those arteries, then,
given that persons with low socioeconomic status tend to be overrepresented as residents of
these roads (Smargiassi, Berrada, Fortier, & Kosatsky, 2006), it would be relevant for
researchers to examine the potential effects of such redirection on inequalities in road
injuries and deaths.
3.2 AIR QUALITY
In urban environments, traffic is responsible for a significant portion of emissions of nitrogen

oxides (NOx), carbon monoxides (CO), volatile organic compounds (VOCs)–including
hydrocarbons (HC), and particulate matter (PM). As an example, in Toronto, motor vehicles
are responsible for 85% of CO, 69% of NOx and 16% of PM2.5 (Toronto Public Health, 2007).
The air pollution to which these emissions contribute has been associated with increases in
the incidence and in the prevalence of certain health problems, such as premature death,
cardiovascular and respiratory diseases, cancers, etc. (Toronto Public Health, 2007). It is
important to distinguish between the emission of air pollutants and their effects on air quality,
because a variation in emissions does not necessarily produce a linear variation in ambient
air quality. In reality, the relationship between the two is mediated by, for example, ambient
temperature, prevailing winds, the presence of other reactive molecules or other emission
sources, etc. This said, various actions can be taken to reduce traffic emissions or the effects
of these emissions on the quality of the air to which populations are exposed. For example,
technology can reduce vehicle emissions (e.g., catalytic converters), and urban planning and
the planning of transportation networks can reduce traffic volumes (e.g., densification, public
transportation) or move traffic away from residential areas (e.g., urban bypasses), etc.
Although traffic-calming strategies are primarily promoted as a way to reduce road collisions,
injuries and deaths, it is likely that they could, in certain contexts, help reduce vehicle
emissions on calmed roads and, sometimes, shift them outside the calmed areas. These
changes could affect the quality of the air to which populations are exposed, and thus their
health.
According to the studies consulted, traffic-calming strategies can influence air pollutant
emissions and ambient air quality by means of three main mechanisms:
a) Reduction of vehicle speeds
The variation of air pollutant emissions according to vehicle speed is well established (Owen,
2005). As illustrated in Figure 15, emissions generally follow fuel consumption, which
appears graphically as a U-shaped curve; that is, fuel consumption and emissions per
kilometre travelled are greater at low and at high speeds (Sergerie et al., 2005;
Environmental Protection Agency [EPA] & U.S. Department of Energy [DOE], 2010; WHO
Regional Office for Europe, 2005). Thus, it is not surprising that the “EPA MOBILE Model”
(http://www.epa.gov/OMS/mobile.htm) predicts that the slowing down of traffic due to traffic-
calming strategies should be accompanied by an increase in emissions (Transportation
Research Board, 1995).

Figure 15 Effect of mean travelling

speed on emission levels from
passenger cars with catalysts
Note: Values for nitrogen oxides (NOx) and hydrocarbons (HC)
have been multiplied by 10 and those for carbon dioxide (CO2)
have been divided by 100.
Sources: Adapted from Ntziachristos & Samaras, 2000, in WHO
Regional Office for Europe, 2005, p. 25. Copyright 2000,
reprinted with permission from Elsevier.
b) Reduction of speed variations

Accelerations, decelerations and idling time also influence vehicle pollutant emissions,
because a combustion engine consumes more fuel when turning slowly, while a vehicle is
idling, or when accelerating, than when it is turning at a constant speed (EPA & DOE, 2010;
Bahar, Smahel, & Smiley, 2009). The increased fuel consumption that results from frequent
variations in driving speed and the less complete combustion that occurs during
accelerations produce an overall increase in emissions (Houwing, 2003; Sergerie et al.,
2005). Frequently varying speed while driving thus tends to produce more air pollutant
emissions than driving at a constant speed, which explains, moreover, the difference in the
emission levels associated with aggressive and calm driving styles (Owen, 2005; EPA &
DOE, 2010; Houwing, 2003). The configuration of calming measures (which generally
determines the speed at which they can easily be negotiated), their spacing (which
determines the potential for acceleration and deceleration between them) and how they
manage right of way at intersections (which determines accelerations, decelerations and
idling time at intersections) should therefore influence their impact on pollutant emissions.
Thus, interventions succeeding in reducing speed variations should contribute to decrease
pollutant emissions by vehicle.
c) Reduction of traffic volume

Air pollutant emissions from motor vehicles vary in accordance with the number of trips taken
and the distance travelled in motor vehicles. Given the reduced effectiveness of catalytic
converters during the first few minutes of driving and the emissions caused by initial
acceleration, WHO Regional Office for Europe estimated in 2005 that approximately 90% of
air pollutants are emitted during the first 200 seconds that a car is in motion when travelling

less than 6 km in a city (WHO Regional Office for Europe, 2005). Since a significant number
of motor vehicle trips taken in urban settings cover short distances, 18 a change in the number
of such trips (due to an increase in walking or cycling trips, for example) should have a
greater effect on total emissions than an increase or decrease in the distances travelled.
When calming strategies succeed in reducing the volume of motor vehicle traffic, by
increasing walking, for example, this should result in a reduction in emissions and an
improvement in air quality, whereas pollutant emissions may simply be displaced if traffic is
diverted from one area to another.
Most of the studies consulted that examine the effects of traffic-calming strategies on air
pollutant emissions and air quality inferred these effects by referring to the three mechanisms
above, without attempting to measure the effects. We have retained for purposes of analysis
only those studies that evaluated the effects of one or of a series of calming measures. As in
the preceding section, we will first present the data on isolated interventions installed at
specific points on the road network or along a single street, followed by the data on area-
wide calming schemes.
All the studies consulted that considered the effects of individual interventions on air quality
evaluated these effects by examining the air emissions of motor vehicles. Probably because
the impact of such interventions is very restricted geographically, only one article (Boulter et
al., 2001) attempted to measure their effects on ambient air quality.
A literature review produced by the Transport Research Laboratory (Boulter & Webster,
1997) demonstrates that the results of seven studies conducted prior to 1997 focusing on
European interventions at isolated points of the road network or on a single street vary
considerably. Table 10 summarizes these results. 19
18
An article estimates, for example, that in the Montréal metropolitan region, 11.7% (n=862,000) of trips carried
out weekly cover distances of less than 1.6 km and that 55% of these are taken in motor vehicles (Morency,
Demers, & Lapierre, 2007). In the United States, according to the authors of one article, 41% of all trips carried
out in urban settings cover distances of less than 3.2 km and 28% cover less than 1.6 km, with 89% of
Americans using their car for trips of between 1.6 and 3.2 km and 66% for trips covering less than 1.6 km
(Pucher & Renne, 2003). In Europe, according to WHO estimates, 30% of car trips cover distances of less than
3 km and 50% cover distances of less than 5 km (WHO Regional Office for Europe, 2011).
19
To facilitate comparison of the various results, all the tables in the present section have the same columns.
This choice explains why some tables have entire columns for which no data are available (“-”). Fuel
consumption and CO2 emissions are presented together because their variations are strongly correlated
(Boulter & Webster, 1997).

Table 10 Effects of various calming measures on air emissions
Type of
Calming vehicle
CO HC CO2/Fuel NOx PM
measures (gas-powered
cars)
with and
Speed without
+70% to +80% +70% to +100% +50% to +60% -20% to 0% -
humps catalytic
converter
with catalytic
+20% - +4% +18% -
converter
One speed
hump without
catalytic +11% - +5% +22% -
converter
with catalytic
+300% - +37% +300% -
converter
Ten speed
humps without
catalytic +200% - +51% +300% -
converter
Various
measures, - - - -9% - -
40-km/h limit
Six speed with catalytic
+300% - +25% +1000% -
humps converter
Five speed
- - - +36% to +73% - -
humps
Two
- - - +33% - -
roundabouts
Source of the data: Boulter & Webster, 1997.
Despite wide variation among the studies, nearly all of them report substantial increases in
the various air pollutants measured. According to the authors, a decrease in driving speeds
and more frequent speed variations following the interventions explain the increases
observed.
To follow up on the variations observed in this literature review, the Transport Research
Laboratory undertook to study the specific impact of various calming measures (flat-top and
round-top speed humps, raised intersections, speed cushions, chicanes, curb extensions,
combination of a road narrowing and a speed cushion, mini-roundabout) on air pollutant
emissions and ambient air quality 20 (Boulter et al., 2001). Actual driving speeds were
measured without the knowledge of drivers, before and after the installation of calming
measures or on street blocks with calming measures and on control street blocks, to
establish typical driving cycles for each calming measure. These cycles were then
20
While the study evaluated the effects of the calming measures taken in isolation, it should be noted that almost
half of these were part of area-wide calming schemes. The manner in which the study presents the results
does not allow for their separation on the basis of the two approaches to traffic calming.

reproduced under laboratory conditions using three types of cars, to measure the associated
emissions. Without offering any real explanation or justification, the authors indicate that
seven of the twelve cars used were replaced by other cars of the same type during the
experiment. Since this may have compromised the validity of the results, these should be
interpreted with caution. Table 11 presents the average variations measured for all the
calming measures taken together.
Table 11 Effects of calming measures on air emissions according to type of

vehicle
Calming Type of vehicle

measure (cars)
Gas with catalytic
+59%* +54%* +26%* +8% -
converter
Various
Gas without
measures +34%* +50%* +20%* +1% -
catalytic converter
Diesel +39%* +48%* +26%* +28%* +30%*
* Significant variation at p <0.05.
Source of the data: Boulter et al., 2001.
These results show that, apart from the NOx emitted by gas-powered cars, the calming
measures, on average, led to significant increases in the emission of most pollutants. To
determine the impact of these increases on air quality 10 metres from road centres, the
authors used a dispersion model. The results take into consideration reduced traffic volume
recorded at one site and increases at four other sites following the interventions, but the
author does not specify when these measurements were taken. The results also take into
account the traffic volume of heavy vehicles, basing calculations on the relationship,
established in a previous study, between the speed and the average emissions of such
vehicles. Based on the proportion of vehicles with and without catalytic converters circulating
in England in 2000, the authors estimated that the interventions had, on average, the
following effects on air quality: +0.4% to +6.7% for CO, +2.7% to +10.1% for benzene, +2.7%
to +9.5% for 1,3-butadiene, and -0.4% to +4.3% for NO2. The reported values seem to
indicate a slight deterioration in air quality, but in the absence of any statistical significance
test, and given the unexplained increases in traffic volumes, these results should be
interpreted with caution.
In 2005, the Society of Automotive Engineers International published a report (Daham,

Andrews, Li, Partridge, Bell, & Tate, 2005) on the impact of speed humps on air pollutant
emissions. However, the method used casts doubt on the validity of the results obtained. To
simulate a road without speed humps, the authors drove at a constant speed of 50 km/h on a
road with seven speed cushions installed on it, and to simulate the presence of seven 80 mm
speed humps (the highest kind), they drove on the same road, slowing down to 16 km/h to
go over the speed cushions and accelerating up to 32-50 km/h between them. Using
equipment installed on the car to measure emissions in real time, they simulated the impact
of speed humps on a car with a very heavy load, thus producing results that are probably not
very representative of a car with an average load. Moreover, by slowing down more than
necessary to easily go over the simulated speed humps, and then accelerating rapidly

between the speed humps, the authors simulated an aggressive driving style, which probably
led to an overestimation of average emissions on the calmed road. Indeed, the authors state
that their results are “probably a representation of an unsmooth driver who is in a hurry to
negotiate a traffic-calmed road driving a heavily laden car” (Daham et al., 2005, p. 14).
Table 12 summarizes the results obtained.
Table 12 Effects of speed humps on air emissions
Type of vehicles
Calming measures CO HC CO2/Fuel NOx PM
(gas-powered cars)
With catalytic
Speed humps +117% +148% +90%/+35% +195% -
converter
Source of the data: Daham et al., 2005.
According to these results, speed humps engender significant increases in air emissions.
However, by changing the driving style (from calm to aggressive) between the control cycle
and the experimental cycle, the authors introduced a confounding factor into their study.
Moreover, these results are probably not representative of a car with an average load.
Therefore, these results should be interpreted with caution.
An article by Ahn & Rakha (2009) presents the results of a study in which speeds and speed
variations were measured on three roads and air pollutant emissions were calculated using a
computer model. The first road had a speed limit of 40 km/h: driver speeds were measured
before and after the installation of five speed cushions. On the second road, whose speed
limit was 25 km/h, speeds were measured before and after the installation of two speed
bumps. 21 On the third road, whose speed limit was 40 km/h, the impact of the calming
measures was determined by having a car driven on the road as illustrated in Figure 16. The
pollutants emitted at intersections without calming measures were compared to those of
calmed intersections.
Figure 16 Diagram of a road with calmed intersections

Source: Reprinted from Ahn & Rakha, 2009, p. 414, Copyright 2009, with permission
from Elsevier.
21
It is important to distinguish between “speed bumps” and “speed humps.” While the former are narrow (often
less than 30 cm across) and are crossed at very low speeds (5-10 km/h), the latter form a broader platform
(typically 3 to 4 m across) and are crossed at higher speeds (typically 30 km/h). The former are usually
installed in alleys and parking lots, whereas the latter are usually installed on local streets, where slow vehicle
speeds are desired.

On the first two roads, the presence of calming measures significantly increased per car
emissions of the various pollutants. On the third road, intersections with calming measures
gave rise to significantly higher emissions than intersections without calming measures or
stop signs. However, on the latter road, the authors did not calculate the emissions of
vehicles on cross streets. Yet, the calming measures introduced probably affected speeds
and speed variations on these streets by changing how the right of way was managed at the
intersections. When calming measures cause traffic to flow more freely on cross streets, by
replacing stop signs with a mini-roundabout, for example, this is likely to lead to a decrease
in emissions on these streets. Because of this omission, the results for the third road cannot
be considered representative of the total impact of the calming measures on air pollutant
emissions. For this reason, the results for the third road (stop signs, mini-roundabouts, speed
humps) must be interpreted with caution. Table 13 summarizes the results for the three
roads studied.
Table 13 Effects of calming measures on air emissions
Type of
Calming measures Road vehicles CO HC CO2/Fuel NOx PM
(cars)
st
Speed cushions 1 Gas +47%* +54%* +48%* +98%* -
nd
Speed bumps 2 Gas +9% +20%* +29%*/+33%* +19%* -
rd
Stop signs 3 Gas +145%* +125%* +112%*/+114%* +264%* -
rd
Mini-roundabouts 3 Gas +20%* +31%* +35%*/+34%* +56%* -
rd
Speed humps 3 Gas +44%* +51%* +52%*/+53%* +110%* -
Source of the data: Ahn & Rakha, 2009.
Whether or not they are considered traffic-calming measures by specialists, stop signs are
often initially seen as a traffic-calming panacea by residents faced with a traffic problem
(Ewing, 1999). Notwithstanding the weaknesses of the article, it is worth noting that this is
the intervention that led to the greatest increase in air pollutants, a result that might calm
residents’ enthusiasm for stop signs and lead one to consider that replacing intersections
controlled by stop signs with mini-roundabouts could, depending on the context, help reduce
vehicle emissions.
3.2.2.2 Subjective study
An article (Morrison, Thomson, & Petticrew, 2004) evaluates the perception of residents
living on a road in Glasgow, Scotland, six months before and six months after the installation
of several calming measures (five pairs of speed cushions, parking bays, and two zebra
crossings). Of the 750 residents randomly selected to fill out the postal questionnaire,
244 answered before the intervention and 185 answered afterwards. They were asked to
evaluate their feelings about problems related to their environment, using a scale of 1 to 7
(1 being the happiest and 7 the unhappiest). Having compared the answers given before and
after the intervention, the authors found a significant improvement of 2.88 points in answers
to the question regarding traffic fumes.

3.2.2.3 Summary of results for isolated traffic-calming measures
The documentary search led to the identification of two articles (Ahn & Rakha, 2009;
Morrison et al., 2004) belonging to the scientific literature, one of which examines residents’
perceptions (Morrison et al., 2004), and three studies (Boulter & Webster, 1997; Boulter et
al., 2001; Daham et al., 2005) drawn from the grey literature, including a traditional literature
review (Boulter & Webster, 1997). These studies evaluate the effects of individual calming
measures on air emissions and air quality. They use different methodological approaches
and are not all equally rigorous, but they present results which, overall, converge and agree
with the recognized mechanisms of action. The key points to recall are the following:
a) Increase in per vehicle emissions
All of the objective studies report that the majority of isolated interventions led to increases of
varying size for per vehicle emissions. The two studies (Boulter et al., 2001; Ahn & Rakha,
2009) that mention statistical significance tests report statistically significant increases for
most of the pollutants measured. These results are linked to reduced traffic speeds and to an
increase in speed variations caused by most of the isolated interventions. Nevertheless, it is
still possible that interventions which reduce speed variations on top of traffic speeds can
also reduce vehicle emissions (Ahn & Rakha, 2009).
b) Slight deterioration of air quality

The only study (Boulter et al., 2001) that evaluates air quality on calmed roads reports a
slight deterioration, on average, of air quality following interventions, but traffic volumes
increased on most of the roads studied, and no information is provided about the statistical
significance of the results.
c) Perception
Contrary to what the objective studies may lead one to expect, one article (Morrison et al.,
2004) reports that residents of a calmed road were significantly less unhappy about vehicle
emissions after calming measures were introduced.
d) Promising calming measures

Without providing conclusive evidence, one article (Ahn & Rakha, 2009) indicates that
replacing an intersection managed with stop signs by a mini-roundabout could, by reducing
vehicle speed variations, reduce per vehicle pollutant emissions.
Area-wide strategies are often used not only to reduce driving speeds, but also to reduce
traffic volumes in the areas where they are implemented. Thus, some studies took into
account reductions in volume when evaluating the effects of area-wide strategies on air
emissions and air quality.

The traditional literature review carried out by Boulter & Webster (1997) examines the six
studies that focus on area-wide strategies separately. The review specifies that the studies
examined the effects of speed without taking into account the reduction in traffic volume that
may have occurred following the interventions. Table 14 presents the variations in emissions
measured following the interventions.
Table 14 Effects of various area-wide strategies on air emissions
Type of
Calming vehicles
measures (gas-powered
cars)
Area with Without
extensive traffic catalytic +7% to +71% -25% to -10% +7% to +19% -60% to -38% -
a
calming converter
Without
30-km/h zone catalytic -20% to +28% -23% to +2% -6% to +14% -31% to -5% -
converter
Speed hump Slight
- No change - Decrease -
scheme increase
Area with
extensive - Increase Increase - Decrease -
calming
21 mini-
roundabouts, - +2% - +<1% +1% -
b
30-km/h limit
a
This is the expression used by the authors of the literature review, although they do not define it.
b
Boulter & Webster (1997) classify this study among those evaluating cases where a series of calming measures were
installed on one road. However, the evaluation is of an area-wide scheme involving the installation of 21 mini-roundabouts on
more than one street in Växjö, Sweden, also evaluated by Hyden & Várhelyi (2000) and Várhelyi (2002); which is why it is
presented in this section.
Source of the data: Boulter & Webster, 1997.
All of these studies report a decrease in NOx emissions or a negligible effect on them, but
note variable results with respect to the other pollutants.

Figure 17 Leigh Park area calming scheme

Source: Department for Transport, 1999, p. 2.
evaluating the environmental impact of an area-wide traffic-calming scheme involving the
installation of several types of calming measures in the Leigh Park area of Havant, England.
Figure 17 shows the calming measures installed and their distribution in the area. The
speeds and speed variations of 20 subjects who drove through the zone were measured to
establish driving cycles. On roads calmed with speed cushions, traffic volume reductions of
15% to 35% and speed reductions of 18-19 km/h were recorded at the locations where the
speed cushions were installed. It was estimated that 10% of traffic was diverted outside of
the calmed sector, the equivalent of approximately 600 fewer vehicles per day in the area.
Variations in vehicle emissions, calculated using a computer model, are presented in
Table 15.

Table 15 Effects of the Leigh Park traffic-calming scheme on air emissions
Vehicle emissions
Type of vehicles CO2/
Air emissions
(gas-powered) CO HC Fuel NOx PM
Emissions on roads with speed
cushions (without taking into Cars +22% +24% +9% -12% -
account volume reduction)
Emissions on roads with speed
Cars and heavy
cushions (taking into account -8% -7% -12% -22% -
vehicles
volume reduction)
Emissions on roads calmed with
Cars and heavy
other measures (taking into account -4% -2% -4% -8% -
vehicles
volume reduction)
Cars and heavy
Emissions for the area -6% -5% -8% -15% -
vehicles
Source of the data: Cloke et al., 1999.
With the exception of NOx emissions per vehicle, which decreased, the calming of roads
resulted in an increase in per vehicle pollutant emissions. Nevertheless, when the reduction
in traffic volumes is taken into account, the result is a decrease in emissions on roads
calmed and in the entire area, although the study does not indicate whether these results are
statistically significant. Based on measurements of NO2 and benzene (a type of HC) taken
using diffusion tubes distributed at 6 sites, including 2 control sites outside the area, before
and after the interventions, a non-significant improvement, at p <0.05, in air quality (-5%
benzene, -1% NO2) was observed in the area, as compared to control sites. Thus, the
creation of the zones did not significantly affect the air quality within them.
Várhelyi (2000) published an article evaluating the effects of a scheme involving the
installation of 21 mini-roundabouts in Växjö, Sweden, on the fuel consumption and emissions
of cars. Driving cycles were observed before (n=600) and after (n=800) the installation of the
mini-roundabouts. Measurements were made using a car equipped to record the distance it
travelled every two seconds, making it possible to determine travelling speed and speed
variations. The driver of this car copied the movements of other randomly selected drivers,
without their knowledge. The information recorded was used to establish typical driving
cycles. The fuel consumption and emissions resulting from these typical driving cycles were
calculated using a model that took into account traffic volumes at the various branches of the
intersections, which did not vary significantly following the intervention. The calculations
simplified reality by assuming that all the vehicles counted were cars (whereas heavy
vehicles represented about 7% of traffic), that they were all gas-powered and that 30% of
them were equipped with a catalytic converter. Table 16 summarizes the results obtained.

Table 16 Effects on air emissions of a scheme involving installation of

21 mini-roundabouts
Type of vehicles
Calming measures CO HC CO2/Fuel NOx PM
(gas-powered cars)
Mini-roundabout
30% with catalytic
replacing an intersection +6% - +3% +4% -
converter
without signals
30% with catalytic
Main road +13%* - +8% +8% -
converter
30% with catalytic
Secondary road -20% - -21% -15% -
converter
Mini-roundabout
30% with catalytic
replacing an intersection -29% - -28% -21% -
converter
with traffic lights
Source of the data: Várhelyi, 2002.
Although only one result was statistically significant, all the results in this article agree with
the outcomes expected based on the mechanisms of action: a decrease in emissions
following improved flow at secondary road intersections previously without signals (where
vehicles systematically yielded passage); a decrease in emissions following a decrease in
speed variations at a mini-roundabout replacing an intersection with traffic lights (where all
vehicles on each of the branches alternated stopping); and an increase in emissions
following the slowing down of traffic at intersections on main roads previously without signals
(where vehicles systematically had right of way). Even though the results concerning the
mini-roundabout replacing an intersection with traffic lights are not statistically significant
(probably because a single roundabout was evaluated), the point estimates suggest that this
intervention holds promise as a way to reduce emissions. The article points to the
importance of the implementation context of an intervention and of the distribution of traffic
volumes on the different branches of an intersection in determining the effects of a calming
measure installed at an intersection, but the results regarding the overall effect of the area-
wide scheme on per vehicle emissions are inconclusive.
In another article, Owen (2005) evaluated the effects on pollutant emissions and air quality of
six 20-mph (32-km/h) zones implemented in North West England. The six zones studied
were relatively small (approximately 0.5 km x 0.5 km). Speed in the zones was controlled by
physical traffic-calming measures, such as speed humps and traffic lights. Measurements of
ambient concentrations of NO2 and benzene were taken before (5 to 9 months) and after
(3 to 12 months) the creation of the zones. These measurements were taken at three sites
within each zone and at one control site using diffusion tubes and thermal desorption tubes,
which are subject to wide margins of error (±25% for diffusion tubes). On the basis of these
measurements, the article concludes that the introduction of calming measures neither
improved nor worsened air quality significantly in the six zones. Basing calculations on
average driving speeds within five of the six zones, the author found that per vehicle
emissions rose (0% to +5% NOx; +11% to +34% benzene), but when variations in traffic
volume were taken into account, emissions were found to have decreased in the majority of
zones (+8%, -18%, -9%, -32% and -80% NOx; +22%, +3%, -15%, -32% and -76% benzene).

However, using average speeds to calculate emissions can lead to the underestimation or
overestimation of emissions by failing to consider the effect of speed variations. A dispersion
model indicated that local traffic was responsible for a small percentage of ambient
concentrations of NOx and, more specifically, of NO2 (4% to 14% of NOx; 0% to 3% of NO2).
This weak contribution to NOx levels and the use of relatively imprecise measuring equipment
probably explains, at least in part, why no significant variation in air quality was measured
following the interventions.
3.2.3.2 Subjective studies
Within the context of the report on Leigh Park (Cloke et al., 1999), discussed previously,
interviews before (n=151) and after (n=150) implementation of the calming measures were
carried out with groups of residents, in particular, to determine their perception of airborne
nuisances in the area. The report does not specify how the interviews were carried out or
how participants were selected. Moreover, only 113 residents participated in the interviews
both before and after the intervention, and the results are not accompanied by statistical
significance tests. Due to these limitations, the results should be interpreted with caution.
The majority of respondents (68%) had perceived no change with respect to airborne
nuisances (dust, dirt, fumes, smoke) when out walking after the interventions. Of those who
perceived a difference, the majority reported an increase in nuisances (21%, versus 10%
who perceived a reduction). However, the percentage of respondents who reported being
“very much” or “quite a lot” bothered by dust or dirt (42% before/46% after) or by smoke
(42% before/38% after) in their homes remained basically the same.
The report prepared by Hemsing & Forbes (2000) for the City of Ottawa and the Ottawa-
Carleton region presents the perception of the residents of calmed roads and zones
according to the type of road on which they live (local, collector or arterial) and the type of
calming measure installed (horizontal, vertical or a combination of both). The survey was
conducted using questionnaires distributed on a non-random basis to residents of already-
calmed roads. Thus, the representativeness of the sample cannot be affirmed, the less so
since on some roads only a few surveys were filled out. For these reasons, the results should
be interpreted with caution. Table 17 summarizes the views of residents, as reported in the
report’s appendices.

Table 17 Effects of calming schemes on air quality as perceived by residents
Types of
Calming measures No change Don’t know Improvement Deterioration
roads
Local 27% 31% - 21%
Horizontal Collector 45% 28% 19% -
Arterial 41% 28% 18% -
Local 13% 55% 13% -
Vertical Collector 39% 24% 26% -
Arterial 48% 27% 12% 12%
Local 37% 18% 34% -
Horizontal and vertical Collector - - - -
Arterial 43% 30% 18% -
Note: No statistical significance test was mentioned, but the lines with underlined numbers indicate a greater proportion of
respondents having perceived an improvement, as opposed to a deterioration, in air quality; whereas those remaining in plain
text indicate the absence of or inability to establish such a relationship or its inverse.
Source of the data: Hemsing & Forbes, 2000.
According to these results, the majority of respondents observed no change in air quality or
did not know if there had been one. Nevertheless, for most road categories, more
respondents perceived an improvement in air quality than perceived a deterioration.
3.2.3.3 Summary of results for area-wide calming schemes
The documentary search led to the identification of two articles (Várhelyi, 2002; Owen, 2005)
belonging to the scientific literature and three studies (Boulter & Webster, 1997; Cloke et al.,
1999; Hemsing & Forbes, 2000) drawn from the grey literature, including one traditional
literature review (Boulter & Webster, 1997) and two studies that examined residents’
perceptions (Cloke et al., 1999; Hemsing & Forbes, 2000). These studies evaluate the
effects of area-wide calming schemes on air emissions and air quality. They use different
methodological approaches and are not all equally rigorous, but they present results which,
overall, agree with the recognized mechanisms of action. The key points to recall are the
following:
a) Variable effects on per vehicle emissions
The traditional literature review (Boulter et al., 1997) reports both increases and decreases
as well as a lack of change with respect to per vehicle emissions of the various pollutants
following the implementation of area-wide schemes, with the exception of NOx emissions for
which it reports no increases. One article (Owen, 2005) reports per vehicle increases in
emissions of all pollutants following implementation of the schemes and one study (Cloke et
al., 1999) concludes likewise, except that it reports a decrease in per vehicle emissions of
NOx. The only article (Várhelyi, 2002) that mentions statistical significance tests points to the
importance of the implementation context, and reports weak, non-significant increases in per
vehicle emissions for a mini-roundabout replacing an intersection without signals, but much
greater non-significant decreases for the replacement of an intersection controlled by traffic
lights. The differences between the results reported by the various studies are likely due, in
part, to the variable effects of area-wide schemes on driving speeds, speed variations and
time spent idling.

b) Relevance of traffic volumes for total emissions

The two studies (Cloke et al., 1999; Owen, 2005) that take into account increases and
decreases in traffic volume in the calmed areas report reductions for emissions of all
pollutants in the majority of areas where traffic volumes decreased, although no statistical
significance tests accompany these results. One article (Várhelyi, 2002) examining a scheme
involving mini-roundabouts points to the fact that interventions can have different effects on
vehicles driving on the various branches of an intersection. Thus, the effect of such
interventions on total emissions would depend on the amount of traffic on the various
branches.
c) Air quality
Despite the variations in per vehicle emissions and in overall emissions in an area, neither of
the two studies (Cloke et al., 1999; Owen, 2005) having measured the effect of calming
schemes on ambient air quality report significant changes for areas, once the variations at
control sites were taken into account. One article (Owen, 2005) suggests that the weak
contribution of road traffic in calmed areas to ambient concentrations of atmospheric
pollutants may, in certain cases, explain why an increase or a decrease in emissions caused
by an area-wide scheme does not necessarily lead to significant variation in measurements
of ambient air quality. However, the imprecision of the measuring devices used could also
explain the absence of significant results.
d) Perception
The two studies (Cloke et al., 1999; Hemsing & Forbes, 2000) that examined residents’
perceptions report that a majority did not observe any change in air quality following the
interventions.

None of the studies consulted compare the effects of area-wide strategies. Nevertheless, two
studies (Cloke et al., 1999; Owen, 2005) report reductions in emissions following the
implementation of schemes that reduced traffic volumes, while one article (Várhelyi, 2002)
implies that mini-roundabouts replacing traffic lights can, when they result in fewer speed
variations and less time spent idling, help reduce per vehicle emissions.
3.2.4 Conclusion of the section on air quality
Traffic-calming interventions are not usually promoted as a way of improving air quality, but
rather as a way of reducing the number of collisions, injuries and deaths. While, generally
speaking, they can be said to improve road safety, the literature is less abundant and less
conclusive concerning their effects on air pollutant emissions and ambient air quality.
Nevertheless, most studies indicate that the implementation of calming measures generally
leads to an increase in per vehicle emissions of the main air pollutants, even though certain
calming measures, such as mini-roundabouts, are more likely to reduce them in certain
contexts. Increases in per vehicle emissions can be explained by a decrease in speeds or an
increase in speed variations and time spent idling, whereas decreases in emissions can be

explained by the inverse phenomena. The studies also indicate that when a traffic-calming
intervention leads to a reduction in traffic volumes, this can compensate for an increase in
per vehicle emissions or even lead to an overall reduction in emissions at the intervention
site. According to the studies consulted, the residents of calmed areas do not always
perceive these changes. Despite increases and decreases in air pollutant emissions, the
studies do not report any significant variation in air quality in the vicinity of the interventions,
but this result may be due to the imprecision of measuring devices. The explanatory
mechanisms lead one to expect that the strategies most effective at reducing emissions at
intervention sites are those that reduce speed variations, idling times and traffic volumes at
those sites. According to the logic of these mechanisms, area-wide strategies, which often
explicitly aim to reduce traffic volumes in an area, seem to have an advantage over targeted
interventions, which do not generally seek to alter traffic volumes. However, area-wide
strategies that divert motorized traffic instead of reducing it (e.g., by promoting active
transportation) risk displacing pollutants rather than reducing pollutant emissions, and, in
certain contexts, this may increase health inequalities.
Effects of local traffic on air quality

Traffic-calming interventions, depending on the measures implemented and the context, can
lead to an increase or a decrease in emissions at intervention sites. While some studies have
sought to quantify these variations, only one article (Owen, 2005) has calculated the
contribution of these local emissions to ambient concentrations of atmospheric pollutants.
The article in question indicates that, for the area-wide schemes it evaluated, local emissions
contributed little to ambient concentrations of pollutants, which may explain why the studies
consulted (Boulter et al., 2001; Cloke et al., 1999; Owen, 2005) report no significant increase
or decrease in ambient air quality following an increase or decrease in these emissions. To
improve understanding of the health effects that may be linked to an increase or a decrease
in local emissions, it would be relevant to conduct studies to determine, in various contexts,
the portion of atmospheric pollutants that can generally be attributed to emissions on the
roads where most calming measures are installed, that is, on local streets lined with homes.
Accuracy of measuring devices

In a similar vein, the three studies (Boulter et al., 2001; Cloke et al., 1999; Owen, 2005)
having evaluated the effects of traffic calming on ambient air quality used rather imprecise
measuring devices (e.g., diffusion tubes). It would be relevant to conduct research using
more precise devices (e.g., continuous analyzers), to determine whether or not interventions
have a significant effect on ambient air quality.
Inequalities
The area-wide approach to traffic calming often aims to redirect a portion of traffic on local
streets toward the arterial network. Given that persons with low socioeconomic status tend to
be overrepresented among residents of these roads (Smargiassi et al., 2006), it would be
relevant to conduct research into the potential effects of area-wide strategies on inequalities
in exposure to air pollutants from road traffic.

Greening
Also, traffic calming often results in the recovery of space formerly devoted to automobile
traffic. This reorganization of public space often presents opportunities for greening which
can, moreover, affect ambient air quality (Beckett, Freer-Smith, & Taylor, 1999; Yang,
McBride, Zhou, & Sun, 2005) and the amount of greenhouse gases released into the
atmosphere (Nowak & Crane, 2002). It would be relevant to conduct research into this
potential opportunity for greening.
Figure 18 A “green” mini-roundabout

Source: SkyscraperPage Forum, 2011. Photographer:
SFUVancouver.
3.3 ENVIRONMENTAL NOISE

Environmental noise generally refers to sounds considered undesirable or harmful emitted by
any source, except noise at the industrial workplace (WHO, 1999). The principal effect of
environmental noise is the disturbance of sleep, which can lead to fatigue and feelings of
depression, and can lower performance levels. On a daily basis, it can interfere with
communication processes, concentration, memory and complex problem solving (WHO,
1999). Prolonged exposure to environmental noise can, notably, increase consumption of
tranquilizers and sleeping pills, psychiatric symptoms, psychiatric hospital admissions,
stress-related hormones, the risk of hypertension, obesity, and ischaemic heart disease
(WHO, 1999; WHO Regional Office for Europe, 2009). Since motorized traffic is one of the
main sources of environmental noise (WHO Regional Office for Europe, 2009), it thus
contributes to the deterioration of quality of life and to increases in the incidence and
prevalence of certain chronic diseases and mental illnesses. Canada does not seem to have
been unaffected by this problem, for it has been estimated that 1.8 million Canadians were
“very” or “extremely annoyed” by traffic noise in 2003 (Michaud, Keith, & McMurchy, 2005).
This said, various actions can be taken to help reduce environmental noise caused by traffic
or the population’s exposure to it. In particular, noise can be curtailed at its source (e.g., quiet
technologies, reduction of traffic volumes), mitigated (e.g., acoustic barriers, insulation of
homes) or distanced from populations (e.g., diversion of traffic). Although traffic-calming
strategies are primarily promoted as a way to reduce road collisions, injuries and deaths, it is
likely that they could also help reduce traffic noise and exposure to this noise, in particular,

by reducing traffic speeds and volumes, and by distancing a portion of traffic from sensitive
locations (e.g., homes, schools, etc.).
According to the studies consulted, traffic-calming strategies can influence motorized traffic
noise by acting on five mechanisms:
a) Reduction of vehicle speeds
Vehicle noise increases with speed. As Figure 19 shows, this association is stronger for cars
than for heavy vehicles (commercial vehicles and buses), whose noise is mainly generated
by the engine and the exhaust system, which does not vary much with speed, unlike the
noise caused by the friction of tires on pavement (Abbott, Tyler, & Layfield, 1995). Thus, by
reducing driving speeds, calming measures should mainly help reduce the amount of noise
generated by cars.
Figure 19 Increase in maximum vehicle noise

level according to speed for three
types of vehicles
Source: Abbott et al., 1995, p. 9.
b) Reduction of speed variations

Increasing the magnitude, the frequency and the speed of accelerations and decelerations
tends to increase the noise generated by accelerating motors and by brakes, for example.
Traffic-calming strategies can lead to speed variations when, for example, a driver is
navigating a speed hump, but they can also encourage driving at slow and constant speeds
when calming measures are close enough to discourage accelerating and decelerating
between them. Thus, the impact of calming strategies on vehicle noise levels tied to speed
variations is likely to depend on the configuration of calming schemes. Those who succeed in
reducing speed variations are likely to contribute to decreasing traffic noise.

c) Reduction of traffic volume

The effect of variations in traffic volume on noise is complex. The presence of fewer vehicles
on a road or in an area can lead, depending on the context, to higher driving speeds. These
can lead to an increase in the maximum noise level of vehicles. Inversely, heavier traffic can
slow vehicles down and thus decrease their maximum noise level. However, heavier traffic
can also lead to congestion and thus increase noise due to braking mechanisms, frequent
stops, etc. Considering another noise measurement, an increase in the number of vehicles
travelling on a road or in an area can also increase the average noise level (see Box 3 for a
discussion of these different measurements). The impact of calming strategies on
environmental noise attributable to a change in traffic volume therefore depends on the
indicators used to measure noise and on several other factors, such as the conditions
determining driving speeds and road capacities before and after calming.
d) Introduction of textured materials

Textured materials, such as paving stones, can increase noise from vehicles travelling over
them by causing their bodywork to vibrate, for example. The use of calming measures
incorporating such materials is thus likely to contribute to an increase in traffic noise.
Figure 20 Textured crosswalk

Source: www.flickr.com. Photographer: Richard
Drdul.
e) Introduction of vertical deflections

A large vertical deflection (e.g., speed hump) can increase suspension noises or noise from
objects carried in a trailer, for example. Strategies that incorporate calming measures
causing such deflections are thus likely to increase noise from vehicles affected by
vibrations. The impact of these calming measures therefore depends on their design and on
the type of vehicles travelling on the roads where they are used.
Most of the studies consulted that examine the effects of traffic-calming strategies on noise
generated by vehicles inferred these effects by referring to the five mechanisms above
without attempting to measure the effects. We have retained for purposes of analysis only
those studies that evaluated the effects of one or of a series of calming measures. As in the
preceding sections, we will first present the data on isolated interventions installed at specific
points on the road network, or on a single street or a test track, followed by the data on area-
wide calming schemes.

BOX 3 MEASUREMENTS OF ENVIRONMENTAL NOISE AND A FEW POINTS OF REFERENCE
There are several indicators that are useful for determining the effect of an intervention on
environmental noise. All of the studies presented in this document used indicators based
on the measurement of noise in A-weighted decibels or dB(A), which is a unit of
measurement weighted according to a filter, A, to take into account the way the human ear
responds to sound frequencies.
Four indicators were used in the studies consulted:

• LAmax: The maximum A-weighted sound pressure level. This indicator should be used to
measure a limited number of discrete sounds, such as the passage of a few cars at
night on a local street with little traffic (WHO, 1999; WHO Regional Office for Europe,
2009).
• LAeq T, LAnight, LAday: The A-weighted equivalent average sound pressure level for a time
period, T, or over an entire night or day. This indicator should be used to measure
relatively continuous noise, such as road traffic on a major artery (WHO, 1999; WHO
Regional Office for Europe, 2009).
• LA10: The A-weighted noise level which is exceeded 10% of the time for a given time
period. This indicator, less commonly used today, was widely used to measure the
relatively continuous noise of road traffic, but it is generally very strongly associated with
the noisiest isolated events, as measured by LAmax (WHO Regional Office for Europe,
2009).
• LA90: The A-weighted noise level which is exceeded 90% of the time for a given time
period. This indicator, less commonly used today, was used to measure background
noise, which excludes the noisiest isolated events, (WHO Regional Office for Europe,
2009).
To provide a few points of reference, the human ear begins to perceive increases or
decreases in noise levels beginning with a variation of 3 dB(A) (Direction de la santé
publique, 2006). During daytime and evening, an outside noise level of 50 dB LAeq,16h, in
residential areas, is associated with moderate annoyance, whereas a level of 55 dB
LAeq,16h is associated with high annoyance (WHO, 1999). A level of 55 dB LAeq,16h also
causes annoyance in school courtyards (WHO, 1999). At night, at a level of more than
30 dB LAnight (the equivalent average level, LAeq, taken at night–ideally–for a period of one
year), measured near the facades of homes, the effects on sleep (e.g., movements,
awakenings) begin to be observed (WHO Regional Office for Europe, 2009). Between
40 and 55 dB LAnight, health effects (e.g., environmental insomnia, hypertension) are
observed (WHO Regional Office for Europe, 2009). Above 55 dB LAnight, the sleep of a
sizable proportion of the population is disturbed, and cardiovascular effects become the
main health concern (WHO Regional Office for Europe, 2009). As regards noise from
discrete events, starting at 53 dB LAmax outside (based on a 21 dB(A) difference between
inside and outside), biological effects can be observed (e.g., movements, duration of sleep
stages), whereas starting at 63 dB LAmax, sleep quality is reduced (e.g., awakenings).

On the basis of these results, the World Health Organization recommends that European
countries aim for a threshold of 40 dB LAnight outside of homes and set an interim threshold
of 55 dB LAnight (WHO Regional Office for Europe, 2009).
In 1995, the Transport Research Laboratory published a report (Abbott et al., 1995)
evaluating the effects of various configurations of calming measures relying on vertical
deflection (wide and narrow speed cushions, round-top and flat-top speed humps) on noise
from one car, from three buses and from six heavy vehicles. The calming measures were
installed on a test track in London, U.K., and the maximum noise level before and after their
installation was measured using a microphone located 7.5 m from the road centre and at
1.2 m from the ground. The vehicles were driven on the test track before and after the
introduction of the calming measures at constant speeds typical of the average speed
observed on roads where the various calming measures had been installed. It is important to
note that the fleet of heavy vehicles selected was noisier than the average fleet used in the
country (metal suspension, panels subject to vibrations, etc.) and that the use of constant
speeds potentially led to an overestimation of noise from suspension systems and loose
metal and an underestimation of acceleration and deceleration noises. This stated, Table 18
summarizes the results obtained.
Table 18 Effects of calming measures on the maximum noise level of vehicles

moving at constant speeds
Calming measures Car Bus Heavy vehicles

Track without calming measure (before)
km/h 45 38 38
dB LAmax 71.0 73.7 79.7
Narrow speed cushions (after)
km/h 30 (-15) 34 (-4) 34 (-4)
dB LAmax 64.4 (-6.6) 74.7 (+1.0) 81.8 (+2.1)
Wide speed cushions (after)
km/h 22 (-23) 24 (-14) 24 (-14)
dB LAmax 62.3 (-8.7) 73.2 (-0.5) 87.6 (+7.9)
Round-top speed humps (after)
km/h 22 (-23) 18 (-20) 18 (-20)
dB LAmax 60.9 (-10.1) 69.8 (-3.9) 77.6 (-2.1)
Flat-top speed humps (after)
km/h 22 (-23) 18 (-20) 18 (-20)
dB LAmax 60.7 (-10.3) 70.0 (-3.7) 85.9 (+6.2)
Source of the data: Abbott et al., 1995.

At a constant speed, the effect of introducing vertical deflections was, thus, to reduce the
maximum noise from the car, to slightly increase or reduce that of the buses and to increase
that of the heavy vehicles, with the exception of the round-top speed humps, which reduced
the maximum noise levels of these vehicles. It is interesting to note that the presence of
round-top speed humps led to reduced maximum noise levels for all three types of vehicles.
On the basis of these results, a model with which to estimate the effects of calming
measures on ambient averaged noise levels (in dB LAeq) based on traffic composition was
constructed. The results of seven traffic flow scenarios are illustrated in Figure 21.
Figure 21 Estimated variations in noise levels after

the installation of traffic-calming measures
according to traffic composition
Sources: Abbott et al., 1995, p. 20 in Department for Transport, 2007, p. 61.
These results demonstrate that when traffic is composed entirely of cars, the installation of
calming measures with vertical deflections should result in a reduction of ambient noise
levels. However, the installed measures should generally increase ambient averaged noise
levels where heavy vehicles are present. Narrow speed cushions and round-top humps
would be preferable where heavy vehicles are present, because they would lead to
decreased noise levels even in the presence of 94% cars, 1% buses and 5% heavy vehicles
for narrow speed cushions, and of 79% cars, 20% heavy vehicles and 1% buses for round-
top humps.
An article by Campolieti & Bertoni (2009) examines the effects on environmental noise levels
of replacing an intersection controlled by traffic lights with a roundabout in Modena, Italy.
Noise levels were measured before and after the installation using three microphones: two
located on the roundabout and one located a few dozen meters away on a branch of the
intersection. Although the article specifies that the analyses take into account changes in
traffic volume and composition, it should be noted that the methodology used is not explicitly
described and that no statistical significance test is mentioned. This said, the article reports
averaged noise level reductions of 1 dB LAday and 2.5 dB LAnight.

3.3.2.2 Subjective study
An article by Morrison et al. (2004) evaluates the perception of residents living on a road in
Glasgow, Scotland, six months before and six months after the installation of several calming
measures (five pairs of speed cushions, parking bays, and two zebra crossings). Of the
750 residents randomly selected to fill out the postal questionnaire, 244 answered before the
intervention and 185 answered afterwards. Respondents were asked to evaluate their
feelings about problems related to their environment, using a scale of 1 to 7 (1 being the
happiest and 7 the unhappiest). Having compared the answers given before and after the
intervention, the authors found a significant improvement of 2.83 points in answers to the
question regarding traffic noise.
The documentary search led to the identification of two articles (Campolieti & Bertoni, 2009;
Morrison et al., 2004) belonging to the scientific literature, one of which examines residents’
perceptions (Morrison et al., 2004), and one study drawn from the grey literature (Abbott et
al., 1995). These studies evaluate the effects of individual calming measures on noise
produced by vehicles. While overall their results converge, by themselves, these three
articles have rather limited significance. The key points to recall are the following:
a) Variable effects of vertical deflections on the maximum noise level of vehicles
One study (Abbott et al., 1995) reports that the installation of calming measures with vertical
deflections reduces the maximum noise level of cars travelling over the deflections at a
constant speed, but that, with the exception of round-top speed humps, these measures tend
rather to increase the maximum noise levels of buses and heavy vehicles. These results are
attributable to reductions in noise levels caused by decreased speeds and to increased noise
from suspension systems and rattling metal when vehicles travel over vertical deflections, but
the results do not take into account the speed variations normally observed when this type of
calming measure is driven over. Nor are they accompanied by statistical significance tests.
b) Perception
One article (Morrison et al., 2004) reports that the residents of a calmed road were
significantly less unhappy about road traffic noise after the road was calmed.

Without providing conclusive evidence, one article (Campolieti & Bertoni, 2009) indicates that
replacing an intersection controlled by traffic lights with a roundabout could, by reducing
vehicle speeds and speed variations, reduce ambient averaged noise levels at the
intersection. Another study (Abbott et al., 1995) indicates that calming measures with vertical
deflections can be effective at reducing noise levels on roads where there are few or no
heavy vehicles.

evaluating the environmental impact of an area-wide traffic-calming scheme involving the
installation of several types of calming measures in the Leigh Park area of Havant, England.
Measurements of average driving speeds were carried out before and after the installations,
in addition to measurements of maximum noise levels (LAmax) by type of vehicle (light, heavy)
and measurements of noise levels that are exceeded 10% (LA10) and 90% (LA90) of the time
during the day and at night. Video cameras made it possible to associate maximum noise
levels with the various types of vehicles. The authors specify that the increase they
measured in the nighttime background noise level (LA90, 6h) was probably due to the fact that
the night on which they took the measurements after the intervention was particularly windy.
Table 19 summarizes the results obtained.
Table 19 Effects of calming measures on vehicle speeds and environmental noise
Light vehicles Heavy vehicles All vehicles
Average Average LA10, 18h LA90, 18h LA10, 6h LA90, 6h

speeds speeds (6:00-0:00) (6:00-0:00) (0:00-6:00) (0:00-6:00)
Calming measures (km/h) LAmax (km/h) LAmax day day night night
Mini-roundabout
(and speed -19.3 to -3.4 to -13.8 to +4.5 to
-4.7 +1.3 -1.1 +7.8
cushions on one -9.3 -1.2 -7.1 +6.2
arm of the junction)
-21.2 to -6.5 to -6.8 to -2.7 to -2.3 to +8 to
Speed cushions - -
-19.3 -5.2 -4.8 -1.0 -0.1 +9.4
Pedestrian refuge -7.4 -0.7 - - -1.9 +3.4 +2.6 +14.1
Raised
- - - - -3.6 -2.5 -3.6 +9.6
intersection
These results show that, in general, the strategy led to speed reductions for all vehicles, but
that above all, it reduced the maximum noise level of cars and increased that of heavy
vehicles. The authors of the report do not explain the increase in the maximum noise level of
heavy vehicles at the mini-roundabout, but, according to the logic of the mechanisms of
action identified at the beginning of the section, it is likely that the mini-roundabout made it
more difficult for heavy vehicles to manoeuvre, thus increasing their speed variations, or that
the presence of speed cushions (vertical deflections) resulted in noise from suspension
systems or bodyworks. With one exception, the intervention decreased the LA10 (the noise
levels exceeded 10% of the time), both during the day and at night. The authors believe that
the 33 to 35% decrease in traffic volume on the calmed roads brought about by the speed
cushions was responsible for a 2 dB LA10, 18h daytime reduction, whereas the 10% decrease in
traffic at the pedestrian refuge was responsible for a 0.5 dB LA, 18h daytime reduction. As
regards the measurements of background noise, that is, the noise level exceeded 90% of the
time, the daytime results are variable and the validity of the nighttime results is doubtful.

The article by Hyden & Várhelyi (2000) measures the averaged ambient noise levels at three
intersections before and after roundabouts were installed, within the context of a scheme
involving the installation of 21 mini-roundabouts in Växjö, Sweden. The scheme replaced one
intersection controlled by traffic lights and 20 intersections without signals with mini-
roundabouts. Although the article specifies that these intersections had daily traffic volumes
as high as 23,500 vehicles, it does not indicate, for the intersections studied, the volumes
before and after the interventions or the type of intersections replaced. The article concludes
that the three mini-roundabouts, by reducing average speeds and speed variations at and
between the intersections, were responsible for a reduction in noise levels at the three
intersections of 1.6 dB, 3.9 dB and 4.2 dB LAeq. There is no mention of statistical significance
tests for these results.
Carleton region presents the perception of residents of calmed roads and zones according to
the type of road on which they live (local, collector or artery) and the type of calming measure
installed (horizontal, vertical or a combination of both). The survey was conducted using
questionnaires distributed on a non-random basis to residents of already calmed roads.
Thus, the representativeness of the sample cannot be affirmed, the less so since on some
roads only a few surveys were filled out. For these reasons, the results should be interpreted
with caution. Table 20 summarizes the views of residents, as reported in the report’s
appendices.
Table 20 Effects of calming schemes on environmental noise as perceived by

residents
Calming measures Types of roads No change Don’t know Increase Decrease

Local 39% - 16% 31%
Horizontal Collector 51% 14% 19% -
Arterial 41% - 21% 19%
Local - - 36% 28%
Vertical Collector 47% 12% 37% -
Arterial 30% - 22% 39%
Local 42% 18% 32% 12%
Horizontal and
Collector - - - -
vertical
Arterial 48% 15% 18% -
respondents having perceived an increase in noise levels, as opposed to a decrease in noise levels; whereas the lines with bold
numbers indicate the inverse relationship and those remaining in plain text indicate the inability to establish such a relationship.
According to these results, the majority of respondents seemed not to have noticed any
change in ambient noise levels or did not know if there had been one. Nevertheless, for most
road categories, more respondents perceived an increase in noise levels than perceived a
decrease.

Within the context of the report on Leigh Park (Cloke et al., 1999), interviews before (n=151)
and after (n=150) implementation of the calming measures were carried out with groups of
residents, in particular, to determine their perception of noise nuisances in the area. The
report does not specify how the interviews were carried out or how participants were
selected. Moreover, only 113 residents participated in the interviews both before and after
installation of the calming measures. Due to these limitations, the results should be
interpreted with caution. After the intervention, the majority of respondents observed no
change either inside their homes (63%), or when walking outside (58%). Moreover, the
percentage of respondents who reported being “very much” or “quite a lot” bothered by noise
in their home (38% before/32% after) or when walking outside (33% before/36% after)
remained basically the same. Despite these results, significantly fewer respondents indicated
that traffic noise prevented them from opening their windows (53% before/28% after).
Significantly fewer people mentioned being bothered by noise (38% before/23% after) and by
screeching of brakes and tires (48% before/28% after) during the day. Although, prior to
implementation of the area-wide scheme, people residing where calming measures were
installed had expressed a fear of being more disturbed by noise from vehicles than residents
living between two calming measures, the report did not find that the distance of
respondents’ homes from the various calming measures significantly affected their answers
concerning noise nuisances.
The documentary search led to the identification of one article (Hyden & Várhelyi, 2000)
belonging to the scientific literature, and two reports (Cloke et al., 1999; Hemsing & Forbes,
2000) drawn from the grey literature. These studies evaluate the effects of area-wide calming
schemes on vehicle noise and residents’ perceptions. While overall their results converge, by
themselves, these three studies have rather limited significance. The key points to recall are
the following:
a) Reduction of maximum noise level of cars
One study (Cloke et al., 1999) reports reductions in the maximum noise level of cars at all
locations where calming measures were installed (mini-roundabout, speed cushions,
pedestrian refuge), but no statistical significance tests accompany these results.
b) Increase in maximum noise level of heavy vehicles

The same study (Cloke et al., 1999) reports an increase in the maximum noise level of heavy
vehicles at a mini-roundabout equipped with speed cushions; however, no statistical
significance tests accompany this result.
c) Reduction of ambient noise levels

Taking into account all noise sources, the same study (Cloke et al., 1999) reports decreases
in the noise levels exceeded 10% of the time (LA10) at nearly all locations where
measurements were taken, both during the day and at night, but the levels exceeded 90% of
the time (LA90) were more variable. One article (Hyden & Várhelyi, 2000) also reports
reductions in the noise levels at three intersections replaced with mini-roundabouts. These
results are not accompanied by statistical significance tests.

d) Perception
Two studies (Cloke et al., 1999; Hemsing & Forbes, 2000) report that the majority of
residents indicated they had not noticed any difference when asked if noise levels had
changed after their area was calmed. Nevertheless, one of these studies, having compared
the answers before and after, found that significantly fewer respondents said they were
disturbed by noise after the interventions (Cloke et al., 1999).

Without providing conclusive evidence, two studies (Cloke et al., 1999; Hyden & Várhelyi,
2000) support the expectation that a scheme involving the installation of mini-roundabouts
that encourage low, constant speeds, in an area where few heavy vehicles circulate, should
reduce the intensity of noise events and ambient averaged noise levels in that area. One
study (Cloke et al., 1999) also indicates that a scheme incorporating speed cushions, among
other measures, to reduce driving speeds and traffic volumes should produce a reduction in
the amount of noise generated by cars.
3.3.4 Conclusion of the section on noise pollution
Traffic-calming interventions are not usually promoted as a way of reducing environmental

noise, but rather as a way of reducing the number of collisions, injuries and deaths. While,
generally speaking, they can be said to improve road safety, the literature is less abundant
and less conclusive concerning their effects on vehicle noise. Nevertheless, most of the
evaluative studies indicate that the implementation of calming measures generally leads to a
reduction in the maximum noise level of cars and an increase in that of heavy vehicles, with
the reported effects on ambient noise levels being more variable. According to the studies
consulted, residents of calmed areas may or may not perceive these changes. According to
the logic of the mechanisms of action identified at the beginning of the section, the increase
in noise from heavy vehicles can be explained, in part, by the presence of vertical
deflections, and also by an increase in speed variations. As regards the reduction in the
amount of noise generated by cars, this can be explained, in part, by reductions in speed, in
speed variations and in traffic volumes. According to the logic of the mechanisms identified,
the strategies most effective at reducing traffic noise should be those that reduce speed
variations and driving speeds, while reducing traffic volumes and avoiding the use of vertical
deflections on routes frequented by heavy vehicles. Based on this same logic, area-wide
strategies would seem to have a theoretical advantage over targeted interventions. Indeed,
area-wide strategies often explicitly aim to reduce traffic volumes in an area and can be
designed to encourage constant speeds, whereas targeted interventions do not generally
seek to alter traffic volumes and are less apt to encourage constant speeds throughout an
area. However, area-wide strategies that divert motorized traffic instead of reducing it (e.g.,
by promoting active transportation) can contribute to an increase in noise levels elsewhere
on the road network, which, in certain contexts, may increase health inequalities.

Speed variations
The frequency and intensity with which vehicles navigating a calming measure accelerate
and decelerate should influence the effects of calming strategies on ambient noise levels.
Thus, it would be relevant to conduct research into the effects of the various strategies and
the calming measures they include on speed variations and the noise generated by cars and
by heavy vehicles.
Materials
Many calming measures can be constructed using more than one material: pedestrian
crossings can, for example, be built from asphalt or from paving stones, just as speed humps
can be fixed and made of asphalt or movable and made of polyurethane. It would be relevant
to conduct research into the effects of these materials on vehicle noise.
Inequalities
in exposure to environmental noise.
Greening
Since the presence of vegetation can help mitigate sound propagation (Fang & Ling, 2003),
traffic-calming strategies that offer opportunities for greening could help reduce
environmental noise generated by road traffic. Ewing (1999) has proposed this hypothesis,
but to our knowledge, no study has tested its validity. It would be relevant to carry out
research for this purpose.
3.4 ACTIVE TRANSPORTATION
Land-use and road-network planning that prioritizes motor vehicle travel over active
transportation (walking, cycling, etc.) encourages a sedentary lifestyle (WHO Regional Office
for Europe, 2000). Although their position is certainly open to debate, the WHO goes so far
as to suggest that the impact of traffic on active transportation constitutes its greatest
negative impact on health (WHO Regional Office for Europe, 2000). It has been
demonstrated that a sedentary lifestyle predisposes one to overweight and obesity, which
increases, in particular, the risk of developing cardiovascular disease, certain cancers, type 2
diabetes, hypertension and certain mental illnesses (Centers for Disease Control and
Prevention, 2010; Transport, Health and Environment Pan-European Programme, 2004;
Desjardins, D'amours, Poissant, & Manseau, 2008). According to the data from 2005, 47% of
Canadians are considered inactive, 35% of adults are overweight and 24% are obese
(Human Resources and Skills Development Canada, 2011a; 2011b). In 1995, the number of
premature deaths attributable to a sedentary lifestyle in Canada was estimated to be 21,000

(Katzmarzyk, Gledhill, & Shephard, 2000). In urban settings, a significant portion of car trips
cover short distances, 22 and given favourable conditions, these trips could be made on foot
or by bicycle, for example, thus contributing to more active lifestyles. Therefore, while traffic-
calming strategies are primarily promoted as a way to reduce road collisions, injuries and
deaths, they are also often proposed as a way to promote physical activity by helping create
an environment that encourages active transportation, whether for utilitarian or recreational
purposes.
According to the studies consulted, traffic-calming strategies can influence the proportion of
trips that involve active transportation (that is, the modal share) by means of four main
mechanisms:
a) Improvement of perceived safety
Perceived danger from motorized traffic is identified in several articles as a major factor, if
not the main factor, that discourages cycling (Pucher, Dill, & Handy, 2010; Pucher, Garrard,
& Greaves, 2011; Reynolds, Harris, Teschke, Cripton, & Winters, 2009; Jacobsen, Racioppi,
& Rutter, 2009; Pucher & Buehler, 2008). Concern about danger is more of a deterrent to
children, the elderly and women, and thus contributes to health inequalities (Pucher et al.,
2011; Jacobsen, 2003; Pucher & Buehler, 2008). The same concern also discourages
people from walking (Jacobsen et al., 2009) and parents from letting their children walk or
cycle to school (Direction de la santé publique, 2006). By thus deterring active transportation,
perceived danger risks giving rise to a feedback loop (see Figure 22) in which perceived
danger leads to a decrease in active travel, producing a concomitant increase in traffic
volume, which in turn causes roads to be perceived as increasingly dangerous. By reducing
traffic volumes and speeds, calming measures could help reverse this dynamic.
22
Morency et al. (2007) estimate, for example, that in the Montréal metropolitan region, 11.7% (n=862,000) of
trips carried out weekly cover distances of less than 1.6 km and that 55% of these are taken in motor vehicles.
In the United States, according to the authors of one article, 41% of all trips carried out in urban settings cover
distances of less than 3.2 km and 28% cover less than 1.6 km, with 89% of Americans using their car for trips
of between 1.6 and 3.2 km and 66% for trips covering less than 1.6 km (Pucher & Renne, 2003). In Europe,
according to WHO estimates, 30% of car trips cover distances of less than 3 km and 50% cover distances of
less than 5 km (WHO Regional Office for Europe, 2011).

Figure 22 The vicious circle linking active transportation to

perceived danger
Sources: Adapted from WHO Regional Office for Europe, 2002, p. 11, in Direction de la
santé publique, 2006, p. 82.
b) Increase of the relative speed of active transportation

Reducing the gap between the speeds of motorized and active travel can make active travel
for utilitarian purposes more attractive. According to the literature, traffic-calming strategies
can reduce this gap in three main ways: 1) by slowing down traffic (Pucher et al., 2010); 2)
by giving cyclists and pedestrians priority by installing facilities such as bike boxes, by
synchronizing traffic lights with respect to cycling speeds, creating “green waves” for cyclists
(Rietveld & Daniel, 2004) or by reserving exclusive traffic light phases for pedestrian
crossing; 3) by providing coordinated networks of facilities designed for active transportation
that reduce active travel distances and increase those of through traffic, as illustrated in
Figure 23 (Bassett, Pucher, Buehler, Thompson, & Crouter, 2008; Pikora, Giles-Corti, Bull,
Jamrozik, & Donovan, 2003; Pucher & Dijkstra, 2003).

Figure 23 Diagram of a road network that

is more permeable to active
means of transportation than
to motor vehicles
Dotted lines show routes restricted to active travel.
Source: Hummel, 2001, p. 30.
c) Reduction of perceived noise and atmospheric nuisances

Air and noise pollution caused by traffic can also deter people from walking or cycling by
making active travel less pleasant (Jacobsen et al., 2009). If traffic-calming strategies
succeed in reducing these irritants, they could encourage active travel.
d) Amelioration of aesthetics
Since the aesthetics of an area is one of the factors that determines the number of
pedestrians and cyclists that go there (Pikora et al., 2003), a traffic-calming strategy
designed to include high quality materials and to take advantage of the space recovered from
motorized traffic by greening roads could also help increase active travel (TDM
Encyclopedia, 2010b).
Most of the studies consulted that examine the effects of traffic-calming strategies on active
transportation inferred these effects by referring to these four mechanisms without attempting
to measure the effects. We have retained for purposes of analysis only those studies that
evaluated the effects of one or of a series of calming measures. As in the preceding sections,
we first present the data on isolated interventions installed at specific points on the road
network or along a single street, followed by the data on area-wide calming schemes.

BOX 4 THEORETICAL LIMITATIONS OF ISOLATED INTERVENTIONS FOR ENCOURAGING

ACTIVE TRANSPORTATION
Active transportation requires coordinated and interconnected networks on which it is

possible and pleasant to travel quickly and safely to multiple destinations (Pucher &
Buehler, 2008; Lee & Moudon, 2008). Except in the case where a network that is generally
suitable for active transportation already exists and where isolated interventions would add
missing connections or improve problem spots, it therefore seems unlikely, a priori, that
isolated interventions could have as great an effect on the volume of active travel as area-
wide strategies.
An article by Morrison et al. (2004) tabulates the number of pedestrians at three sites six
months before and six months after the installation of a series of calming measures (five
pairs of speed cushions, parking bays, and two zebra crossings) on the main road of a
neighbourhood in Glasgow, Scotland. The authors note that a difference between the
weather conditions during the first count (cloudy) and the second count (intermittent rain)
may have led to an underestimation of the overall increase in pedestrian activity observed six
months after the intervention. The presence of this confounding factor and the fact that only
two one-day counts were carried out limit the significance of the results of this study.
However, they are consistent with the self-reported significant increase in foot travel following
the intervention (see below). Table 21 presents the results of the counts according to
pedestrian age.
Table 21 Overall increase in the number of pedestrians following calming of a

road
Site 1 Site 2 Site 3

{CI 95%} {CI 95%} {CI 95%}
Children (0-16 years old) +18.0%* {+15.4 to +20.6} +44.1%* {+40.8 to +47.4} +40.0%* {+2.8 to +9.0}
Adults (16-60 years old) +12.3%* {+10.3 to +14.3} +54.9%* {+52.2 to +57.6} +11.4%* {+29.3 to +43.3}
Pensioner (60+ years old) +5.9%* {+2.8 to +9.0} +36.3%* {+29.3 to +43.3} -53.8%* {-59.3 to -48.3}

Source of the data: Morrison et al., 2004.
These results indicate a significant overall increase in pedestrian activity following calming of
the road. The study does not explain the variation among the three sites and does not
provide an explanation for the significant reduction in the number of pensioners walking
observed at site 3.

BOX 5 PARTICULAR IMPORTANCE OF SUBJECTIVE STUDIES TO UNDERSTANDING THE

EFFECTS OF TRAFFIC CALMING ON ACTIVE TRANSPORTATION
The effect of calming strategies on active travel is dependent on the decisions of residents
and road users who can choose to walk or cycle to destinations rather than use a motor
vehicle. It also depends on the same people choosing to walk or cycle for recreational
purposes. Since the desired effect is mediated by choice, it is likely to be influenced by the
perceptions of residents and road users regarding, in particular, safety, amenability
(aesthetics, air quality and ambient noise), and the time required to complete a utilitarian
trip using active transportation (relative speeds).
The article by Morrison et al. (2004), presented in the previous section, also evaluates the
perception of residents living on a road in Glasgow, Scotland, six months before and six
months after the installation of calming measures (five pairs of speed cushions, parking bays,
and two zebra crossings). Of the 750 residents randomly selected to fill out the postal
questionnaire, 244 answered before the intervention and 185 answered afterwards. They
were asked to evaluate their feelings about problems related to their environment, using a
scale of 1 to 7 (1 being the happiest and 7 the unhappiest). Having compared the answers
given before and after the intervention, the authors found significant improvements of
0.24 and 3.60 points in answers to the questions regarding safety for cyclists and drivers
respectively and a non-significant improvement of 0.76 points for pedestrians. They also
indicate a significant improvement of 2.19 points for questions regarding the ease of crossing
the road and significant improvements of 2.88 and 2.83 points for air quality and
environmental noise respectively. The improvement in feelings of safety and in the perceived
ease of crossing the road, as well as the reduction of perceived nuisances, may have
contributed to the significant increase in pedestrian activity observed six months after the
intervention. The second questionnaire asked respondents to evaluate the changes in their
travel behaviour that were attributable to the intervention. Table 22 presents the results
reported in the article, which, however, does not indicate whether some residents reported or
would have reported having travelled less by active means.
Table 22 Self-reported changes in the travel behaviour of residents of a calmed

road
Travel behaviour {CI 95%}

More walking 20.0%* {14.1 to 25.9}
More cycling 3.8%* {0.8 to 6.8}
Allowing children to play outside more 11.8%* {6.7 to 16.9}
Allowing children to walk more 12.5%* {7.2 to 17.8}
Allowing children to cycle more 11.6%* {6.6 to 16.6}
Source of the data: Morrison et al., 2004.

The results of the second questionnaire seem to indicate that the significant increase in
pedestrian activity observed on the road is due to an increase in active travel and not to
increased use of the road resulting from decreased use of other roads. However, in the
absence of data on those who reported or would have reported limiting their active travel,
these results must be interpreted with caution. The article also reports a significant
improvement in the self-reported physical health of residents six months after the
intervention, but no significant change in their mental health.
Watkins (2000) gave a conference presentation on a study evaluating the perceived safety of
residents of a street in Cambridge, Massachusetts (U.S.) on which curb extensions, a raised
intersection, a pedestrian crossing and improved crosswalk markings had replaced traffic
lights. It should be noted that the study indicates neither the sampling design used nor the
number of respondents. The results, presented in Table 23, should therefore be interpreted
with caution.
Table 23 Residents’ perception of safety following calming of their street
Better Worse
Pedestrians 57% 13%
Cyclists 33% 8%
Motorists 46% 10%
Note: No statistical test for evaluating the significance of changes is mentioned.
Source of the data: Watkins, 2000.
The results indicate that more residents perceived an improvement in safety, for all types of
road users, and in particular, for pedestrians and motorists. However, the study does not
indicate whether these changes are significant.
Gibbard et al. (2004) produced a report for which they surveyed 393 relatively experienced
cyclists using a questionnaire posted online on the websites of cycling associations based in
England. The objective of the questionnaire was to determine how safe cyclists feel when
they encounter calming measures involving road narrowing or horizontal deflection, such as
chokers or curb extensions. Their results indicate that 78.4% of the cyclists who responded
find these calming measures problematic, whereas only 17% are unaffected by them and
3.8% find them helpful. Many respondents said they stop at road narrowings (over 40%), that
they sometimes move onto the sidewalk (46.8%) and that they sometimes choose another
route to avoid them (48.6%). From among a list of phenomena resulting in road narrowings
(various traffic-calming measures, parked vehicles, bus stops…), respondents identified
pedestrian refuge islands or raised medians (38.9%), as well as chicanes (8.9%) as
problematic. According to respondents, they perceive narrowings as dangerous because of
the greater proximity of moving vehicles, especially heavy vehicles. Indeed, whereas only
5.3% of relatively experienced cyclists indicated feeling stressed or intimidated when
negotiating this type of calming measure in the presence of motorcycles, this percentage
rose to 33.6%, 46.6%, 59.8%, and 61.6% in the presence of cars, buses, light vans, and
medium or heavy trucks, respectively. It should be noted that women indicated feeling more
stressed or intimidated than men. Since it is often possible to adapt the configuration of

calming measures to cyclists’ needs, as shown by Figure 24, it would be relevant to

determine if cyclists feel differently about narrowings or horizontal deflections that do not
force them to go closer to motor vehicles. Unfortunately, the report does not allow for this
distinction.
Figure 24 Two road narrowings configured differently

The choker on the left forces cyclists to come closer to moving vehicles, but the one
on the right does not.
Sources: On the left, http://www.pedbikeimages.org. Photographer: Dan Burden. On
the right, http://www.cyclestreets.net. Photographer: unknown.
The documentary search led to the identification of one article that had both objective and
subjective dimensions (Morrison et al., 2004) belonging to the scientific literature, and two
subjective studies (Watkins, 2000; Gibbard et al., 2004) drawn from the grey literature. These
studies evaluate the effects of individual calming measures on the volume of active travel
and the perceived safety of public road users. While their results converge, by themselves,
these three studies have rather limited significance. The key points to recall are the following:
a) Increase in active transportation
One article (Morrison et al., 2004) reports a significant increase in the number of pedestrians
using a calmed road and significantly more residents who reported walking or cycling more
and letting their children play outside, walk, and cycle more, after the interventions.
b) Perceived safety
Two studies (Watkins, 2000; Morrison et al., 2004) indicate that the residents of calmed
roads perceived an improvement in safety, for the various types of road users (drivers,
cyclists and pedestrians), following the interventions. However, another study (Gibbard et al.,
2004) reports that cyclists feel less safe in the presence of calming measures that introduce
road narrowings or horizontal deflections forcing them closer to moving vehicles.

The studies consulted do not allow for a comparison of the effectiveness of isolated calming
strategies at encouraging active transportation.

BOX 6 THEORETICAL ADVANTAGES OF AREA-WIDE INTERVENTIONS FOR ENCOURAGING

ACTIVE TRANSPORTATION AND METHODOLOGICAL LIMITATIONS
A priori, area-wide strategies seem better adapted to influencing the volume of active travel
than targeted interventions, because area-wide measures can contribute more to the
establishment of coordinated and interconnected networks on which it is possible and
pleasant to travel quickly and safely to multiple destinations. However, authors seeking to
empirically evaluate the impact of area-wide calming strategies encounter, in particular, the
following three methodological difficulties:
a) Threshold effects
Being dependent on coordinated and interconnected networks, the volume of active travel
is likely to progress in accordance with the attainment of thresholds, with more significant
increases achieved once a network reaches critical density and range, and not linearly,
with each block of cycle track or each speed hump added, for example. This implies that
evaluating a network being implemented in stages before it has reached a critical threshold
could result in failure to measure an increase in active travel, even if the network were one
step away from producing a significant increase in active travel (Pucher et al., 2010).
b) Synergies
One literature review concludes that the effect of various policies encouraging cycling
depends on the synergies between them (Pucher et al., 2010). An area-wide strategy for
calming traffic on residential streets coupled with the installation of cycle tracks on main
arteries and bike parking facilities at various destinations, for example, is likely to have a
much greater effect than the sum of each of these interventions taken separately. Based
on 14 case studies, the literature review produced by Pucher et al. (2010) leads to the
conclusion that the simultaneous presence of measures derived from a range of policies
promoting cycling, which includes area-wide calming strategies, can considerably increase
the volume of bicycle travel. However, methodologically it is difficult, if not impossible, to
precisely determine the specific effect of an area-wide strategy in such a context, for it
would be necessary to separate its effect from that of the other interventions, which likely
produce their own effects, and also to measure the synergistic effects (Pucher et al.,
2010).
c) Temporality
It is often the case that area-wide calming strategies are implemented gradually over a
period of several years, especially when they extend over a large geographic area. This
temporal dimension requires that evaluative studies be able to control for possible
confounding factors (underlying trend, nearby interventions on the road network) over long
periods of time.

The documentary search led to the identification of four studies having evaluated the effects
of calming measures on geographic areas comprising more than one road. Only one of these
focuses explicitly on calming measures that were systematically planned and implemented
on a network of roads. This study is presented first. The three studies that do not specify the
logic of the interventions examined are presented next.
and after (n=150) the implementation of calming measures were carried out with groups of
residents, in particular, to shed light on their travel behaviour. The report does not specify
how the interviews were carried out or how participants were selected. Moreover, only
113 residents participated in the interviews both before and after the intervention. Due to
these limitations, the results should be interpreted with caution. Table 24 summarizes the
results of the survey according to the mode of travel used to reach various destinations.
Table 24 Modal share of residents’ trips, by destination, before (n=151) and after
(n=150) implementation of an area-wide calming strategy
a
On foot By bike By bus
Destination Before After Before After Before After
Businesses 42% 27% 1% 1% 7% 11%
Work 6% 5% 1% 1% 1% 2%
Friends’ homes 27% 18% 1% 0% 13% 15%

a
Travel by bus is reported here because it has been shown that half of public transport users walk more than 19 minutes per
day in the United States going to and from service points (Besser & Dannenberg, 2005), while public transport users in the
Greater Montréal region perform, on average, 25% of the recommended amount of daily activity going to and from service
points (Morency, Trépanier, & Demers, 2011).
Note: No statistical test for evaluating the significance of “before-after” differences is mentioned.
These results indicate a slight decrease in foot travel and a slight increase in travel by bus
after the area was calmed.
One article (Carver, Timperio, & Crawford, 2008) evaluates the impact of the road
environment on the physical activity of children (5 to 6 years old) and adolescents (10 to
12 years old) in Melbourne, Australia. The children (n=295) and adolescents (n=919) were
recruited from 19 elementary schools. Parents were asked to report on the travel habits of
their children and adolescents were asked to self-report. Moderate or vigorous physical
activity outside of school hours was recorded using accelerometers. The road environment
within an 800-metre radius of their homes was audited. The number of speed humps, the
number of gateways (installed at the entrance to residential or calmed zones, for example)
and the number of road narrowings, among other things, were determined and these were
geocoded using a geographic information system (GIS). Multiple regression analyses were
carried out to search for significant associations between elements of the road environment
and (1) children and adolescents making at least seven walking or cycling trips per week and
(2) moderate or vigorous physical activity. It should be noted that the choice of seven or
more trips as a threshold could obscure the effects of the road environment on less frequent
travel and that the article’s analyses do not take into account the proximity of other facilities,

such as parks or businesses. This said, the analyses did not reveal any significant
associations between the presence of calming measures and (1) children making at least
seven walking or cycling trips per week and (2) their level of physical activity. As regards
adolescents, the presence of an average number of speed humps (2 to 7) was significantly
associated (OR: 0.38, CI 95%: 0.15 to 0.97) with a decreased likelihood of boys making at
least seven walking or cycling trips per week, whereas a high number of speed humps (8 to
99) was significantly associated (OR: 2.95, CI 95%: 1.34 to 6.51) with an increased likelihood
of girls making as many trips. The number of speed humps present proved to be significantly
and positively associated (r=0.210) with moderate to vigorous physical activity for boys
during the evening, but negatively associated (r=-0.073) with physical activity for girls before
school hours. As regards the number of road narrowings present, this was significantly and
negatively associated (r=-5.197) with physical activity for boys during weekends.
Kamphuis et al. (2008) published an article evaluating potential significant associations

between various characteristics of the road environment in Melbourne, Australia, and the
recreational cycling behaviour of residents. Of the 4170 census collector districts less than
20 km from downtown Melbourne, 50 were randomly selected from samples stratified
according to percentage of low-income homes. In each of these districts, covering an
average area of 0.34 km2, an audit was conducted of characteristics of the road environment
within a 400 m radius of a randomly selected home. Based on the characteristics observed,
three notable categories emerged: 1) traffic-calming measures influencing traffic speeds or
volumes (e.g., speed humps); 2) crossing aids for pedestrians (e.g., raised medians,
pedestrian refuges, curb extensions); 3) cycle tracks or bike lanes installed on roads. It
should be noted that, referring to the examples provided in the article, it seems that calming
measures that create horizontal deflections have been classified as crossing aids for
pedestrians and calming measures that create vertical deflections have been referred to as
calming measures. The article, however, does not explicitly make this distinction. It is also
worth noting that, although cycle tracks or bike lanes may be introduced to calm traffic by
narrowing roads (Macbeth, 1998), the article does not indicate why they were introduced or
how they changed the road environment. Randomly selected residents (n=4005) of these
districts were mailed a questionnaire asking whether they had cycled during the past month
for recreational purposes. The responses from valid questionnaires (n=2203) and the results
of the audit were analyzed using multilevel logistic regression models to identify significant
associations. Calming measures were significantly and positively associated with
recreational cycling (OR: 2.90, CI 95%: 1.19 to 7.02), as were cycle tracks and bike lanes
(OR: 5.40, CI 95%: 1.29 to 22.60). As regards calming measures identified as crossing aids
for pedestrians, no significant association was found (OR: 0.73, CI 95%: 0.30 to 1.78).
An article focused on walking behaviour and physical activity evaluates the influence of
calming measures on the travel behaviour of residents of Minneapolis-St. Paul, in Minnesota,
in the United States (Forsyth, Hearst, Oakes, & Schmitz, 2008). Using a sampling design
stratified according to median street block length and residential density, 36 square areas of
805 x 805 m were randomly selected, and 715 residents of these areas were recruited for the
study. Streets including at least one calming measure were identified in each of the areas
studied, but the concept of traffic calming used in the article excluded devices that act on
volume (e.g., forced-turn islands) as well as some measures that can affect road width (such

as bike lanes). The travel behaviour of residents was measured for a period of seven days
using accelerometers and self-completed travel diaries. The data gathered were analyzed
using multilevel logistic regression. Positive significant associations were found between the
percentage of street blocks with calming measures and (1) daily walking distances
(r=0.3629) and (2) daily utilitarian walking distances (r=0.3674). No association was found
involving recreational walking or physical activity in general.
and after (n=150) the implementation of calming measures were carried out with groups of
residents, in particular, to determine their perception of the area. The report does not specify
how the interviews were carried out or how participants were selected. Moreover, only
113 residents participated in the interviews both before and after installation of the calming
measures. Due to these limitations, the results should be interpreted with caution. After the
intervention, significantly fewer respondents reported being “very much” or “quite a lot”
bothered by vehicles travelling at high speeds (83% before/55% after) or by traffic volumes
(74% before/62% after). Also, significantly fewer respondents reported being “very much” or
“quite a lot” bothered by the danger or difficulty of crossing the street on foot (81%
before/57% after) or by the danger posed to children (93% before/70% after). However, the
majority of respondents had perceived no change with respect to air quality (68%) or noise
nuisances (58%) when out walking after the interventions. Even though the mechanisms
identified at the beginning of the section would lead one to expect an increase in active
transportation, the changes in residents’ perceptions did not produce an increase in the
number of respondents who reported having opted for active transportation.
Carleton region reports on residents’ perception of the safety of calmed roads and zones,
according to the type of road on which they live (local, collector or artery) and the type of
calming measure installed (horizontal, vertical or a combination of both). The survey was
conducted using questionnaires distributed on a non-random basis to residents of already
calmed roads. Thus, the representativeness of the sample cannot be affirmed, the less so
since on some roads only a few surveys were filled out. For these reasons, the results should
be interpreted with caution. Table 25 summarizes the views of residents, as reported in the
report’s appendices, regarding the safety of pedestrians and cyclists following the
interventions.

Table 25 Effects of an area-wide strategy on the safety of pedestrians and cyclists

as perceived by residents
Safety of pedestrians
Safety of pedestrians walking on the
Calming Types of crossing the street sidewalk Safety of cyclists
measures roads % % %
a a a
NC Incr. Decr. NC Incr. Decr. NC Incr. Decr.
Local 31 48 14 34 43 17 24 25 28
Horizontal Collector 49 26 20 37 33 21 19 - 45
Arterial 32 25 29 >50 27 23 - - >50
Local 37 46 18 31 46 14 36 32 13
Vertical Collector 36 29 24 27 22 29 34 - 24
Arterial 25 62 - 25 56 12 - 49 15
Local 44 26 14 51 27 9 27 16 24
Horizontal
Collector - - - - - - - - -
and vertical
Arterial 32 45 13 37 39 9 29 18 26
a
NC = “no change” perceived.
respondents having perceived an increase (“Incr.”) in safety, as opposed to a decrease (“Decr.”) in safety; whereas the lines
with bold numbers indicate the inverse relationship and those remaining in plain text indicate the inability to establish such a
relationship.
Overall, it seems that more residents perceived an increase in pedestrian safety, than
perceived a decrease in safety, whereas the inverse seems to be the case for cyclist safety.
The report’s appendices also contain information about residents’ perceived changes in their
travel behaviour. In general, residents did not think they had changed their travel behaviour.
The documentary search led to the identification of three articles (Carver et al., 2008;
Kamphuis et al., 2008; Forsyth et al., 2008) belonging to the scientific literature, and two
reports (Cloke et al., 1999; Hemsing et Forbes, 2000) drawn from the grey literature. These
studies evaluate the effects of area-wide calming schemes on the volume of active travel and
on public road users’ perception of safety. By themselves, these studies have limited
significance, in particular, because of their small number and their sometimes divergent
results. The key points to recall are the following:
a) Variable effects on active transportation
One article reports that the presence of calming measures close to home is associated with a
significant increase in recreational cycling (Kamphuis et al., 2008) and another (Forsyth et
al., 2008) that it is associated with a significant increase in walking to utilitarian destinations,
but that it is associated with neither a significant increase nor a significant decrease in
recreational walking. A third article (Carver et al., 2008) reports that the presence of calming
measures in the immediate environment of children (5-6 years old) is not significantly
associated with their travel behaviour, but that significant positive and negative associations
with the travel behaviour of adolescents boys and adolescent girls respectively can be

observed. One study (Cloke et al., 1999) examining the implementation of an area-wide
scheme reports a slight decrease in trips made on foot and a slight increase in trips made by
bus, but no statistical significance tests accompany these results.
b) Variable effects on physical activity

One article (Carver et al., 2008) reports that the presence of calming measures in the
immediate environment of children (5-6 years old) is not significantly associated with their
level of physical activity, but that significant positive and negative associations with the levels
of physical activity of adolescents boys and adolescent girls respectively can be observed.
Despite having observed a significant positive association between the presence of calming
measures near the home and utilitarian walking distances, another article (Forsyth et al.,
2008) found no association with physical activity in general.
c) Variable effects on perceived safety

Two studies (Hemsing & Forbes, 2000; Cloke et al., 1999) report that residents of calmed
areas perceive calmed roads as being safer for pedestrians. One of these (Cloke et al.,
1999) also reports that children were thought to be safer on calmed roads, whereas the other
(Hemsing & Forbes, 2000) indicates that cyclists were thought to be less safe. The study
(Cloke et al., 1999) whose results are accompanied by statistical significance tests reports
significant improvements after the interventions.
d) Promising area-wide strategies

The studies consulted do not allow for a comparison of the effectiveness of the various
strategies for encouraging active transportation. Without providing conclusive evidence, one
article (Kamphuis et al., 2008) nevertheless points to the benefits of using cycle tracks and
bike lanes on roads to encourage bike travel. These measures can be used, among other
things, to decrease the width of lanes devoted to motorized traffic (Macbeth, 1998).
3.4.4 Conclusion of the section on active transportation
Although traffic-calming interventions are mainly promoted as a way of reducing the number
of collisions, injuries and deaths, they are also frequently proposed as a way to encourage
active transportation. However, although the mechanisms of action described support the
logic of intervention, especially with regard to area-wide strategies, which can be designed to
establish coordinated, safe and pleasant networks for walking or cycling travel, the methods
and the rigorousness of the evaluative studies consulted, their number and their results do
not support the conclusion that traffic-calming interventions encourage active transportation.
Nevertheless, traffic calming seems to improve all road users’ perception of safety, with the
exception of cyclists, who report feeling less safe in the presence of certain calming
measures (e.g., road narrowings). Since traffic calming, and in particular the area-wide
approach, is usually integral to the policies of cities that have succeeded in increasing the
modal share of public and active transportation and at decreasing that of cars (Pucher et al.,
2010), it seems likely that the existing lack of evidence is partly due to the methodological
difficulties described in Box 6.

Travel behaviour and perception

In general, more research is needed to better document the effects of traffic-calming
interventions on travel behaviour and on the perception of various public road users.
Ideal strategy
According to Pucher & Buehler (2008), the ideal strategy for encouraging bike travel involves
the calming of residential neighbourhoods in a manner that allows motor vehicles and
cyclists to safely coexist, coupled with the installation of cycle tracks on arteries to separate
drivers and cyclists. In light of certain European experiments that reflect this approach (in
Amsterdam and in Copenhagen, in particular), this type of intervention seems promising;
however, based on the results of the literature consulted, its effects on bike travel remain to
be demonstrated.
Special needs of certain users

According to some authors (Bahar et al., 2009; Litman, 1999; Gibbard et al., 2004), certain
calming measures, such as road narrowings and roundabouts, can prove problematic for
some public road users, such as cyclists and visually-impaired pedestrians. For example,
when encountering calming measures that narrow roads, cyclists may be forced closer to
moving vehicles. Visually-impaired pedestrians, for their part, may find it difficult to find the
right moment to cross a roundabout where the traffic is quasi continuous. The same authors
suggest that these problems can usually be attributed to configurations of calming measures
that do not take into account the needs of these users, whereas there exist alternative
configurations worthy of consideration. Thus, it would be relevant to compare the effects of
problematic configurations and of existing solutions on active transportation and on various
users’ perception of safety.
Inequalities
linked to the travel behaviour of these persons.

4 CONCLUSION OF THE LITERATURE REVIEW

4.1 REVISITING THE WORKING PROPOSAL
The literature review was organized so as to allow for comparison of the effects on four
health determinants of strategies based on the two approaches to traffic calming,
implemented in an urban setting. In the introduction, we stated that area-wide strategies
likely had more potential for improving population health than strategies based on the black-
spots approach, because they favour interventions implemented on a larger geographic
scale (networks comprising more than one street), they are often driven by broader
objectives and goals (road safety and improvement of the living environment) and they tend
to involve both the main mechanisms of action associated with calming measures (reduction
of traffic speeds and of traffic volumes) rather than just one. A detailed examination of the
relationship between the mechanisms of action and the health determinants being
considered revealed another potential advantage of the area-wide approach. By intervening
systematically in an area, they seem a priori better suited to encouraging driving at constant
speeds in that area. The reduction of speed variations has, in turn, been identified as a
transversal mechanism likely to simultaneously reduce collisions, air emissions and traffic
noise, which, by acting on public road users’ perceptions, is also likely to encourage active
transportation. Thus, in theory, it seems that area-wide traffic-calming strategies that seek to
reduce traffic volumes and encourage driving at low and constant speeds offer the most
potential for improving population health.
As regards the evaluative studies consulted and analyzed, these allow one to conclude that,
in general, traffic calming is an effective way to reduce the number of collisions, injuries and
deaths. This holds true both for interventions that target high risk “black spots” and those that
cover areas comprising more than one street. Probably because traffic calming is mainly
promoted as a way to improve road safety, the literature examining this aspect is abundant
and conclusive. While this is less true of the literature on the effects of traffic calming on air
pollutant emissions and ambient air quality, the literature does indicate that calming
measures, whether part of a black-spots or an area-wide approach, generally lead to an
increase in per vehicle pollutant emissions, even though certain measures, such as mini-
roundabouts, have more of a tendency to reduce them. However, the literature also indicates
that area-wide strategies that reduce traffic volumes can compensate for an increase in per
vehicle emissions or even reduce total emissions in the intervention area. Despite these
increases and decreases in emissions, the studies do not report any significant variation in
air quality in the vicinity of interventions, regardless of whether these are part of a black-
spots or an area-wide strategy. As regards effects on nearby residents’ perception of air
quality, the results of studies are variable. With regard to the literature evaluating the effects
of traffic calming on traffic noise, it is also less abundant and conclusive than that concerning
road safety. Nevertheless, it indicates that the implementation of calming measures generally
leads to a reduction in the maximum noise level of cars and an increase in that of heavy
vehicles, regardless of whether the measures are part of an area-wide or a black-spots
approach. Studies report more variable results concerning the effects of interventions on
ambient noise levels and on nearby residents’ perception of noise levels. As regards the
effects of traffic calming on active transportation, the evaluative literature does not support

the conclusion that either area-wide or isolated interventions encourage this, even though, in
general, traffic calming seems to improve all road users’ perceptions of safety, with the
exception of cyclists, who report feeling less safe in the presence of certain calming
measures (e.g., road narrowings).
While the mechanisms of action of the calming measures used in both strategic approaches
seem to confer a theoretical advantage on the area-wide approach, it is important to stress
that none of the studies consulted was specifically designed to measure and compare the
effects of the two approaches to traffic calming. Within the current literature, only the results
concerning the effects of traffic calming on one determinant, “air quality,” allow for a direct
comparison. With respect to this, it is possible to conclude that the area-wide approach, by
reducing traffic volumes in an area, can effectively reduce total emissions produced there,
whereas the black-spots approach has more of a tendency to increase total emissions,
except when it involves the use of certain calming measures, such as mini-roundabouts.
However, it must be recalled that the interventions evaluated were conducted in specific
contexts and that not all their effects are known, not even all their effects on air quality. In
fact, it is not known if the area-wide strategies evaluated simply shifted pollutants elsewhere
along with diverted traffic, if they diverted traffic in a manner that reduces overall exposure to
pollutants, or if they reduced absolute traffic volumes (by increasing the modal share of
active transportation, for example). Thus, in general, the results of the evaluative studies do
not lead to the unequivocal conclusion that one approach to traffic calming has more
potential than the other to affect the four health determinants examined. On the other hand,
the results do not call into question the logic underpinning the theoretical advantage of area-
wide calming measures aimed at reducing traffic volumes and encouraging driving at low,
constant speeds.
4.2 CONTRIBUTIONS AND LIMITATIONS OF THE LITERATURE REVIEW
The value of this literature review is threefold:

1) It makes it possible to predict, based on the mechanisms of action described and the
results of the evaluative studies, the positive and negative effects on health determinants
of certain combinations of calming measures (from the wide range available) and
intervention contexts (e.g., replacement of traffic lights versus replacement of stop signs).
These combinations serve as points of reference for the development of traffic-calming
intervention strategies that promote health, at a time when this type of intervention in the
built environment seems to be gaining in popularity in the country’s large urban centres.
2) It brings together in a single document the principal mechanisms of action and the results
of evaluative studies focused on the effects of the two broad approaches to traffic calming
on four health determinants, whereas most synthesis documents do not distinguish
between the two approaches and address only the number and severity of collisions.
3) It sets out the strengths and weaknesses of the existing literature on the effects of traffic-
calming interventions on population health determinants.

Limitations:
1) The working proposal of the literature review made it necessary to classify the evaluative
studies, which could not always be done on the basis of explicit references to the logic of
interventions or on the basis of a detailed description of the traffic-calming interventions
evaluated. Since the studies that failed to provide these details were not eliminated from
the literature review, it is possible that some of them were misclassified.
2) To obtain an overview of the effects of traffic-calming interventions, instead of just
studying the effects of individual interventions carried out in specific contexts, it was
necessary to group together diverse interventions implemented in various contexts–which
can make generalization hazardous.
3) The effects reported in this literature review apply to health determinants and do not
correspond directly to effects on population health.
4) With only two exceptions (Zein et al., 1997; Hemsing & Forbes, 2000), the interventions
evaluated by the studies examined in this literature review were implemented in countries
other than Canada.
4.3 RESEARCH GAPS
In addition to the research gaps specific to each of the four health determinants covered in
the literature review and described at the end of each section, certain factors would be
worthy of study:
a) Canadian research
The vast majority of the empirical studies referred to in this literature review examine
interventions implemented in European countries or in the United States. Only one article
(Zein et al., 1997) and one report (Hemsing & Forbes, 2000) present evaluations of calming
strategies implemented in Canada. Although these other countries were deemed sufficiently
similar to Canada for the results of the studies to be applicable and relevant here, it would be
interesting to see studies that document Canadian traffic-calming interventions, within their
implementation contexts, and evaluate their effects.
b) Methodology
The majority of the empirical studies consulted sought to establish causal relationships
between traffic-calming strategies and the four health determinants covered in this literature
review. However, these causal relationships are often very complex. In fact, because the
effects of strategies depend on the specific characteristics of interventions and their contexts,
it is often difficult to establish, measure, and make generalizations about such relationships.
Thus, there exists not only a need for evaluation methods adapted to these difficulties, but
also for methods that can enhance understanding of the dynamic relationships between the
built environment, the populations that live in it and their lifestyles (mode and frequency of
travel, perception of facilities for walking and cycling and the safety of doing so, etc.) without
reducing complex dynamics (co-determination of relationships, threshold and synergy
effects, feedback loops) to linear causal relationships. Without proposing a method, the
report by the Groupe de recherche Ville et mobilité (2008) on children using active travel to
go to school provides an interesting example of research that sheds light on such dynamics

by simultaneously examining the effectiveness of various interventions that promote active

transportation, the evolution of policy and legal contexts that have either encouraged or
discouraged the active travel of young Quebeckers, the system of actors closely or
tangentially involved in how these children travel to school, and parents’ perception of their
children’s mobility.
c) Accuracy of studies
Many of the studies consulted do not precisely describe the interventions evaluated (traffic-
calming measures and strategies) and do not mention whether other interventions modified
the environment at the same time (e.g., demand management policy) or if other relevant
contextual changes may have taken place (e.g., increased fuel costs, recession, etc.). Future
studies should attempt to clearly identify these relevant factors and to take them into account
during data collection and analysis.
d) Temporality
Although several authors mention the importance of not evaluating the impact of an
intervention immediately after implementation, so as to allow public road users time to adjust
their behaviour to the new environment, none of the studies specifically documented the
medium and long-term impacts of calming strategies on the four health determinants covered
in this literature review. It would be relevant to research, for example, whether the
effectiveness of calming strategies at reducing collisions and their severity tends to increase,
decrease or remain the same after a certain number of months or years.
e) Policy context
Traffic-calming interventions fit within a context of diverse municipal policies which they may
affect and which, inversely, may determine which interventions are favoured. It would be
relevant to document how traffic-calming policies can fuel or impede other policies that have
a positive or negative impact on health, as well as which municipal policies favour or impede
the adoption of traffic-calming policies.
f) Citizen participation
In most cases, citizen participation constitutes an important component of the process of
planning traffic-calming interventions (Ewing, 1999; TAC & CITE, 1998; Ewing & Brown,
2009). These processes are often initiated at the request of citizens or in response to their
concerns about traffic volumes and speeds. Citizens are also frequently invited to play an
active role in the planning of interventions by public authorities, who thus ensure the social
acceptability of the calming measures selected. Such involvement by citizens can potentially
make them feel more in control of their environment. Moreover, greater feelings of control
have been associated with a decrease in the incidence and in the prevalence of numerous
chronic diseases and mental illnesses (Paquet, Dubé, Gauvin, Kestens, & Daniel, 2010). It
would be interesting to evaluate the positive, or even the negative, effects (in the case of
significant community conflict, for example) on health of participative planning processes for
traffic-calming interventions.

g) Social network
In the same vein, the protective effects of social networks on mental and physical health are
relatively well documented (Holt-Lunstad, Smith, & Layton, 2010). Thus, just as citizen
involvement in the process of planning interventions leads to social interaction and
networking, so quieter, friendlier neighbourhoods that encourage walking and cycling can
potentially increase the number of opportunities for social interaction (Leyden, 2003). It would
therefore be relevant for studies to evaluate the potential of calming strategies to encourage
the creation and strengthening of social relationships.
h) Heat islands
The literature review mentions several times the fact that the installation of calming
measures, by recuperating urban space previously devoted to cars, offers opportunities for
greening, which can have various desirable effects on noise, air quality and active
transportation. Since plants can be used to combat the heat island effect (Shashua-Bar &
Hoffman, 2000), it would be appropriate to evaluate the effect that the potential for greening
created by traffic calming (when combined with other interventions) can have on urban heat
islands.

BIBLIOGRAPHY
Abbott, P., Tyler, J., & Layfield, R. (1995). Traffic calming: vehicle noise emissions alongside
speed control cushions and road humps (Report No. TRL 180). Crowthorne, Berkshire:
Transport Research Laboratory.
Ahn, K. & Rakha, H. (2009). A field evaluation case study of the environmental and energy
impacts of traffic calming. Transportation Research Part D: Transport and Environment,
14, 411-424. doi: 10.1016/j.trd.2009.01.007.
Bahar, G., Smahel, T., & Smiley, A. (2009). Study of the environmental, economic, safety &
social benefits of roundabouts for Transport Canada. Human Factors North Inc. and
Navigats Inc.
Bassett, D. R., Pucher, J., Buehler, R., Thompson, D. L., & Crouter, S. E. (2008). Walking,
Cycling, and Obesity Rates in Europe, North America, and Australia. Journal of Physical
Activity and Health, 5(6), 795-814.
Beckett, K. P., Freer-Smith, P. H., & Taylor, G. (1999). Urban woodlands: their role in
reducing the effects of particulate pollution. Environmental Pollution, 99(3), 347-360.
doi: 10.1016/S0269-7491(98)00016-5.
Besser, L. M. & Dannenberg, A. L. (2005). Walking to Public Transit: Steps to Help Meet
Physical Activity Recommendations. American Journal of Preventive Medicine, 29(4),
273-280. doi: 10.1016/j.amepre.2005.06.010.
Boulter, P. G., Hickman, A. J., Latham, S., Layfield, R., Davidson, P., & Whiteman, P. (2001).
The impacts of traffic calming measures on vehicle exhaust emissions (Report No. TRL
482). Crowthorne, Berkshire: Transport Research Laboratory.
Boulter, P. G. & Webster, D. C. (1997). Traffic calming and vehicle emissions: A literature
review (Report No. TRL 307). Crowthorne, Berkshire: Transport Research Laboratory.
Bunn, F., Collier, T., Frost, C., Ker, K., Roberts, I., & Wentz, R. (2003). Traffic calming for the
prevention of road traffic injuries: systematic review and meta-analysis. Injury
Prevention, 9, 200-204.
Bunn, F., Collier, T., Frost, C., Ker, K., Steinbach, R., Roberts, I. & Wentz R. (2009). Area-
wide traffic calming for preventing traffic related injuries. Cochrane Database of
Systematic Reviews 2003, 1, 1-36. doi: 10.1002/14651858.CD003110.
Bureau de prévention des accidents. (2008a). Route et lois physiques. Mieux comprendre
pour mieux conduire (avec formules). Retrieved from: http://www.bfu.ch/PDFLib/1051_4
2.pdf.
Bureau de prévention des accidents. (2008b). Zones 30. Brochure technique. Retrieved
from: http://www.bfu.ch/PDFLib/1069_42.pdf.
Campolieti, D. & Bertoni, D. (2009). The action plan for noise reduction in Modena: methods,
effects and perspectives. Radiation Protection Dosimetry, 137(3-4), 252-255.
doi: 10.1093/rpd/ncp209.

Canadian Institute for Health Information. (2010). National Trauma Registry 2009 Report:
Major Injury in Canada (Includes 2007-2008 Data) Ottawa, Ontario: CIHI. Retrieved
from: http://secure.cihi.ca/cihiweb/products/ntr_major_injury_in_canada_2009_en.pdf.
Carver, A., Timperio, A. F., & Crawford, D. A. (2008). Neighborhood Road Environments and
Physical Activity Among Youth: The CLAN Study. Journal of Urban Health, 85(4),
532-544. doi: 10.1007/s11524-008-9284-9.
Centers for Disease Control and Prevention. (2010). CDC Recommendations for Improving
Health through Transportation Policy. Retrieved from: http://www.cdc.gov/transportation/
docs/FINAL%20CDC%20Transportation%20Recommendations-4-28-2010.pdf.
Cloke, J., Webster, D., Boulter, P., Harris, G., Stait, R., Abbott, P. & Chinn, L. (1999). Traffic
Calming: Environmental assessment of the Leigh Park Area Safety Scheme in Havant
(Report No. TRL 397). Crowthorne, Berkshire: Transport Research Laboratory.
Cubbin, C. & Smith, G. S. (2002). Socioeconomic Inequalities in Injury: Critical Issues in

Design and Analysis. Annual Review of Public Health, 23, 375. doi: 10.1146/annurev.pu
blhealth.23.100901.140548.
Daham, B., Andrews, G. E., Li, H., Partridge, M., Bell, M. C., & Tate, J. (2005). Quantifying
the Effects of Traffic Calming on Emissions Using On-road Measurements (Report
No. 2005-01-1620). Warrendale, U.S.: SAE International. Retrieved from: http://eprints.w
hiterose.ac.uk/2050/1/2005-01-1620SOrion_peedbump.pdf.
Department for Transport. (1999). Leigh Park Area Safety Scheme, Havant, Hampshire
(Report No. Traffic Advisory Leaflet 2/99). London: TSO. Retrieved
from: http://www.ukroads.org/webfiles/TAL%202-
99%20Leigh%20Park%20Area%20Safety%
20Scheme%20-%20Havant%20-%20Hampshire.pdf.
Department for Transport. (2007). Traffic Calming (Report. No. Local Transport Note 1/07).
London: TSO. Retrieved from: http://www2.dft.gov.uk/pgr/roads/tpm/ltnotes/pdfltn0107tra
fficcalm.pdf.
Desjardins, N., D'amours, G., Poissant, J., & Manseau, S. (2008). Avis scientifique sur les
interventions efficaces en promotion de la santé mentale et en prévention des troubles
mentaux. Québec: Institut national de santé publique du Québec. Retrieved
from: http://www.inspq.qc.ca/pdf/publications/789_Avis_sante_mentale.pdf.
Direction de la santé publique. (2006). Le transport urbain, une question de santé. Rapport
annuel 2006 sur la santé de la population montréalaise. Montréal: Direction de la
santé publique. Retrieved from: http://publications.santemontreal.qc.ca/uploads/tx_asss
mpublications/2-89494-491-8.pdf.
Dumbaugh, E. & Rae, R. (2009). Safe Urban Form: Revisiting the Relationship Between
Community Design and Traffic Safety. Journal of the American Planning Association,
75(3), 309-329. doi: 10.1080/01944360902950349.

Edmonton, the City of. (2008). Motor Vehicle Collisions 2007. Edmonton. Retrieved
from: http://www.edmonton.ca/transportation/RoadsTraffic/2007_Annual_Collision_Repo
rt_FINAL.pdf.
Elvik, R. (1997). Evaluations of road accident blackspot treatment: A case of the iron law of
evaluation studies? Accident Analysis & Prevention, 29(2), 191-199.
doi: 10.1016/S0001-4575(96)00070-X.
Elvik, R. (2001). Area-wide urban traffic calming schemes: a meta-analysis of safety effects.
Accident Analysis & Prevention, 33, 327-336.
Environmental Protection Agency & U.S. Department of Energy. (2010). Fuel Economy
Guide. U.S. Environmental Protection Agency and U.S. Department of Energy. Retrieved
from: http://www.fueleconomy.gov/feg/pdfs/guides/FEG2010.pdf.
Ewing, R. (1999). Traffic Calming: State of the Practice. Washington, DC, USA: Institute of
Transportation Engineers/Federal Highway Administration. Retrieved
from: http://www.ite.org/traffic/tcstate.asp.
Ewing, R. (2000). Impacts of Traffic Calming. Conference proceedings. “1st Urban Street
Symposium”. Dallas, TX, 28-30 June 1999. Retrieved from: http://www.urbanstreet.info/1
st_symp_proceedings/Ec019_i1.pdf.
Ewing, R. & Brown, S. J. (2009). U.S. Traffic Calming Manual. Chicago, Il.; Reston, VA:
American Planning Association; ASCE Press.
Ewing, R. & Edwards, P. (2009). The Built Environment and Traffic Safety: A Review of
Empirical Evidence. Journal of Planning Literature, 23(4), 347-367. doi: 10.1177/088541
2209335553.
Fang, C. F. & Ling, D. L. (2003). Investigation of the noise reduction provided by tree belts.
Landscape and Urban Planning, 63(4), 187-195. doi: 10.1016/S0169-2046(02)00190-1.
Forsyth, A., Hearst, M., Oakes, J. M., & Schmitz, K. H. (2008). Design and Destinations:
Factors Influencing Walking and Total Physical Activity. Urban Studies, 45, 1973-1996.
doi: 10.1177/0042098008093386.
Gibbard, A., Reid, S., Mitchell, J., Lawton, B., Brown, E., & Harper, H. (2004). The effect of
road narrowings on cyclists (Report No. TRL 621). Crowthorne, Berkshire: Transport
Research Laboratory. Retrieved from: http://www.transport-research.info/Upload/Docum
ents/200607/20060728_163846_65628_UG171_Final_Report.pdf.
Grand Lyon, Communauté urbaine de. (2007). Zone 30. Grand Lyon. Retrieved
from: http://www.grandlyon.com/fileadmin/user_upload/Pdf/activites/deplacements/2007
0308_gl_perimetre_zone30_lyoncentre.pdf.
Groupe de recherche Ville et mobilité. (2008). Le transport actif et le système scolaire à

Montréal et à Trois-Rivières. Analyse du système d'acteurs concernés par le transport
actif des élèves des écoles primaires au Québec. Montréal. Retrieved
from: http://mapageweb.umontreal.ca/lewisp/GVM%20rapport%20final.pdf.

Grundy, C., Steinbach, R., Edwards, P., Green, J., Armstrong, B., & Wilkinson, P. (2009).
Effect of 20 mph traffic speed zones on road injuries in London, 1986-2006: controlled
interrupted time series analysis. BMJ, 339, b4469. doi: 10.1136/bmj.b4469.
Grundy, C., Steinbach, R., Edwards, P., Wilkinson, P., & Green, J. (2008a). The Effect of
20 mph zones on Inequalities in Road Casualties in London: A report to the London
Road Safety Unit London: London School of Hygiene and Tropical Medicine. Retrieved
from: http://www.tfl.gov.uk/assets/downloads/the-effect-of-20-mph_-zones-on-inequali
ties_in-road-casualties-in-london.pdf.
Grundy, C., Steinbach, R., Edwards, P., Wilkinson, P., & Green, J. (2008b). 20 mph Zones
and Road Safety in London: A report to the London Road Safety Unit. London: London
School of Hygiene and Tropical Medicine. Retrieved from: http://www.20splenty
forus.org.uk/UsefulReports/20-mph-zones-and-road-safety-in-london.pdf.
Haddon, W. (1980). Advances in the Epidemiology of Injuries as a Basis for Public Policy.
Landmarks in American Epidemiology, 95(5), 411-421.
Hemsing, S. & Forbes, G. (2000). Ottawa-Carleton Traffic Calming Evaluation Study (Report
No. 99041). Synectics.
Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social Relationships and Mortality Risk:
A Meta-analytic Review. PLoS Medicine, 7(7), e1000316. doi: 10.1371/journal.pmed.100
0316.
Houwing, S. (2003). Traffic calming: engineering measures. In Ingrid van Schagen (ed.),
Traffic calming schemes: Opportunities and implementation strategies (pp. 27-34). The
Netherlands: Swov Institute for Road Safety Research. Retrieved
from: http://www.swov.nl/rapport/R-2003-22.pdf.
Human Resources and Skills Development Canada. (2011a). Indicators of well-being in

Canada: Health-Physical Activity. Consulted on June 7, 2011: http://www4.hrsdc.gc.ca/.3
nd.3c.1t.4r@-eng.jsp?iid=8.
Human Resources and Skills Development Canada. (2011b). Indicators of well-being in

Canada: Health - Physical Obesity. Consulted on June 7, 2011: http://www4.hrsdc.gc.ca/
.3nd.3c.1t.4r@-eng.jsp?iid=6.
Hummel, T. (2001). Land use planning in Safer Transportation Network Planning. Safety
principles, planning framework, and library information. The Netherlands: Swov Institute
for Road Safety Research. Retrieved from: http://www.swov.nl/rapport/D-2001-12.pdf.
Hyden, C. & Várhelyi, A. (2000). The effects on safety, time consumption and environment of
large scale use of roundabouts in an urban area: a case study. Accident Analysis &
Prevention, 32, 11-23.
Jacobsen, P. L. (2003). Safety in numbers: more walkers and bicyclists, safer walking and
bicycling. Injury Prevention, 9, 205-209. doi: 10.1136/ip.9.3.205.

Jacobsen, P. L., Racioppi, F., & Rutter, H. (2009). Who owns the roads? How motorised
traffic discourages walking and bicycling. Injury Prevention, 15(6), 369-373.
doi: 10.1136/ip.2009.022566.
Joly, M.-F., Bourbeau, R., Bergeron, J., & Messier, S. (1992). Analytical approach to the
identification of hazardous road locations: A review of the literature. Montréal : Centre de
recherche sur les transports.
Jones, S. J., Lyons, R. A., John, A., & Palmer, S. R. (2005). Traffic calming policy can reduce
inequalities in child pedestrian injuries: database study. Injury Prevention, 11, 152-156.
doi: 10.1136/ip.2004.007252.
Kamphuis, C. B. M., Giskes, K., Kavanagh, A. M., Thornton, L. E., Thomas, L. R., van
Lenthe, F. J., Turrel, G. (2008). Area variation in recreational cycling in Melbourne: a
compositional or contextual effect? Journal of Epidemiology and Community Health, 62,
890-898. doi: 10.1136/jech.2007.067116.
Katzmarzyk, P. T., Gledhill, N., & Shephard, R. J. (2000). The economic burden of physical
inactivity in Canada. Canadian Medical Association Journal, 163(11), 1435-1440.
Laflamme, L., Hasselberg, M., & Burrows, S. (2010). 20 Years of Research on

Socioeconomic Inequality and Children's Unintentional Injuries–Understanding the
Cause-Specific Evidence at Hand. International Journal of Pediatrics, 1-23. doi:
10.1155/2010/819687.
Lee, C. & Moudon, A. V. (2008). Neighbourhood design and physical activity. Building
Research and Information, 36(5), 395-411. doi: 10.1080/09613210802045547.
Leyden, K. M. (2003). Social Capital and the Built Environment: The Importance of Walkable
Neighborhoods. American Journal of Public Health, 93(9), 1546-1551.
Lindenmann, H. P. (2005). The Effects on Road Safety of 30 Kilometer-Per-Hour Zone

Signposting in Residential Districts. ITE Journal, 75(6), 50-54.
Litman, T. (1999). Traffic Calming: Benefits, Costs and Equity Impacts. Victoria Transport
Policy Institute. Retrieved from: http://www.vtpi.org/calming.pdf.
Lovegrove, G. R. & Sayed, T. (2006). Macro-level collision prediction models for evaluating
neighbourhood traffic safety. Canadian Journal of Civil Engineering, 33(5), 609-621.
doi: 10.1139/l06-013.
Macbeth, A. (1998). Calming Arterials in Toronto. Conference proceedings. “68th Annual

Meeting of the Institute of Transportation Engineers”. Toronto, Ontario, August 9-12,
1998. Retrieved from: http://www.ite.org/traffic/documents/AHA98C19.pdf.
Michaud, D. S., Keith, S. E., & McMurchy, D. (2005). Noise Annoyance in Canada. Noise &
Health, 7(27), 39-47.
Ministère des Transports du Québec. (2007). Roundabouts. Consulted on September 14,

2011: http://www.mtq.gouv.qc.ca/portal/page/portal/entreprises_en/camionnage/reseau_
routier/carrefours_giratoires.

Morency, C., Demers, M., & Lapierre, L. (2007). How Many Steps Do You Have in Reserve?
Thoughts and Measures About a Healthier Way to Travel. Transportation Research
Record: Journal of the Transportation Research Board, (2002), 1-6. doi: 10.3141/2002-
01.
Morency, C., Trépanier, M., & Demers, M. (2011). Walking to transit: An unexpected source
of physical activity. Transport Policy, 18(6), 800-806. doi: 10.1016/j.tranpol.2011.03.010.
Morency, P. & Cloutier, M.-S. (2005). Distribution géographique des blessés de la route sur
l'île de Montréal (1999-2003). Cartographie pour les 27 arrondissements. Direction de
santé publique. Agence de développement de réseaux locaux de services de santé et
de services sociaux de Montréal. Retrieved from: http://publications.santemontreal.qc.ca/
uploads/tx_asssmpublications/2-89494-460-8.pdf.
Morency, P. & Cloutier, M.-S. (2006). From targeted "black spots" to area-wide pedestrian
safety. Injury Prevention, 12(6), 360-364. doi: 10.1136/ip.2006.013326.
Morrison, D. S., Thomson, H., & Petticrew, M. (2004). Evaluation of the health effects of a
neighbourhood traffic calming scheme. Journal of Epidemiology and Community Health,
58, 837-840. doi: 10.1136/jech.2003.017509.
Mountain, L. J. & Fawaz, B. (1992). The Effects of Engineering Measures on Safety at

Adjacent Sites. Traffic Engineering & Control, 33(1), 15-22.
Mountain, L. J., Hirst, W. M., & Maher, M. J. (2005). Are speed enforcement cameras more
effective than other speed management measures? An evaluation of the relationship
between speed and accident reductions. Accident Analysis & Prevention, 37, 731-741.
doi: 10.1016/j.aap.2005.03.017.
Nowak, D. J. & Crane, D. E. (2002). Carbon storage and sequestration by urban trees in the
USA. Environmental Pollution, 116(3), 381-389.
Ntziachristos, L. & Samaras, Z. (2000). Speed-dependent representative emission factors for

catalyst passenger cars and influencing parameters. Atmospheric Environment, 34(27),
4611-4619. doi: 10.1016/S1352-2310(00)00180-1.
Owen, B. (2005). Air quality impacts of speed-restriction zones for road traffic. Science of
The Total Environment, 340, 13-22. doi: 10.1016/j.scitotenv.2004.08.011.
Paquet, C., Dubé, L., Gauvin, L., Kestens, Y., & Daniel, M. (2010). Sense of Mastery and
Metabolic Risk: Moderating Role of the Local Fast-Food Environment. Psychosomatic
Medicine, 72, 324-331. doi: 10.1097/PSY.0b013e3181cdf439.
Persaud, B. & Lyon, C. (2007). Empirical Bayes before-after safety studies: Lessons learned
from two decades of experience and future directions. Accident Analysis & Prevention,
39(3), 546-555. doi: 10.1016/j.aap.2006.09.009.
Pikora, T., Giles-Corti, B., Bull, F., Jamrozik, K., & Donovan, R. (2003). Developing a
framework for assessment of the environmental determinants of walking and cycling.
Social Science & Medicine, 56(8), 1693-1703. doi: 10.1016/S0277-9536(02)00163-6.

Pucher, J. & Buehler, R. (2008). Making Cycling Irresistible: Lessons from The Netherlands,
Denmark and Germany. Transport Reviews, 28(4), 495-528. doi: 10.1080/01441640701
806612.
Pucher, J. & Dijkstra, L. (2003). Promoting Safe Walking and Cycling to Improve Public
Health: Lessons From The Netherlands and Germany. American Journal of Public
Health, 93(9), 1509-1516.
Pucher, J., Dill, J., & Handy, S. (2010). Infrastructure, programs, and policies to increase
bicycling: An international review. Preventive Medicine, 50(Supplement 1), S106-S125.
doi: 10.1016/j.ypmed.2009.07.028.
Pucher, J., Garrard, J., & Greaves, S. (2011). Cycling Down Under: A Comparative Analysis
of Bicycling Trends and Policies in Sydney and Melbourne. Journal of Transport
Geography, 19(2), 332-345. doi: 10.1016/j.jtrangeo.2010.02.007.
Pucher, J. & Renne, J. L. (2003). Socioeconomics of Urban Travel: Evidence from the 2001
NHTS. Transportation Quarterly, 57(3), 49-77.
Retting, R. A., Bhagwant, P. N., Garder, P. E., & Lord, D. (2001). Crash and Injury Reduction
Following Installation of Roundabouts in the United States. American Journal of Public
Health, 91(4), 628-631.
Reynolds, C., Harris, M., Teschke, K., Cripton, P., & Winters, M. (2009). The impact of
transportation infrastructure on bicycling injuries and crashes: a review of the literature.
Environmental Health, 8(1), 47. doi: 10.1186/1476-069X-8-47.
Rietveld, P. & Daniel, V. (2004). Determinants of bicycle use: do municipal policies matter?
Transportation Research Part A: Policy and Practice, 38(7), 531-550.
doi: 10.1016/j.tra.2004.05.003.
Rosales, J. A. (2007). Road Diet Handbook - Overview. Conference proceedings. District 6

Annual Meeting “Prepare. Evaluate. Innovate. Sustain.” Portland, Oregon, July 15-18,
2007. Retrieved from: http://www.oregonite.org/2007D6/paper_review/D4_201_Rosales
_paper.pdf.
Sergerie, D., King, N., Drouin, L., Fortier, I., Smargiassi, A., & Maurice, P. (2005). Road
Speed: Health Impact and Counteractive Measures. Scientific Review. Institut
national de santé publique du Québec. Retrieved from: http://www.inspq.qc.ca/pdf/public
ations/847_road_speed.pdf.
Shashua-Bar, L. & Hoffman, M. E. (2000). Vegetation as a climatic component in the design

of an urban street: An empirical model for predicting the cooling effect of urban green
areas with trees. Energy and Buildings, 31(3), 221-235. doi: 10.1016/S0378-
7788(99)00018-3.
SkyscraperPage Forum. (2011). The Great Canadian Cycling Thread. Consulted on

August 25, 2011: http://forum.skyscraperpage.com/showthread.php?t=191039.
Smargiassi, A., Berrada, K., Fortier, I., & Kosatsky, T. (2006). Traffic intensity, dwelling value,
and hospital admissions for respiratory disease among the elderly in Montreal (Canada):
a case-control analysis. Journal of Epidemiology and Community Health, 60(6), 507-512.
doi: 10.1136/jech.2005.037044.

Société de l’assurance automobile du Québec. (2011). Speeding Reduces Your Field of

Vision. Consulted on August 2, 2011: http://www.saaq.gouv.qc.ca/en/accident_preventio
n/speed/slowing_down/vision.php.
Stein, H., Kittelson, W., Newton, J., & Hottmann, K. (1992). Portland's Successful Experience
with Traffic Circles. Conference proceedings. “Institute of Transportation Engineers
Annual Meeting”. Washington, D.C., August 9-12, 1992. ITE Compendium of
Technical Papers, 39-44. Retrieved from: http://www.ite.org/traffic/documents/AHA92A3
9.pdf.
Stewart, A. E. & Lord, J. H. (2002). Motor vehicle crash versus accident: a change in
terminology is necessary. Journal of Traumatic Stress, 15(4), 333-335.
doi: 10.1023/A:1016260130224.
Stout, T. B., Pawlovich, M., Souleyrette, R. R., & Carriquiry, A. (2006). Safety impacts of
"road diets" in Iowa. Institute of Transportation Engineers. ITE Journal, 76, 24-27.
Taylor, M. C., Lynam, D. A., & Baruya, A. (2000). The effects of drivers' speed on the
frequency of road accidents (Report No. TRL 421). Crowthorne, Berkshire: Transport
Research Laboratory. Retrieved from: http://www.20splentyforus.org.uk/UsefulReports/T
RLREports/trl421SpeedAccidents.pdf.
Tester, J. M., Rutherford, G. W., Wald, Z., & Rutherford, M. W. (2004). A Matched Case-
Control Study Evaluating the Effectiveness of Speed Humps in Reducing Child
Pedestrian Injuries. American Journal of Public Health, 94(4), 646-650.
Toronto Public Health. (2007). Air Pollution Burden of Illness from Traffic in Toronto:
Problems and solution. Toronto, Canada. Retrieved from: http://www.toronto.ca/health/h
phe/pdf/air_pollution_burden.pdf.
Transport Canada. (2005). Traffic Calming In Canadian Urban Areas (Report No. TP 14420
E). Retrieved from: http://www.tc.gc.ca/media/documents/programs/cs35e_trafficcalming
.pdf.
Transport, Health and Environment Pan-European Programme. (2004). Transport-related

Health Effects with a Particular Focus on Children - Towards an Integrated Assessment
of their Costs and Benefits. State of the Art Knowledge, Methodological Aspects and
Policy Directions. Topic Report: Physical Activity.
Transportation Association of Canada & Canadian Institute of Transportation Engineers.

(1998). Canadian Guide to Neighbourhood Traffic Calming. Ottawa, Canada:
Transportation Association of Canada.
Transportation Demand Management Encyclopedia. (2010a). Traffic Calming: Roadway

Design to Reduce Traffic Speeds and Volumes. Victoria Transport Policy Institute.
Consulted on July 8, 2010: http://www.vtpi.org/tdm/tdm4.htm.
Transportation Demand Management Encyclopedia. (2010b). Health and Fitness: Strategies

That Improve Public Health Through Physical Activity. Victoria Transport Policy Institute.
Consulted on July 8, 2010: http://www.vtpi.org/tdm/tdm102.htm.

Transportation Research Board. (1995). Expanding Metropolitan Highways: Implications for

Air Quality and Energy Use (Special report No. 245). Washington, D.C.: National
Academy Press. Retrieved from: http://www.nap.edu/openbook.php?record_id=9676.
Várhelyi, A. (2002). The effects of small roundabouts on emissions and fuel consumption: a
case study. Transportation Research Part D, 7, 65-71.
Watkins, K. F. (2000). Cambridge's Traffic Calming Program: Pedestrians are the Focus.
Conference proceedings. “ITE 2000 Annual Meeting and Exhibit” Nashville,
Tennessee August 6-8, 2000. Retrieved from: http://www.ite.org/traffic/documents/AB00
H3702.pdf.
World Health Organization. (1999). Guidelines for Community Noise. Geneva. Retrieved
from: http://www.who.int/docstore/peh/noise/guidelines2.html.
World Health Organization. (2004). World report on road traffic injury prevention. Geneva.
Retrieved from: http://whqlibdoc.who.int/publications/2004/9241562609.pdf.
World Health Organization. (2009). Global Status Report on Road Safety: Time for Action.
Geneva. Retrieved from: http://whqlibdoc.who.int/publications/2009/9789241563840_en
g.pdf.
World Health Organization Regional Office for Europe. (2000). Transport, environment and
health. WHO Regional Publications, European Series, No. 89. Copenhagen.
Retrieved from: http://www.euro.who.int/__data/assets/pdf_file/0003/87573/E72015.pdf.
World Health Organization Regional Office for Europe. (2002). A Physically Active Life
Through Everyday Transport with a Special Focus on Children and Older People and
Examples and Approaches from Europe. Copenhagen. Retrieved from:
http://www.euro.who.int/__data/assets/pdf_file/0011/87572/E75662.pdf.
World Health Organization Regional Office for Europe. (2005). Health effects of transport-
related air pollution. Copenhagen. Retrieved from: http://www.euro.who.int/__data/assets
/pdf_file/0006/74715/E86650.pdf.
World Health Organization Regional Office for Europe. (2009). Night Noise Guidelines for
Europe. Copenhagen. Retrieved from: http://www.euro.who.int/__data/assets/pdf_file/00
17/43316/E92845.pdf.
World Health Organization Regional Office for Europe. (2011). Transport and health–
Physical activity. Copenhagen. Consulted on June 22, 2011: http://www.euro.who.int/en/
what-we-do/health-topics/environment-and-health/Transport-and-health/facts-and-
fiigures/physical-activity2.
Yang, J., McBride, J., Zhou, J., & Sun, Z. (2005). The urban forest in Beijing and its role in air
pollution reduction. Urban Forestry & Urban Greening, 3(2), 65-78.
doi: 10.1016/j.ufug.2004.09.001.
Zein, S. R., Geddes, E., Hemsing, S., & Johnson, M. (1997). Safety Benefits of Traffic
Calming. Transportation Research Record, 1578(1), 3-10. doi: 10.3141/1578-01.

APPENDIX 1
INTERVENTION LOGIC OF
THE TWO APPROACHES TO TRAFFIC CALMING
INTERVENTION LOGIC OF THE TWO APPROACHES TO TRAFFIC CALMING
Approaches Mechanisms of action a Effects on health determinants Effects on health
a
In this box, texts encircled by a solid line refer to the principal mechanisms of action of traffic-calming measures and strategies. Texts encircled by a broken line (dashes) indicate
intermediate mechanisms acting between the principal mechanisms and the effects on the determinants of health.

APPENDIX 2
GLOSSARY OF TRAFFIC-CALMING MEASURES

GLOSSARY OF TRAFFIC-CALMING MEASURES
Bike box/
Sas vélo, Sas cyclable
A bike box is a facility that allows cyclists
to position themselves in front of vehicles
stopped at an intersection with traffic lights.
This painted space on the pavement
makes cyclists more visible and ensures
them start-up priority when the light turns
green.
Source: www.flickr.com.
Photographer: Richard Drdul.
Bike lane, Cycle lane/

Bande cyclable
A bike lane is a portion of the road
reserved for the exclusive or preferential
use of cyclists. Unlike a cycle track, which
is physically separated from motor vehicles
using the road, a bike lane is delimited by
road markings. The space needed for the
bike lane is generally obtained by
eliminating one traffic lane, by narrowing
one or several lanes, or by eliminating
Source: www.pedbikeimages.org.
parking spaces for cars. Photographer: Steven Faust.
Chicane, Serpentine, Reversing curve,

Twist/
Chicane
A chicane is a series of horizontal
deflections (usually three in a row) installed
on an otherwise straight road to create an
“S” shaped traffic lane.
Photographer: Dan Burden.

Choker, Mid-block narrowing, Pinch

point, Mid-block yield point,
Constriction/
Goulot d’étranglement
A choker is an isolated narrowing of one or
several traffic lanes created by the
installation of horizontal deflections in the
centre or on the sides of the road. The
term "choker," like its equivalents, is
usually reserved for narrowings located
other than at intersections.
Source: www.cyclestreets.net.
Photographer: unknown.
Crosswalk, Zebra crosswalk, Zebra

crossing/
Passage piéton, Traverse piétonne,
Traversée piétonne
A crosswalk is a facility designed to make
crossing the road easier for pedestrians by
delimiting a space with road markings to
indicate that it is meant to be shared with
pedestrians.
Curb extension, Bulb-out, Bulbout/

Saillie de trottoir, Avancée de trottoir
A curb extension is a continuation of the
sidewalk at an intersection intended to
make pedestrians more visible and
decrease their exposure to collisions by
reducing crossing distances. A curb
extension can also be used to reduce the
width or the number of traffic lanes.

Cycle track/
Piste cyclable
A cycle track is a portion of the road
reserved for the exclusive use of cyclists.
Unlike a bike lane, which is delimited by
road markings, a cycle track is physically
separated from motorized traffic by
bollards, medians, parking spaces, etc.
The space needed for the cycle track is
generally obtained by eliminating a traffic
lane, by narrowing one or several lanes, or
by eliminating parking spaces for cars.
Photographer: Eric Gilliland.
Diagonal diverters, Full diverters,

Diagonal road closures/
Terre-plein diagonal, Îlot diagonal
A diagonal diverter is a raised island
placed diagonally at an intersection so as
to allow only right turns. Diagonal diverters
can be designed to allow pedestrians and
cyclists to continue on their way
unobstructed. Source: www.flickr.com.
Photographer: UrbanGrammar.
Forced-turn island, Right-turn island,

Forced turn lane, Deflector island,
Forced turn channelization/
Îlot de canalisation, Îlot tourne-à-droite,
Îlot tourne-à-gauche
A forced-turn island is a median positioned
at the approach to an intersection that
orients vehicles in the desired direction or
directions.
Source: Ewing, 1999, p. 29.

Full closure, Full street closure, Cul-de-

sac, Dead-end/
Fermeture de rue, Impasse, Cul-de-sac
Full closures often take the form of barriers
that prevent motor vehicles from continuing
along the road, but allow pedestrians and
cyclists to pass.
Gateway/
Portail d’entrée, Porte d’entrée
Gateways are facilities designed to
indicate entrance to a calmed area.
Mini-roundabout, Mini-traffic circles,

Intersection islands/
Minigiratoire, Îlot circulaire
A mini-roundabout is an intersection with a
central island that is usually raised and
circular. Vehicles entering the circle must
yield passage to those already inside and
must travel around in a counterclockwise
direction.

One-way street/
Rue à sens unique
A one-way street is a street on which
vehicles are authorized to travel in only
one direction. One-way streets can be
used, with minimal cost, to prevent through
traffic from using local residential streets
instead of roads designed to handle larger
traffic volumes (collector roads and
arteries) to cross an area. For example,
the installation of two facing one-way
streets going in opposite directions can
force drivers to turn onto an intersecting Photographer: Guillaume Goyette.
artery, and prevent vehicles from
continuing in a straight line along local
streets.
Pedestrian refuge, Median refuge/

Refuge piéton
A pedestrian refuge is a median typically
located in the middle of the road to allow
pedestrians to cross in two stages.
Source: NCCHPP.
Photographer: Olivier Bellefleur.

Raised crosswalk, Raised zebra

crossing, Raised crossing, Hump
pelican/
Passage piéton surélevé, Traverse
piétonne surélevée, Traversée
surélevée
A raised crosswalk is a facility designed to
make crossing the road easier for
pedestrians and which typically raises the
pavement to the level of the sidewalks.
Raised crosswalks are often made of a
textured and coloured material to indicate
clearly that the space is meant to be Photographer: Dan Burden.
shared with pedestrians.
Raised intersection, Raised junction,

Intersection hump, Table, Plateau/
Intersection surélevée
A raised intersection is an intersection
where the pavement has been raised
relative to the level of the roads leading to
it. The platform created by the vertical
deflection is often made of a textured
material and is raised to the level of the
sidewalks to indicate clearly that the space
is meant to be shared with pedestrians.
Source: www.cyclestreets.net.
Photographer: unknown.
Raised median, Centre island

narrowing, Traffic island/
Terre-plein central, Îlot central
A raised median is a raised island usually
built down the central axis of two-way
roads to separate traffic going in opposite
directions and reduce lane widths.

Road diet, Lane reduction/

Régime routier
A road diet usually refers to the conversion
of a four-lane road into a three-lane road,
with one lane for traffic going in each
direction and a central lane reserved for
left turns from either direction. The space
recuperated can be used to add bike
lanes, sidewalks, or vegetation.
Roundabout, Modern roundabout/

Carrefour giratoire, Giratoire
A roundabout is an intersection at which
vehicles entering must yield right of way to
vehicles already circulating around a
central circular or oval-shaped island. To
slow down traffic and induce drivers to
yield right of way, there are horizontal
deflections at the entrances which position
vehicles to rotate in the correct direction.
Roundabouts generally replace Source: www.flickr.com.
Photographer: WSDOT.
intersections with traffic signals on roads
designed for quite high traffic volumes
(collectors, arteries).
Speed bump, Bump/

Dos d’âne
Speed bumps, not to be confused with
speed humps, are narrow vertical
deflections that generally extend less than
30 centimetres across. In cars, it is easy to
travel over them at very low speeds
(5-10 km/h) or very high speeds, in which
case the suspension system can absorb
the deflection. Thus, their use is generally
restricted to areas where high speeds are
impractical, such as parking lots or Source: www.flickr.com.
alleyways. Photographer: Bridget Ames.

Speed Camera/
Radar photo
Speed cameras are devices that allow
vehicles exceeding the speed limit to be
identified automatically.
Photographer: B.T. Indrelunas.
Speed cushions, Speed lumps/

Coussins berlinois
Speed cushions are vertical deflections
designed to act on cars in the same way
as speed humps, while having a minimal
effect on heavy vehicles, such as
emergency vehicles (fire truck, ambulance,
etc.) and buses.
Speed hump, Road hump, Hump/

Dos d’âne allongé
Speed humps, not to be confused with
speed bumps, are wide vertical deflections
that typically extend three to four metres
along the road. They can only be travelled
over comfortably at low speeds (15-
30 km/h). Thus, their use is widespread on
local streets in residential neighbourhoods,
in school zones, around parks, etc.

Speed limit painted on the asphalt/

Marquage au sol indiquant la limite de
vitesse
Road markings indicating the speed limit
are often used in conjunction with other
calming measures, such as vertical or
horizontal deflections.
Photographer: Ian Britton.
Speed table, Trapezoidal hump, Speed

platform/
Plateau ralentisseur
A speed table is a vertical deflection
spanning the pavement, whose top is
usually flat and extends far enough along
the road for a car or even a heavy vehicle
to rest on it. The vertical contour of speed
tables allows them to be easily crossed at
faster speeds than speed humps allow,
Source: Boulter et al., 2001, p.11.
which is why they are generally used on
collector roads and arteries.
Speed-activated sign/
Signal lumineux activé par la vitesse
A speed-activated sign is a device that
usually indicates the speed of vehicles and
whether they are travelling under or over
the speed limit.
Photographer: Eric Allix Rogers.

Stop sign/
Panneau d’arrêt
A stop sign is a traffic sign indicating that
drivers must stop their vehicle and wait
until the lane is free before continuing on
their way. Its purpose is usually to manage
right of way for users of an intersection.
However, it is also sometimes used as a
traffic-calming measure. For example, stop
signs have been used in the past to slow
down traffic in certain areas, and thus
make them less attractive to through traffic. Photographer: Bridget Ames.
Textured crosswalk, Textured crossing/

Passage piéton texturé, Traverse
piétonne texturée, Traversée texturée
A textured crosswalk is a facility designed
to make crossing the road easier for
pedestrians and which is made from a
textured, and often coloured, material to
indicate clearly that the space is meant to
be shared with pedestrians.

APPENDIX 3
SUMMARY TABLES OF EVALUATIVE STUDIES

SUMMARY TABLES OF EVALUATIVE STUDIES
The tables below constitute a synthesis of the evaluations of traffic-calming interventions

included in this literature review, including our comments. Readers will find here a synthesis
of each study (research questions, methodology, results), along with a column containing
remarks about the conceptual validity, the internal validity and/or the reliability of each one.
Although they are presented individually, the studies are grouped into three broad
intervention categories: those evaluating individual traffic-calming measures (when these are
not explicitly identified as part of an area-wide intervention); those evaluating a series of
measures installed on a single road; and those evaluating a series of measures in a
geographic area including more than one road (whether they were planned to function in a
systemic manner on a road network or were installed without an explicitly identified
intervention logic 24).
24
It is arguable whether studies evaluating a series of measures contained within a given geographic area should
be treated in the same manner as studies evaluating area-wide interventions when the intervention logic of the
former has not been specified. To allow readers to interpret the information differently, an explanatory note has
been included in the “Remarks” column to clearly identify such studies.

1 Evaluations of individual traffic-calming measures
b c
AUTHOR RESEARCH METHODOLOGY RESULTS REMARKS
a
(YEAR) QUESTION
(Abbott et How do Design: experimental before-after Maximum vehicle noise level: At a constant Experimental fleet of heavy vehicles
al., 1995) maximum noise study. speed of 25 km/h, the implementation of noisier than the average fleet in the
levels emitted by traffic-calming measures on a level country (overestimation of noise).
different types of Road types simulated in the London roadway leads to fewer increases in
vehicles on Transport Research Laboratory (U.K.) maximum noise levels (in dB LAmax) than Noise measurements based on
different types of with narrow and wide speed cushions, road deterioration, for all types of vehicles. typical constant speeds for each type
roads speed humps, poorly-maintained and At typical constant speeds before and after of traffic-calming measure instead of
maintained in level roads. Vehicles tested: light the installation of traffic-calming measures, taking into account speed variations
different (cars and milk trucks), buses and the introduction of vertical deflections has caused by traffic-calming measures
conditions vary? heavy vehicles (from 16 to 38 t). the following effect: reduces the average (risk of overestimating noise from
maximum noise level for light vehicles suspension systems and rattling
What effects do Maximum noise level measured at Ŧ metal while underestimating
7.5 m from road centre and at 1.2 m (-10.2 to -6.6 dB LAmax); slightly increases
traffic-calming Ŧ
(+1.0 dB LAmax for narrow speed cushions) acceleration and deceleration noise).
measures that from ground (at a constant speed of Ŧ
25 km/h and at constant speeds or lowers (-3.6 to -0.5 dB LAmax for others) No statistical significance test
vertically deflect that of buses; increases that of heavy
vehicles have on based on typical average speeds Ŧ
mentioned.
before and after installation of the vehicles (+2.0 to +7.9 dB LAmax), except in
traffic noise the presence of round-top humps (-2.1 dB
Ŧ
levels? traffic-calming measures tested) and
modelling of sound levels (based on LAmax).
seven traffic composition scenarios). Noise levels: Depending on the
composition of traffic, the introduction of
vertical deflections has the following effect
on ambient noise conditions: improves it
when there are only light vehicles
Ŧ
(-7.1 to -4.8 dB LAeq ); improves it, has no
Ŧ
effect, or worsens it (-1.3 to +7 dB LAeq )
when there are 10% heavy vehicles and
1% buses; worsens it (+0.1 to
a
The question or questions listed here are not necessarily isomorphic with those of the evaluations. When a study examined a broader subject than that of traffic calming, for
example, the relationship between the built environment and physical activity, we formulated the questions such that they pertain to traffic calming.
b
In general, when statistical significance tests are mentioned in the studies, only the significant results are reported here. Systematically, the term “significant,” as well as the “*”
symbol are used to identify significant results of at least p <0.05. When non-significant results are reported, they are identified by the term “non-significant,” as well as by the
absence of the “*” symbol. Results for which studies mention no statistical significance test are identified by the “Ŧ” symbol, and a comment is included in the “Remarks” column.
c
The remarks included in this column concern the internal validity of the study and its reliability. Considerations related to external validity generally concern all the studies and are
therefore addressed in the conclusion of the literature review.

1 Evaluations of individual traffic-calming measures (cont.)

(YEAR) QUESTION
Ŧ
(Abbott et +8.7 dB LAeq ) when there are 1% buses
al., 1995) and 25% heavy vehicles. Narrow speed
(cont.) cushions and round-top humps would be
preferable where heavy vehicles are
present, because they would allow for
decreased noise levels even in the
presence of 94% cars, 1% buses and 5%
heavy vehicles for narrow speed cushions,
and of 79% cars, 20% heavy vehicles and
1% buses for round-top humps.
(Boulter & What are the Design: traditional literature review. Effects on gas-powered cars of an The methodology used to locate the
Webster, effects of intervention at a single point on a road case studies is not described.
1997) various traffic- Data from 12 case studies (seven network:
calming different countries) are reported and No statistical significance test
measures and the interventions are briefly described. • Speed hump, car with catalytic converter mentioned.
Ŧ
strategies on the One case study concerns an (WCC): +20% Carbon monoxides (CO),
Ŧ
emission of air intervention at a single point on a +4% Carbon dioxide (CO2)/gas,
Ŧ
pollutants by road network, six case studies +18% Nitrogen oxides (NOx);
motor vehicles? examine the effects of introducing a
series of measures on a road and five • Speed hump, car with no catalytic
Ŧ Ŧ
case studies concern area-wide converter (NCC): +11% CO, +5%
Ŧ
interventions. CO2/gas, +22% NOx.
Ŧ Ŧ
(Campolieti What are the Design: before-after. Reduction of 1 dB LAday and of 2.5 LAnight. Controlled for volume and
& Bertoni, effects of composition of motor traffic.
2009) replacing traffic Noise levels recorded by
lights with a microphones located at three sites Tables and calculations are
roundabout on (two at the roundabout and one a few somewhat obscure: data that would
ambient noise dozen metres away on a street be needed to reproduce calculations
levels in leading to it). on which conclusions are based are
Modena (Italy)? missing.
No statistical significance test
mentioned.


(YEAR) QUESTION
(Gibbard et What are the Design: (1) Internet survey, (1) 78.4% of respondents indicated that the (1) Experienced cyclists; mainly male
al., 2004) effects on space (2) before-after and after-only. traffic-calming measures were problematic (15.5% women, 84.5% men).
sharing by motor (because, according to them, they
vehicles, cyclists (1) Survey by questionnaire between encourage motor vehicles to travel closer (2) Two of the sites were part of
and pedestrians July 10, 2002 and August 13, 2002. to cyclists); 17% said they were unaffected; area-wide schemes, although they
of traffic-calming Administered through the websites of 3.8% said the measures helped them. were evaluated in isolation; the
measures that cyclists’ associations, so respondents From among a list of phenomena resulting number of cyclists and conflicts
narrow roads at were relatively experienced cyclists in road narrowings (various traffic-calming observed was too low to produce
specific points (n=393). measures, parked vehicles, bus stops, statistically significant results.
and of measures (2) Analysis of conflicts between etc.), respondents selected pedestrian
for assisting motor vehicles and cyclists at five refuge islands or raised medians (38.9%),
cyclists at these sites where centre islands (raised as well as chicanes (8.9%) as problematic.
points? medians) had been installed to Over 40% of respondents said they
provide a safe crossing point for stopped at a road narrowing, 46.8% said
pedestrians. A before-after study for they sometimes move onto the sidewalk at
two sites and after-only for the other narrowings, and 48.6% indicated they
three. Data on: number of cyclists; sometimes choose another route to avoid
number of conflicts; whether or not narrowings. A few cyclists (5.3%) indicated
motor vehicles slowed down; where feeling stressed or intimidated when
vehicles overtook cyclists in relation negotiating this type of traffic-calming
to the road narrowing; distance measure in the presence of motorcycles,
between vehicles and cyclists; but this percentage climbs to 33.6% when
distance of cyclists from the curb. cars are present, 46.6% for buses, 59.8%
Sites with varying characteristics (for for light vans, and 61.6% for medium or
example, presence or absence of bike heavy trucks. Women indicated feeling
lane; regular bike lane vs. lane more stressed or intimidated than men.
painted green). Daytime video survey (2) Cyclists appear to have a tendency to
(12 hours) for two days after (in all move closer to the curb when they are
cases) and two days before (in two overtaken near a pedestrian island. The
cases). Control for speed and volume presence of a mandatory bike lane, without
of traffic. facilities allowing cyclists to bypass the
narrowing, seems to encourage motor
vehicle drivers to travel closer to cyclists.
The presence of warning signs indicating
an approaching island seems to prompt


(YEAR) QUESTION
(Gibbard et drivers to overtake cyclists before the

al., 2004) island and to leave less space between
(cont.) themselves and cyclists when overtaking–
except when the bike lane is painted green,
which seems to prompt drivers to overtake
cyclists after the island and to leave more
room between their vehicles and cyclists.
In areas where traffic was heavier and/or
vehicles were sometimes parked along the
edge of the road, cyclists tend to travel
further from the curb as they pass through
the road narrowing and drivers are less
likely to overtake cyclists before or while
passing through the narrowing.
(Mountain What is the Design: before-after with control sites Vertical deflections are more effective The grouping together of horizontal
et al., 2005) average effect of (empirical Bayes method: change in (-44%*, confidence interval [CI] 95%: -54 to deflections and road narrowings with
various speed safety = expected number of -34) than horizontal deflections (-29%*, speed-activated signs (n=4) and
management collisions vs. number recorded). CI 95%: -48 to -8) and speed cameras markings indicating the speed limit
measures 79 sites with speed cameras; 31 sites (-22%*, CI 95%: -30 to -13) for reducing (n=1) is unusual. The authors specify
(speed cameras; using one or several horizontal collisions. These reductions appear to that they were grouped together
vertical deflections, road narrowings, speed correspond to reductions in average speed because their results were similar.
deflections; activated signs (n=4) or speed limit (vertical: -13.5 km/h*, CI 95%: -16.6 The inclusion of sites using vertical
horizontal painted on asphalt (n=1); 39 sites to -10.5; horizontal: -5.3 km/h*, and horizontal deflections with sites
deflections and using vertical deflections (or a CI 95%: -7.1 to -3.7; speed using vertical deflections may
road narrowings combination of vertical and horizontal camera: -6.6 km/h*, CI 95%: -7.6 to -5.5). obscure the specific effects of the
and speed- deflections). For KSI collisions, only vertical deflections combination.
activated signs) Data from police and other local produced significant reductions (-35%*, The analyses include control for
on collision authorities on collisions, personal CI 95%: -54 to -18), whereas horizontal regression to the mean only for total
frequency and injury collisions (PICs) and collisions deflections (-14%, CI 95%: -44 to +32) and collisions.
vehicle speeds? with persons killed or seriously injured speed cameras (-11%, CI 95%: -26 to +6)
produced non-significant reductions. For The article does not describe the
(KSI). Data for 3 years before and, on implementation contexts of the
average, 2.5 years after. Speed speed cameras and vertical deflections,
respectively, -6%* (CI 95%: -9 to -3) traffic-calming measures evaluated.
cameras: collisions recorded within a It is therefore impossible to
and -6%* (CI 95%: -10 to -3) collisions are
due to reduced traffic volume.


(YEAR) QUESTION
(Mountain distance of 1 km upstream and 1 km Children, as cyclists and as pedestrians, as determine whether some of them
et al., 2005) downstream from the cameras. well as pedestrians, appear to benefit most were planned and implemented so
(cont.) from vertical and horizontal deflections. as to function in a systemic manner
(The reductions concerning them are within a road network.
significant, but the results presented by
type of road user do not take into account
either regression to the mean or the
underlying trend.)
As regards the number of PICs avoided
and not the percentage of reduction, speed
cameras installed on high-volume roads
registering a higher number of PICs allow
for the avoidance of as many injuries
(-1.00*/km/year, CI 95%: -1.4 to -0.6) as
vertical deflections (-1.03*/km/year,
CI 95%: -1.4 to -0.8) usually installed on
local roads with lower traffic volume.
(Retting et What are the Design: before-after with control sites Reduction of 38%* in the total number of Explanatory mechanism: reduced
al., 2001) nature and the (empirical Bayes method: change in collisions (TCs), of 76%* in personal injury speeds and decrease in conflict
extent of safety = expected number of collisions (PICs) and of 89%* in fatal and points.
reductions in collisions vs. number recorded). incapacitating injury collisions (FIICs).
collisions and Analysis of data from police reports Single-lane roundabout, in an urban
their severity or, when these were unavailable, from environment, replacing stop signs: -61%*
following the report summaries concerning (TCs) and -77%* (PICs). Similar results in
installation of 24 modern roundabouts. rural environments: -58%* (TCs)
modern and -82%* (PICs). Multi-lane roundabouts,
roundabouts in in an urban environment, replacing stop
the United signs: -5% (TCs). Single-lane roundabout,
States? in an urban environment, replacing traffic
lights: -35%* (TCs) and -74%* (PICs).


(YEAR) QUESTION
Ŧ
(Stout et al., What are the Design: before-after with control sites (1) Reduction of 25.2% in the number of The expression “possible injury
2006) effects on road and empirical Bayes method (change collisions per kilometre (collision density) collisions” refers to collisions
Ŧ
safety of road in safety = expected number of and of 18.8% in the collision rate following which victims complained
diets (reduction collisions vs. number recorded). (controlled for volume) at the experimental of pain, but where no other
from four to sites, as compared to the control sites. symptoms were visible.
three lanes on Two methods using the same
Ŧ
bidirectional 15 experimental sites and 15 control (2) Reduction of 28% in the number of No statistical significance test
roads)? sites: collisions per kilometre (collision density), mentioned.
Ŧ
of 21% in the collision rate (controlled for
Do the two (1) Empirical Bayes Method. Analysis Ŧ
volume), of 34% for injury collisions, of
analysis of monthly data from the Iowa Ŧ
11% for fatal and serious collisions, of
methods used Department of Transportation, for Ŧ
30% for minor injury collisions and of
produce similar 23 years. Ŧ
31% for possible injury collisions.
results? (2) Before (5 years)-after (5 years)
study with control sites.
(Tester et How does living Design: observational case-control Among control children, 23% (n=46) lived The protective effect of speed humps
al., 2004) on a street block study. near a speed hump, whereas among case within a radius of 0.4 km may
with a speed children, the rate was 14% (n=14). The depend on the proximity of other
hump installed Use of admission data for a hospital matched-pairs analysis indicated that the traffic-calming measures installed in
affect the emergency department in the city of probability that a child living near a speed the sector. However, the article does
probability that a Oakland (U.S.A.) over a period of five hump will be injured by a car near home not indicate the implementation
child will be years (1995-2000). Identification of (radius of 0.4 km) and admitted to hospital contexts of the speed humps
injured or killed persons aged under 15 years (n=100) d
is significantly lower (odds ratio [OR] : studied. It is therefore impossible to
by a car when living on local streets who had been 0.50*, CI 95%: 0.27 to 0.89) than that of determine whether some of them
walking close to admitted after being struck by a car children not living near this kind of traffic- were planned and implemented so
home? while walking near their home (radius calming measure. When the analysis as to function in a systemic manner
of 0.4 km). Excluded were injuries considers only those children injured on the within a road network.
caused by cars backing out of a street block where they live (n=49), the
driveway or a parking space. Each
d
OR or odds ratio: An odds ratio is a statistic produced through logistic regression that expresses the relationship between the probability that an event will occur and the probability
that it will not occur. An odds ratio can be used to determine if, all other things being equal, persons living in an environment defined by certain characteristics are more or less likely
to move around by bicycle, for example. An odds ratio of 1 would indicate that the characteristic had no effect, whereas a ratio of less than 1 would indicate that the characteristic in
question lowers the likelihood that such persons would move around by bicycle and a ratio of more than 1 would indicate an increase in the likelihood. The further the odds ratio is
from 1, the greater the effect on the probability of the event occurring.


(YEAR) QUESTION
(Tester et case patient was matched with two protective effect of speed humps is even
al., 2004) control children (admitted on the greater (OR: 0.38*, CI 95%: 0.15 to 0.90).
(cont.) same day for other medical reasons,
matched in terms of gender and age Univariate analyses indicated significant
and also living on a local street). Case associations for the variables of ethnicity
children and control children were and race.
classified into two categories: those The multivariate analysis (adjusted for
living on a street block with at least ethnicity and race) indicated that the
one speed hump installed and those protection provided by proximity to a speed
living on a street block without any hump remains statistically significant and
speed humps. similar for children injured within a radius of
Matched-pairs analysis to determine 0.4 km (OR: 0.47*, CI 95%: 0.24 to 0.95),
whether proximity of speed humps is however, although the point estimate still
a significant risk reduction factor. indicates a protective effect, the confidence
Univariate analyses to determine intervals indicate that the relationship is no
whether other variables (race, longer statistically significant for children
ethnicity, household income and injured on the street block where they live
medical insurance) are also (OR: 0.40, CI 95%: 0.15 to 1.06).
significantly associated. Multivariate On the basis of these results, the article
analysis to isolate the effect of speed concludes that speed humps make the
humps. environment safer for children walking.

2 Evaluations of a series of traffic-calming measures implemented on one road
b c
a
(YEAR) QUESTION
(Ahn & What are the energy Design: before-after and cross- (1) Intersections with traffic-calming Data were collected on weekends to
Rakha, and environmental sectional. measures are associated with minimize interaction with other
2009) impacts of certain significantly greater gas consumption vehicles.
traffic-calming Study of behaviour on road network and higher emission levels than
measures? (and not on an experimental track) uncontrolled intersections and with (1) In comparisons of the traffic-
comparing: significantly less consumption and calming measures at intersections,
lower emission levels than intersections variations in emission levels on cross
(1) at the intersections of a road:
with stop signs. streets were not taken into account.
different traffic-calming measures
(mini-roundabouts, speed humps, (2) Traffic-calming measures
speed cushions), uncontrolled significantly raise gas consumption and
intersections and an intersection with emission levels because they lead to
stop signs; variations in speed.
(2) on two other road sections: before
and after the installation of five speed
cushions on one road and two speed
bumps on another.
Combination of second-by-second
GPS data (to capture variations in
real speed) and microscopic energy
and emission models.
Different types of vehicles.
a
b
c

2 Evaluations of a series of traffic-calming measures implemented on one road (cont.)

(YEAR) QUESTION
(Boulter & What are the effects Design: traditional literature review. Effects on gas-powered cars of a series The methodology used to locate the
Webster, of various traffic- of traffic-calming measures on a road: case studies is not described.
1997) calming measures Data from 12 case studies (seven
and strategies on different countries) are reported and • Speed humps (SHs) with and No statistical significance test
the emission of air the interventions are briefly described. without catalytic converter (CC): +70 mentioned.
Ŧ
pollutants by motor One case study concerns an to +80% Carbon monoxides (CO),
Ŧ
vehicles? intervention at a single point on a +70 to +100% Hydrocarbons (HC),
Ŧ
road network, six case studies +50 to +60% Carbon dioxide
Ŧ
examine the effects of introducing a (CO2)/gas, -20 to 0% Nitrogen
series of measures on a road and five oxides (NOx);
case studies concern area-wide Ŧ
• Ten SHs with CC: +300% CO,
interventions. Ŧ Ŧ
+37% CO2/gas, +300% NOx;
Ŧ
• Ten SHs without CC: +200% CO,
Ŧ Ŧ
+51% CO2/gas, +300% NOx;
• Various traffic-calming measures
and 40-km/h speed limit:
Ŧ
-9% CO2/gas;
Ŧ
• Six SHs with CC: +300% CO,
Ŧ Ŧ
+25% CO2/gas, +1000% NOx;
Ŧ
• Five SHs: +36 to +73% CO2/gas;
Ŧ
• Two roundabouts: +33% CO2/gas.
(Boulter et What are the effects Design: before-after and cross- Emissions: Average variations for all The method used to model the typical
al., 2001) of various traffic- sectional. the traffic-calming measures taken speed profiles is not always made
calming measures together: Gas-powered cars NCC: clear. For purposes of readability,
(75 mm flat-top Emissions: Evaluation based on the +34%* CO, +50%* HC, +1% NOx, steps that do not contribute to the final
speed hump; modelling of typical driving cycles +20%* CO2; Gas-powered cars WCC: results because they were discarded
80 mm round-top before and after installation of traffic- +59%* CO, +54%* HC, +8% NOx, and replaced by other procedures
speed hump; calming measures on existing roads. +26%* CO2; Diesel-powered car: during the course of the experiment
100 mm raised Exceptions: for chicane, curb +39%* CO, +48%* HC, +28%* NOx, are not mentioned in the present
intersection; 1.9 m extension, mini-roundabout and +26%* CO2, +30%* Particulate matter document.
and 1.7 m wide speed cushions the data for modelling (PM).
driving cycles before the interventions


(YEAR) QUESTION
Ŧ
(Boulter et speed cushions; was taken from other roads with Air quality: 1998: +0.9 to +8.0% CO, The emission measurements based on
Ŧ
al., 2001) chicane; curb similar characteristics. Combinations +1.4 to +11.9% benzene, +2.9 to the before and after speed profiles
Ŧ
(cont.) extension; of radar-type instruments to record +11.2% 1,3-butadiene, -1.2 to were not performed on the same cars.
Ŧ Ŧ
combination of drivers’ speeds without their +5.1% NO2; 2000: +0.4 to +6.7% CO, Seven of the twelve cars were
Ŧ
bollards for knowledge, counters to record traffic +2.7 to +10.1% benzene, +2.7 to replaced by other cars in the same
Ŧ
narrowing a road volume and video cameras to +9.5% 1,3-butadiene, -0.4 to category for the after tests.
Ŧ Ŧ
and a speed determine the composition of motor +4.3% NO2; 2005: -0.5 to +5.1% CO, The results concerning air quality must
Ŧ
cushion; mini- traffic on the selected road sections. +1.2 to +7.7% benzene, +1.2 to be interpreted with caution because:
Ŧ
roundabout) on the Composition of sample speed profiles +7.2% 1,3-butadiene, +0.5 to (1) the report does not indicate when
Ŧ
emission of air on the road sections under study and +3.3% NO2. the before-after counts were taken.
pollutants by motor modelling of typical driving cycles. The authors imply that seasonal
vehicles? Measurement on a chassis differences characterizing the before
What are the dynamometer of emissions for three and after results or a general trend
expected effects of categories of cars as they drove toward an increase in motorized travel
these emissions on through each of the typical driving can explain the increased traffic
ambient air quality? cycles (12 gas-powered cars used, volume at four sites following the
including 2 small cars with catalytic interventions.
converters [WCC], 2 small cars with
no catalytic converters [NCC], (2) No statistical significance test
2 medium cars WCC, 2 medium cars mentioned.
NCC, 2 large cars WCC, 2 large cars
NCC and 3 medium diesel-powered
cars).
Air quality: Compiled emission data
were analyzed using a dispersion
model to determine effects at 10 m
from road centres. The results take
into consideration decreased traffic
volume at one site and increases at
four sites after installation (but the
report does not specify when). The
results also take into account the
traffic volume of heavy vehicles, the
emission data for which are taken
from a previous study focused on the


(YEAR) QUESTION
(Boulter et relationship between the speed of

al., 2001) these vehicles and their emissions.
(cont.) The effect is estimated for three years
(1998, 2000 and 2005), based on the
estimated proportion of vehicles with
and without catalytic converters in
England, but not based on the
evolution of traffic volumes.
(Daham et What effect do Design: experimental, aimed at Substantial increase in emission of all It would not be cautious to use these
Ŧ
al., 2005) speed humps have simulating before-after pollutants measured: +90% CO2; results to characterize present or
Ŧ Ŧ Ŧ
on the emission of measurements. Measurements made +117% CO; +195% NOx; +148% HC. future emissions of air pollutants by
air pollutants? using a device installed on the vehicles travelling over speed humps
automobile. or speed cushions because:
On-road measurements of emissions (1) use of an exceptional driving style
were made on an existing road with (the most aggressive) and presence of
an average of seven speed aggravating conditions (e.g., vehicles
cushions/km. Control runs were made were very heavy due to the
on the calmed road at a constant measurement equipment; vehicles
speed of 50 km/h. Experimental runs were slowed down more than is
were made to simulate the presence necessary to simulate comfortably
of 80 mm speed humps (the highest); travelling over speed humps);
specifically, the vehicle was slowed (2) control runs on road with speed
down to 16 km/h and accelerated cushions instead of on road without
back up to 32-50 km/h in second gear calming measures and test runs over
at each speed cushion. Drive cycles speed cushions instead of over speed
tested: 1.1 km, roundtrip = 2.2 km, humps;
14 speed cushions. (3) no statistical significance test
mentioned.


(YEAR) QUESTION
(Morrison What are the effects Design: before (6 months)–after Significant increase in pedestrian Weak response rate (39% for the first
et al., on physical and (6 months). activity. questionnaire, once undelivered
2004) mental health of the envelopes were taken into account,
installation of Cohort of 750 participants recruited Perception: Significant reduction in (244/624) and 32% for the second
several traffic- by random sampling (244 responded traffic-related problems (e.g., parents (185/576).
calming measures before and 185 after) to fill in postal allowing children to walk, cycle or play
(five pairs of speed questionnaires about their transport alone outside more, and less bothered The authors did not ask respondents if
cushions, two zebra habits, their state of physical and by the speed of vehicles, fumes, noise, they had been unaffected or if they
crossings, and mental health (SF-36 version 2), their vibrations, etc. (refer to full table on were walking or cycling less.
parking bays) on the perception of the neighbourhood, and p. 838 of the article).
of traffic problems and road safety. The results of the pedestrian counts
main road running Significant improvement in self-reported are congruent with the responses to
through a Two full-day pedestrian counts at state of physical health (note: no the questionnaires with regard to
disadvantaged three locations, with a one-year significant difference between those physical activity, but these results are
residential area in interval. who reported walking more and the based solely on two one-day counts.
Glasgow, UK? others).
The inclement weather for walking
No significant effect on mental health. during the second count may have led
to an underestimation of the increase
in pedestrian activity.
(Watkins, What are the effects Design: before-after and survey sent Drivers yielding to pedestrians: Raised No statistical significance test
Ŧ
2000) of an intervention by post after. crossing: 13% before, 53% after . mentioned.
comprising several Raised intersection: 18% before, 54%
measures (curb Measurement of percentage of drivers Ŧ
after .
extensions, raised yielding to pedestrians before-after at
pedestrian crossing, a raised crossing and raised Perception of road safety: 69% better,
raised intersection intersection. 15% worse, 4% no change, 8% don’t
and chicane) on a know. For pedestrians: 57% better and
Survey by postal questionnaire 13% worse. For cyclists: 33% better
street in Cambridge, between six and nine months after
Massachusetts and 8% worse. For motorists: 46%
completion of construction work better and 10% worse.
(USA)? (perception of safety for various users
How do residents of the street).
perceive the impact
on safety of an


(YEAR) QUESTION
(Watkins, intervention
2000) comprising several
(cont.) measures (curb
extensions, raised
pedestrian crossing,
raised intersection
and improved
crosswalk markings
for pedestrians)
replacing a traffic
light.

3 Evaluations of a series of traffic-calming measures implemented within a geographic area comprising more than one
road
b c
a
(YEAR) QUESTION
(Boulter & What are the Design: traditional literature review. Effects of area-wide interventions on gas- The methodology used to locate the
Webster, effects of various powered cars: case studies is not described.
1997) traffic-calming Data from 12 case studies (seven
measures and different countries) are reported and • Area with extensive traffic calming, No statistical significance test
strategies on the the interventions are briefly without catalytic converter (CC): +7 to mentioned.
Ŧ
emission of air described. One case study concerns +71% Carbon monoxides (CO), -25
Ŧ
pollutants by an intervention at a single point on a to -10% Hydrocarbons (HC), +7 to
Ŧ
motor vehicles. road network, six case studies +19% Carbon dioxide (CO2)/gas, -60
Ŧ
examine the effects of introducing a to -38% Nitrogen oxides (NOx);
series of measures on one road and • 30-km/h zone, without CC: -20 to
five case studies concern area-wide Ŧ Ŧ
+28% CO, -23 to +2% HC, -6 to
interventions. Ŧ Ŧ
+14% CO2/gas, -31 to -5% NOx;
• Speed humps scheme: Slight
Ŧ Ŧ
increase CO, no change HC,
Ŧ
decrease NOx;
• Area with extensive traffic calming:
Ŧ Ŧ
Increase CO, increase HC,
Ŧ
decrease NOx;
• 21 mini-roundabouts and 30-km/h
d Ŧ Ŧ
limit : +2% CO, +<1% CO2/gas,
Ŧ
+1% NOx.
a
b
c
d
Boulter and Webster (1997) classify this study among those evaluating cases of a series of calming measures installed on one road. However, the reference is to an evaluation of
an area-wide scheme involving the installation of 21 mini-roundabouts in Växjö, Sweden, also evaluated by Hyden and Várhelyi (2000) and Várhelyi (2002). For this reason the
study is included in this section.

road (cont.)

(YEAR) QUESTION
(Bunn et al., Can area-wide Design: systematic review with Fatal collisions: Rate ratio (before/after/ The rates indicate a trend toward
2003) interventions meta-analysis. control zones): 0.63 (confidence interval improvement, but the confidence
reduce road [CI] 95%: 0.14 to 2.59) or -37% intervals (CI) indicate that they are not
collision-related Inclusion of randomized trials and (CI 95%: -86 to +159). statistically significant.
deaths and controlled before-after studies.
injuries? Results pertain to 12 distinct Personal injury collisions (fatal and non- Groups together evaluations of highly
controlled before-after studies fatal): 0.89 (CI 95%: 0.80 to 1.00) or -11% diverse area-wide interventions, which
evaluating a total of 16 area-wide (CI 95%: -20 to 0). are only very briefly described (only
strategies. the types of measures). The tests of
Collisions: 0.95 (CI 95%: 0.81 to 1.11) heterogeneity between the results of
Data collected from police or from or -5% (CI 95%: -19 to +11). different studies frequently proved
“local authorities” on the number of statistically significant. A random
collisions, fatal collisions and Collisions involving a pedestrian: 1.00
(CI 95%: 0.84 to 1.18) or identical to the effects model that pooled results was
personal injury collisions. used to account for this heterogeneity.
situation prior to the intervention
For each intervention evaluated, rate (CI 95%: -16 to +18%). All the studies included focus on
ratios were calculated for interventions carried out in the 1970s
before/after/control zones. and 1980s.
Having not controlled for exposure to
risk, the authors suggest that there
may possibly be more pedestrians in
the calmed zones and that this would
partially explain the absence of a
reduction in collisions involving a
pedestrian.
(Bunn et al., Can area-wide Design: systematic review with Fatal collisions: Rate ratio Same remarks as for 2003, except
2009) interventions meta-analysis. (before/after/control zones): 0.79 that the confidence intervals (CI)
reduce road (confidence interval [CI] 95%: 0.23 to indicate that the reduction in personal
collision-related Update of the 2003 meta-analysis 2.68) or -21% (CI 95%: -77 to +168). injury collisions is statistically
deaths and (Bunn et al., 2003): inclusion criteria significant.
injuries? modified to include “reduced speed Personal injury collisions (fatal and non-
limit zones.” Results pertain to fatal): 0.85* (CI 95%: 0.75 to 0.96) The dates of the added interventions
18 distinct controlled before-after or -15%* (CI 95%: -25 to -4). are not indicated.
studies evaluating a total of 22 area-

road (cont.)

(YEAR) QUESTION
(Bunn et al., wide strategies. Collisions: 0.89 (CI 95%: 0.76 to 1.05)
2009) or -11% (CI 95%: -24 to +5).
(cont.)
Collisions involving a pedestrian: 1.01
(CI 95%: 0.88 to 1.16) or +1%
(CI 95%: -12 to +16).
(Carver et What is the Design: cross-sectional study. Adolescent boys in neighbourhoods with No significant association between the
al., 2008) relationship an average number of speed humps (from road environment and the likelihood of
between traffic- Recruitment from 19 state primary 2 to 7) are significantly less likely than young children (aged 5-6 years)
calming measures schools in Melbourne (Australia) in those in neighbourhoods with a low making seven or more walking/cycling
and active areas of varying socioeconomic number of speed humps (from 0 to 1) to trips per week. No significant
transportation for status (SES). Transport habits of make seven or more walking/cycling trips association between the presence of
youth in young children (aged 5–6 years; per week (odds ratio [OR]: 0.38*, traffic-calming measures and physical
Melbourne? n=295) reported by parents and self- confidence interval [CI] 95%: 0.15 to activity for young children.
reported by adolescents (aged 10– 0.97). However, adolescent girls in
What is the 12 years; n=919). Moderate or neighbourhoods with the most speed The article does not take into
relationship vigorous physical activity outside of humps (from 8 to 99) are significantly consideration the proximity of other
between traffic- school hours recorded using features such as parks and
more likely to make seven or more
calming measures accelerometers. Features of the walking/cycling trips per week (OR: 2.95*, businesses.
and physical road environment within a radius of CI 95%: 1.34 to 6.51).
activity for youth 800 m around the home of each In seeking associations between the
in Melbourne? participant were identified using a The number of speed humps is road environment and seven or more
geographic information system significantly and positively associated with walking/cycling trips per week, the
(GIS). moderate to vigorous physical activity for authors may have missed other
adolescent boys during evenings factors associated with fewer trips.
Multiple regression analyses. (r=0.210*), but negatively with activity This article evaluates the effects of
before school hours for adolescent girls traffic-calming measures on an area
(r=-0.073*). The number of road including more than one road, but it
narrowings is significantly and negatively does not indicate whether or not the
associated with physical activity for measures were planned and
adolescent boys during weekends implemented so as to function in a
(r=-5.197*). systemic manner.

road (cont.)

(YEAR) QUESTION
(Cloke et What are the Design: before-after and before-after Collisions and injuries: The number of Air: No statistical test related to
al., 1999) effects of an area- with control site. personal injury collisions is stable 3 years emissions is mentioned. Wide margin
wide strategy on before and 20 months after. No observable of error for instrument used to
collisions and Evaluation of a traffic-calming change involving motorcyclists, cyclists, measure air quality. The
injuries, on air scheme comprising various and adult pedestrians, but a decrease of measurements from the control sites
quality and on measures (raised intersection, Ŧ
50% per year in personal injury collisions neutralize the results.
noise? speed cushions, pedestrian refuges, involving child pedestrians.
raised medians, curb extensions, Noise: The results are based on only
How are mini-roundabouts, raised pedestrian Air: Increase in emissions for most two days and two nights of recording.
residents’ crossings, gateways) aimed at vehicles, but decrease for the area (due No control for weather variation. It is
perceptions reducing the speed of vehicles and to the decrease in traffic volume). Non- noted that the second night was very
affected? at discouraging through traffic on significant improvement in air quality. windy, which could explain the
residential streets in the Leigh Park increase in background noise
What are the area of Havant (UK). Noise: In general, decrease in maximum recorded.
effects on noise level (LAmax) of cars and increase in
transport habits? Collisions: Data for two 3-year periods maximum noise level of heavy vehicles. Perceptions: Significant
before and for 20 months after. Overall decrease in noise level exceeded methodological imprecision (the report
10% of the time during the day (-4.7 does not specify how the interviews
Air: For emissions, experimental Ŧ
to -1.9 dB LA10, 18h) and at night (-3.6 were carried out or how participants
analysis of driving cycles and Ŧ
to -0.1 dB LA10, 6h), except for noise at were chosen). The results are based
modelling of HC, CO and NOx. For night at a pedestrian refuge (+2,6Ŧ dB on the responses of 151 residents
air quality, diffusion tubes at six sites LA10, 18h). Mixed results with regard to before and 150 after, of which only
(including two control sites) to background noise (LA90) during the day 113 did both before and after
measure NO2 and benzene. and significant increase at night (+7.8 to interviews. These results should be
Ŧ
Noise: Analysis of noise levels +14.1 dB LA90, 6h), but doubtful validity. interpreted with reserve.
(LAmax, LA10, 6h, LA10, 18h, LA90, 6h, Perceptions: Perceptions vary from zone
LA90, 18h) using the Statistical Pass- to zone. In general, significantly fewer
By method near three calming persons bothered by speeding vehicles,
measures and residences. Video to the amount of traffic, danger or difficulty in
determine vehicle types. crossing the road, or danger to children.
Perceptions and transport habits: Perceived improvement in road safety.
Interviews before and after of groups Little effect on perception of other
of residents selected among zones dimensions documented (noise, air
within the area to link results to the pollution, etc.).
various measures.

road (cont.)

(YEAR) QUESTION
(Elvik, 2001) What are the Design: systematic review with Travel habits: No notable effect. The results of the studies are
effects of area- meta-analysis. relatively consistent across decades
wide schemes on For the entire calmed area: -15%* and countries, which seems to
road safety? Review of 33 studies (1971–1994) (confidence interval [CI] 95%: -17 to -12) indicate that the reported
that provide information about personal injury collisions; -16%* (CI 95%: effectiveness is not the result of
numbers of collisions. Data on: -19 to -13) collisions without injury. confounding factors, as these would
design (all are before-after studies, have had to influence results in a
with or without control sites– For local roads in the calmed area: -24%*
(CI 95%: -29 to -18) personal injury consistent and uniform manner in
heterogeneity test of effects over varying implementation and
time and across location [several collisions; -29%* (CI 95%: -25 to -22)
collisions without injury. evaluation contexts.
countries] and of publication bias);
traffic volumes; types of roads; For main roads in the calmed area: -8%*
personal injury collisions; number of (CI 95%: -12 to -5) personal injury
collisions; property damage only collisions; -11%* (CI 95%: -16 to -6)
collisions. collisions without injury.
Retaining only the studies with the most
robust designs (before-after with control
sites) recording personal injury collisions,
only the reductions for the entire calmed
area (-12%*, CI 95%: -21 to -1) are
statistically significant.
(Forsyth et What is the Design: cross-sectional study. Significant positive association between The concept of traffic-calming was
al., 2008) relationship the percentage of calmed roads in an operationalized in a restrictive manner
between walking Multilevel regression analysis to area (as well as the presence of that includes certain measures that
and physical discern associations between over sidewalks, street lights and several can slow down motorized traffic.
activity and 200 environmental variables divided indicators of street connectivity) and both Excluded are measures acting on
certain into four categories (density; street total walking distances (r=0.3674*) and traffic volume and some that can
characteristics of pattern or connectivity; pedestrian transportation walking distances influence speed by narrowing roads,
the built infrastructure and amenities (r=0.3629*) covered by its residents, but such as bike lanes (Macbeth, 1998)
environment, [including calming measures]; and no significant association with physical (see online operationalization
including traffic- destinations) and walking and activity in general or leisure walking. protocol www.designforhealth.net/pdfs
calming physical activity. /GIS_Protocols/NEAT_GIS_V5_0_26
measures? Sample of 36 zones of Nov2010FIN.pdf).
805 m x 805 m in Minneapolis- Activity recorded from April to

road (cont.)

(YEAR) QUESTION
(Forsyth et St. Paul (USA) randomly selected November only.

al., 2008) among strata formed from pairings
(cont.) of median block size (small, This article evaluates the effects of
medium, large) and residential traffic-calming measures on an area
density (low, medium, high). including more than one road, but it
Random sampling of does not indicate whether or not the
715 participants, divided almost measures were planned and
equally among the zones. implemented so as to function in a
systemic manner.
Questionnaire focused on
socioeconomic characteristics.
Accelerometer and self-reported
travel/walking diary. Environmental
features measured through surveys
and variables based on Geographic
Information System (GIS) data from
existing databases, interpretation of
photos taken from satellites and a
field-based urban design inventory.
(Grundy et What are the Design: longitudinal and cross- London-wide: Annual reductions were Controlled for the underlying
al., 2008a) effects on sectional study. significantly greater in less-deprived SES downward trend in road injuries.
inequalities in quintiles for: personal injury Controlled for regression to the mean.
road traffic Inequalities measured: Five levels of collisions (PICs); child PICs; pedestrian Not controlled for underlying trend in
injuries of the socioeconomic status (SES) based PICs; cyclist PICs; powered two wheeler road injury inequalities among ethnic
399 20-mph on the Index of Multiple Deprivation PICs; and car occupant PICs. Annual groups; these results are therefore
(32-km/h) zones 2004 (data for 1987–2006) and on reductions were significantly greater less reliable.
in London (UK)? “ethnicity,” that is, white, black and among those identified as white than
Asian (data for 1996–2006). among Asian or black people for: PICs; The comparison among zones
Note: In 2008, the child PICs; collisions with persons killed installed in areas with different SES
percentage of Personal injury collisions: Geocoded was based on effectiveness
police data. or seriously injured (KSI); pedestrian
roads covered by PICs; child pedestrian PICs; cyclist PICs; expressed as a percentage, as is
the zones varied Time series for intervention zones powered two wheeler PICs; and car standard practice in the literature on
by quintile (from and adjacent areas. Calculation of occupant PICs. On the basis of a less this subject; it was not based on the
2.5% for Q1 [the the difference between the predicted number of injuries, of serious injuries,

road (cont.)

(YEAR) QUESTION
(Grundy et least deprived] to outcome without intervention and conservative risk reduction factor than and of deaths avoided. This practice
al., 2008a) 27.5% for Q5 [the the situation recorded in 2006. Thus, that used in 2009 (Grundy et al., 2009), is not without consequence for the
(cont.) most deprived]). a longer or shorter time after the the authors conclude that having targeted results, for there are often more
In 2000 coverage intervention, depending on the zone. deprived areas for installation of 20-mph collisions, injuries and deaths in
was less than 2% (32-km/h) zones can be viewed as having deprived areas than in the least
for all quintiles. In prevented 1193 persons from being deprived areas (Laflamme et al.,
terms of zones, injured each year, with almost half of 2010; Cubbin & Smith, 2002; Morency
6% of zones were these being in the most deprived quintile. & Cloutier, 2005). Thus, it is likely that
in Q1 and 35% in Regarding the number of injuries avoided, interventions whose effectiveness,
Q5 in 2008, the installation of these zones can be expressed as a percentage, is the
whereas in 1995, viewed as having reduced the widening same for areas with different SES
40% were located gap between PICs in the least deprived would prevent more collisions, injuries
in Q1 and 10% in and the most deprived quintiles by about and deaths if they were implemented
Ŧ
Q5. 15% . in areas with a lower SES.
Across 20-mph (32-km/h) zones: Annual Use of a less conservative hypothesis
reductions (expressed as percentages) in than in 2009 (Grundy et al., 2009) to
PICs and KSI collisions are similar for assess the number of injuries
zones and adjacent areas with varying avoided.
SES, with the exception of powered two
wheeler PICs and car occupant PICs, for
which reductions are significantly higher
in sectors adjacent to the most deprived
areas. Annual reductions in PICs and KSI
collisions in the zones and adjacent areas
are similar for different ethnic groups.
However, in calmed zones, significantly
less reduction in KSI collisions, in
pedestrian PICs and in child pedestrian
PICs is observed among those identified
in the study as black people than is seen
among those identified as Asian or white.

road (cont.)

(YEAR) QUESTION
(Grundy et What are the Design: longitudinal and cross- General trend in London: From over There appears to be no significant
al., 2008b) effects on the sectional study. 50,000 (1987) personal injury collisions to difference between the large calmed
number and Description of interventions a little over 30,000 (2006), with a plateau zones (more than 3.6 km of road) and
seriousness of assisted by use of a geographic from 1993 to 2001. the small calmed zones (3.6 km and
collisions of the information system (GIS). Calmed Pooled data from zones: Total less) as regards the effectiveness of
399 20-mph zones range from a road segment collisions -37.5%* (confidence interval [CI] reducing personal injury collisions and
(32-km/h) zones of 0.07 km to one of 37 km 95%: -43.4 to -31.6); personal injury collisions with persons killed or
that have been (median: 3.6 km). collisions (PICs); -41.9%* (CI 95%: -47.8 seriously injured.
installed in Quantification of the effects of each to -36.0%); PICs involving children -48.5%* Having not controlled for exposure to
London (UK) zone on collisions and risk of injury (CI 95%: -55.0 to -41.9); collisions with risk in their analyses, the authors
gradually since (data collected by police) within the persons killed or seriously injured suggest that there may possibly be
1990? zones, on adjacent roads, and (KSI); -46.3%* (CI 95%: -54.1 to -38.6); more pedestrians in the calmed zones
outside areas to verify whether KSI collisions involving children -50.2%* and that this would partially explain
collision “migration” occurred. (CI 95%: -63.2 to -37.2); pedestrian PICs the smaller reduction in collisions
Annual rates over 20-year period -32.4%* (CI 95%: -37.7 to -27.1); child involving pedestrians.
(1986–2006) used in main pedestrian PICs -46.2%* (CI 95%: -55.5 to No migration of collisions was
analyses. -36.8); pedestrian KSI -4.8%* observed, but the results do not take
Time series analysis used to check (CI 95%: -47.5 to -22.1); child pedestrian into account measures that may have
for regression to the mean. KSI -43.9%* (CI 95%: -61.3 to -26.6); been taken to calm traffic on adjacent
Individual analysis of zones and cyclist PICs -16.9%* (CI 95%: -29.0 roads. Thus, the authors may
pooled results. Calculation of the to -4.8); cyclist KSI -37.6%* overestimate the effectiveness of
difference between the predicted (CI 95%: -60.9 to -14.4); powered two 20-mph (32-km/h) zones in reducing
outcome without intervention and wheeler PICs -32.6%* collisions on these roads.
the situation recorded in 2006 (CI 95%: -43.4 to -21.7); powered two Data one year before and one year
(thus, a longer or shorter time after wheeler KSI -39.1%* (CI 95%: -59.1 after presented in the body of the text,
e
the intervention, depending on the to -19.0); car occupant PICs -52.5%* but up to five years before and five
zone). (IC 95%: -62.4 to -42.5); car occupant KSI years after included in appendices
-61.8%* (CI 95%: -71.7 to -52.0). (note: little difference between the
Adjacent areas: All collisions -7.4%* two).
(CI 95%: -11.0 to -3.8); PICs -8.0%*
(CI 95%: -11.5 to -4.4); KSI
collisions -7.9%* (CI 95%: -13.5 to -2.2);
car occupant PICs -11.5%*
(CI 95%: -16.5 to -6.4).
e
Occupants include drivers and passengers of automobiles.

road (cont.)

(YEAR) QUESTION
(Grundy et What are the Design: longitudinal and cross- Condensed form of the 2008 report; thus, Controlled for regression to the mean:
al., 2009) effects of sectional study. results are the same (Grundy et al., little effect.
399 20-mph Observational study of geocoded 2008b). Result added for personal injury Controlled for location of zones (city
(32-km/h) zones police data (1986-2006) covering collisions involving child cyclists centre vs. periphery): no effect.
on the number of 119,029 road segments with at least (0–15 years): -27%* (confidence interval
collisions causing one collision (out of a total of [CI] 95%: -49.1 to -6.3).
minor injury, 298,644 segments included in the On the basis of a more conservative risk
serious injury and database). Estimation of the effects reduction factor than that used in 2008
death in London in the area where the zones where (Grundy et al., 2008a), the authors
(UK)? introduced and on adjacent areas estimate that the 20-mph (32-km/h) zones
and adjustment for underlying in London prevent 203 persons, including
downward trend. 51 pedestrians from being injured each
year, and that, of these, 27 would have
been seriously injured or killed.
(Hemsing & What do residents Design: survey and interview Perception of air quality: The majority of Non-random survey. Thus, there is a
Forbes, and road users afterward. respondents observed no change or did risk that it may not be representative
2000) perceive to be the Survey by questionnaire. Comments not know if there had been one. For most of “the population” of Ottawa. On
effects of traffic- from public during public hearings. road categories, more respondents some roads, only a few surveys were
calming measures Interviews with individuals, perceived an improvement than filled out. No statistical significance
on walking and organizations and special interest deterioration. test mentioned. The extrapolation of
cycling safety, on groups. Perception of environmental noise levels: the results should therefore be
air quality, and on Results were pooled according to The majority of respondents observed no interpreted with caution.
noise levels in the categories of calmed roads change or did not know if there had been
f
Ottawa ? (local, collector, regional) and the one. For most road categories, more
type of traffic-calming measures respondents perceived an increase in
installed (vertical, horizontal, vertical noise levels than a decrease.
and horizontal).
f
The report covers numerous other dimensions that could very well have been the subject of our literature review, such as the perceived effects of traffic calming on social cohesion,
for example. However, this is the only study we found that examines these other dimensions. Given that the results concerning these dimensions are not in themselves conclusive,
we decided not to review them. As regards air quality, environmental noise levels, and transport habits, there exists a body of literature which, although not very extensive, is at
least sufficiently large to allow for analysis. This is why we have included these dimensions in the literature review.

road (cont.)

(YEAR) QUESTION
(Hemsing & Have residents Perception of pedestrian and cyclist

Forbes, been walking and safety: Overall, more residents perceived
2000) cycling more an increase in pedestrian safety and a
(cont.) since the decrease in cyclist safety.
interventions?
Travel habits: In general, residents did not
think they had changed their habits.
(Hyden & What are the Design: before-after. Conflict study: Stable total, car-car The article points to the importance of
Várhelyi, effects of an increase, car-bicycle and car-pedestrian design in the construction of traffic-
2000) experimental Several measurements before (four decrease. Less serious conflicts (from calming measures and in the
scheme involving months) and after (four months for frontal to angular and at lower speeds). evaluation of effects (one of the mini-
the installation of all and a follow-up four years later Estimated number of personal injury roundabouts, whose construction
21 mini- for four of the 21 mini-roundabouts). Ŧ
collisions: -44% (cyclists: -60% ,
Ŧ
differed from that of the other mini-
Ŧ Ŧ
roundabouts (at Measurements of speed, conflicts pedestrians: -80% , drivers: +12% ). roundabouts included in this
intersections (at 12 intersections), videotapes, Note: one of the mini-roundabouts comparison study, was found to have
whose daily road user counts, interviews with produced an increase in the risk of increased the risk of personal injury
Ŧ Ŧ
volume can go as pedestrians. personal injury collisions of +200% at collisions by about 200% ).
high as one intersection. At a four-year follow-up,
23,500 vehicles) Note: The calculation of emissions is one mini-roundabout, whose centre island The anomaly represented by this
in Växjö, Sweden, discussed in detail in Várhelyi had been enlarged, was found to have roundabout also reduced the positive
on the risk of (2002). The same results are complicated the route of cyclists. effects on conflicts.
personal injury reported. The very small number of pedestrians
collisions, on Yielding: More respect for the priority of
other drivers (70% before and 91% questioned regarding their perception
yielding, on CO Ŧ of the ease of crossing makes it
and NOx after) , of cyclists (13% before and 77%
Ŧ
after) and of pedestrians (24% before necessary to interpret these results
emissions and on Ŧ with caution.
noise levels? and 51% after) .

road (cont.)

(YEAR) QUESTION
(Hyden & What was the Emissions: See Várhelyi (2002).

Várhelyi, perceived effect Ŧ Ŧ
2000) on ease of Noise: Reduction of 3.9 , 4.2 and
Ŧ
(cont.) crossing for 1.6 dB LAeq for the three intersections
pedestrians? studied.
Perception: Of the 26 pedestrians
questioned four years later, 40% found
that crossing was easier at mini-
roundabouts than at other intersections
and 20% found it more difficult.
(Jones et Does the Design: Ecological study of small City A: Areas in the most deprived fourth The audit of measures was limited to
al., 2005) distribution of area of intervention (two cities in the had 4.8* (confidence interval [CI] 95%: speed humps, road narrowings, and
traffic-calming UK). Longitudinal analysis of injury 3.71 to 6.22) times the number of traffic- road closures.
measures reflect rates with cross-sectional control for calming measures/1000 residents as
social and modes of travel to school. those in the most affluent fourth. The The article does not provide
economic personal injury collision rate for the whole information about similarities or
inequalities? Participants: Sampling of children city decreased from 6.98 to 4.84 between differences between road networks in
aged 4 to 16 years old between 1992–1994 and 1998–2000, which the most and least deprived areas
Are they related 1992 and 2000. corresponds to a significant decrease of (often the most affluent areas are
to personal injury 2.14* (CI 95%: -2.81 to -1.48). In the most already calmed, for example by grids
collision rates for Indicators: Distribution of made up of cul-de-sacs, a possibility
interventions/socioeconomic status affluent areas, the rate went from 9.53 to
child pedestrians? 5.85, which corresponds to a significant here, given the significant differences
of areas and changes in personal
reduction of 3.68* (CI 95%: -5.28 to in the length of road networks in the
injury collision rates/1000 residents most and least deprived areas,
(police data). -2.13). The rate for the most deprived
areas went from 3.21* (CI 95% 2.27 to particularly in city A).
4.54) times that of the most affluent areas This article evaluates the effects of
to 2.01* (CI 95%: 1.45 to 2.87), which traffic-calming measures on an area
corresponds to a non-significant reduction including more than one road, but it
in inequalities. does not indicate whether or not the
City B: Areas in the most deprived fourth measures were planned and

road (cont.)

(YEAR) QUESTION
(Jones et had 1.88* (CI 95%: 1.46 to 2.42) times the implemented so as to function in a
al., 2005) number of traffic-calming measures/1000 systemic manner.
(cont.) residents as those in the most affluent
fourth. The personal injury collision rate
did not vary significantly between 1992-
1994 and 1998–2000 and inequalities
were not reduced significantly, the rate for
the most deprived areas having gone
from 4.27* (CI 95%: 2.51 to 7.28) times
that of the most affluent areas to 3.96*
(CI 95%: 2.26 to 6.95).
Variations in the personal injury collision
rates are inversely correlated to the
density of traffic-calming measures (n.b.
of traffic-calming measures/km of road)
(r=-0.769*, p=0.026).
Similar proportions of children walked to
school in both cities.
(Kamphuis How are Design: cross-sectional study. Traffic-calming measures are one of the Bike paths and lanes can be
et al., 2008) individual and four characteristics of road design that are introduced as traffic-calming
environmental Measurement of the practice of significantly associated with recreational measures (Macbeth, 1998). This
characteristics cycling for recreational purposes: at cycling (odds ratio [OR]: 2.90*, confidence article does not indicate why they
(including traffic- least once a month vs. never. interval [CI] 95%: 1.19 to 7.02). Bike were introduced or how they modified
calming Survey by postal questionnaire on paths and lanes on roads, which can be the road environment (e.g., by
measures) cycling habits and individual used to calm traffic, are another (OR: narrowing the road).
associated with characteristics (age, sex, education, 5.40*, CI 95%: 1.29 to 22.60).
area Also, some traffic-calming measures
occupation); 2349 participants, (e.g., curb extensions) were classified
socioeconomic 2203 valid questionnaires.
inequalities in the under “safety” as “crossing aids,”
Melbourne region Environmental audits using a which excludes them from the
(Australia) and Geographic Information System analysis of traffic-calming measures
between-area (GIS) for some environmental (considered alone, this category

road (cont.)

(YEAR) QUESTION
(Kamphuis differences in characteristics (road network design produces no significant association,

et al., 2008) recreational [includes some calming measures], but it could be otherwise were it
(cont.) cycling? safety [includes other calming combined with other characteristics
measures], destinations, and classified as “calming” features).
aesthetics).
This article evaluates the effects of
Analysis using multilevel logistic traffic-calming measures on an area
regression. including more than one road, but it
does not indicate whether or not the
measures were planned and
implemented so as to function in a
systemic manner.
(Owen, What are the Design: before-after study with Ambient concentrations of air pollutants Controlled for background
2005) effects on air control site. measured were not significantly affected concentrations and weather.
quality of six by the interventions.
20-mph (32-km/h) The article considered Controlled for traffic volumes, but use
zones 0.5 km x 0.5 km zones with signs For the five zones for which calculations of average speeds instead of real
implemented in and traffic-calming measures that were performed, emissions per vehicle speeds (thus eliminating the effect of
Ŧ Ŧ
North West modified road design (e.g., speed rose (0 to +5% NOx, +11 to +34% speed variations).
England? humps). benzene), but when variations in traffic
volume were taken into account, Imprecision of instruments measuring
Concentrations of NO2 and benzene emissions were shown to have decreased air quality (±25% for the diffusion
measured before (5 to 9 months) in the majority of zones (+8, -18, -9 , -32 tubes).
and after (3 to 12 months) at three Ŧ
and -80% NOx; +22, +3, -15, -32
sites within each zone and at one Ŧ
and -76% benzene).
outside control site. Use of diffusion
tubes and thermal desorption tubes. The dispersion model indicates that the
contribution of traffic within the zone to
Analysis of standard deviation and the ambient concentrations of air
temporal variation. pollutants measured is weak (4 to 14%
Ŧ
Ŧ
Vehicle emissions (NOx and of NOx; 0 to 3% of NO2).
benzene) estimated on the basis of
average speeds and traffic volumes.

road (cont.)

(YEAR) QUESTION
(Owen, Use of a dispersion model to

2005) determine the contribution of traffic
(cont.) in the zone to ambient
concentrations of air pollutants.
(Várhelyi, What are the Design: before-after. At the 20 mini-roundabouts replacing The article points to the importance of
2002) effects of an intersections without traffic lights: the implementation context of
experimental Recording of driving cycles before significant increase in CO (+13%*) and interventions (e.g., is an intersection
scheme involving (n=600) and after (n=800) the non-significant increase in NOx (+8%) and with or without traffic lights being
the installation of installation of mini-roundabouts to gas (+8%) for users of main roads, and replaced?) and the varying potential
21 mini- construct standard driving cycles. non-significant decrease for those using effects at intersections (main road vs.
roundabouts (at Vehicles were selected randomly secondary roads secondary road).
intersections and followed by a car equipped to (-20% CO, -15% NOx, -21% gas).
whose daily record distance travelled twice per Combined, the mini-roundabouts Calculations were limited to gas-
volume can go as second. The driver of the car with produced a non-significant increase in powered cars, despite the fact that
high as the measuring equipment copied the emissions (+6% CO and +4% NOx) and heavy vehicles comprised 7% of
23,500 vehicles) movements of the other drivers gas consumption (+3%). With regard to traffic. Calculations were based on the
in Växjö, Sweden, without their knowledge. the mini-roundabout replacing an assumption that 30% of cars were
on gas intersection with traffic lights, non- equipped with catalytic converters.
Traffic counting performed with
consumption and automatic counters and manually. significant reductions in emissions and Only one statistically significant result.
CO and gas consumption were calculated
NOx emissions? Model used to calculate gas (-29% CO, -21% NOx, -28% gas). The article does not indicate precisely
consumption and emissions. when, with respect to the installation
No significant change in traffic volume. of the mini-roundabouts, the effects
were measured.
(Zein et al., What effects do Design: before-after. The four interventions reduced the This article does not meet recognized
Ŧ
1997) traffic-calming frequency of collisions (-40% on scientific standards. In particular, it
schemes have on Four areas in the Greater Vancouver average) and the annual cost of collisions provides very little information about
the frequency and and Victoria regions were examined. Ŧ
(-38% on average).These averages data sources and about the years
severity of The interventions are described with obscure marked differences between the being compared, which makes it
collisions? the help of maps showing the types interventions (for example, a stop-sign impossible to reproduce and compare
and locations of measures. Before- scheme in Burnaby and a scheme the measured reductions in general
after study (the authors reviewed involving highly diverse physical trends.
data for one year before the measures in the West End area),

road (cont.)

(YEAR) QUESTION
(Zein et al., interventions and collected data significant differences between areas (for However, the interventions are very
Ŧ
1997) following the intervention, but example, -18% for collision frequency in well described and the level of detail
Ŧ
(cont.) without indicating when). West End, Vancouver, and -60% in provided in the analysis tables makes
Burnaby) and between indicators (in the results more informative than
Indicators: collision frequency Ŧ
Burnaby, -60% for collision frequency many studies meeting the recognized
(police data) and collision severity Ŧ
and -48% for claims costs). scientific standards.
(annual costs of insurance claims
according to the Insurance
Corporation of British Columbia).

o
Publication N : 1424
www.ncchpp.ca

ReviewLiteratureTrafficCalming en

Uploaded by

Copyright:

Available Formats

ReviewLiteratureTrafficCalming en

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ReviewLiteratureTrafficCalming en

Uploaded by

Copyright:

Available Formats

National Collaborating Centre

for Healthy Public Policy

URBAN TRAFFIC CALMING

URBAN TRAFFIC CALMING

National Collaborating Centre for Healthy Public Policy I

ABOUT THE NATIONAL COLLABORATING

National Collaborating Centre for Healthy Public Policy III

National Collaborating Centre for Healthy Public Policy V

VI National Collaborating Centre for Healthy Public Policy

National Collaborating Centre for Healthy Public Policy VII

National Collaborating Centre for Healthy Public Policy IX

National Collaborating Centre for Healthy Public Policy XI

National Collaborating Centre for Healthy Public Policy 1

2 National Collaborating Centre for Healthy Public Policy

Figure 1 An example of the black-spots

National Collaborating Centre for Healthy Public Policy 3

To this end, the installation of calming measures on a road network is systematically

Figure 2 An example of the area-wide

4 National Collaborating Centre for Healthy Public Policy

Table 1 Principal characteristics of the two approaches to traffic calming

Black-spots approach Area-wide approach

Systematic bias of roads and of the public road

Improvement of road safety and, more broadly,

1.2 INTERVENTION LOGIC

National Collaborating Centre for Healthy Public Policy 5

Vertical Horizontal Reduction of Alteration of

Figure 3 Classification of principal traffic-calming measures

Although the effectiveness of traffic-calming measures at reducing traffic speeds and

TRAFFIC-CALMING PRINCIPAL MECHANISMS EFFECTS ON HEALTH

Reduction of environmental noise

Figure 4 Intervention logic of traffic calming

6 National Collaborating Centre for Healthy Public Policy

1.3 WORKING PROPOSAL AND ORGANIZATION OF THE LITERATURE REVIEW

National Collaborating Centre for Healthy Public Policy 7

2.1 DOCUMENTARY SOURCES AND RESEARCH METHODS

The following databases were searched:

National Collaborating Centre for Healthy Public Policy 9

• Transportation Demand Management Encyclopedia, Victoria Transport Policy Institute

2.2 INCLUSION AND EXCLUSION CRITERIA

10 National Collaborating Centre for Healthy Public Policy

Methodological quality of studies

National Collaborating Centre for Healthy Public Policy 11

2.3 SELECTED DOCUMENTS

2.4 PERUSAL AND EXTRACTION OF EVALUATIVE DATA

2.5 CLASSIFICATION OF EVALUATIVE STUDIES

12 National Collaborating Centre for Healthy Public Policy

The classification of interventions involving several calming measures installed on a street

This classification of studies according to the geographic extent of the interventions

National Collaborating Centre for Healthy Public Policy 13

3 SYNTHESIS OF DATA DRAWN FROM THE LITERATURE

3.1.1 Mechanisms of action

National Collaborating Centre for Healthy Public Policy 15

Figure 5 Narrowing of field of vision as

Figure 6 Probability that a pedestrian will die

16 National Collaborating Centre for Healthy Public Policy

c) Reduction of motorized traffic volume

d) Reduction of the number of points of conflict

Figure 7 Number of points of conflict (black dots on the

National Collaborating Centre for Healthy Public Policy 17