The impact of COVID-19 on road safety in Canada and the United States

Abstract

The COVID-19 pandemic has led to the implementation of unprecedented public health measures. The effect of these lockdown measures on road safety remain to be fully understood, however preliminary data shows reductions in traffic volume and increases in risky driving behaviors. The objective of the present study is to compare self-reported risky driving behaviors (speeding, distracted driving, drinking and driving, and drugged driving) during the pandemic in Canada and the U.S. to determine what differences exist between these two countries. Data was collected using the Road Safety Monitor (RSM), an annual online public opinion survey that investigates key road safety issues, administered to a representative sample of N = 1,500 Canadian drivers and N = 1,501 U.S. drivers. Respondents were asked about the likelihood of engaging in risky driving during the pandemic as compared to before COVID-19. Results show the majority of respondents indicated their behavior did not change, and most positively, a small proportion reported they were less likely to engage in these risky driving behaviors. However, notable proportions indicated they were more likely to engage in risky driving behaviors during the pandemic, as compared to before COVID-19. Of those who indicated this, U.S. drivers had significantly higher percentages compared to their Canadian counterparts. Behaviors most often reported by this sub-section of drivers who admit to being more likely to engage in risky driving during the pandemic were speeding (7.6%) and drinking and driving (7.6%) in the U.S., and speeding (5.5%) and distracted driving (4.2%) in Canada. Logistic regression results confirm that country was a significant factor, as U.S. drivers had greater odds of reporting they were more likely to engage in these risky driving behaviors, with the exception of speeding. Age also had a significant effect, as increasing age was associated with lower odds of reporting that these risky driving behaviors were more likely during the pandemic. Conversely, sex did not have a significant effect. Overall, the current findings suggest that a small proportion of drivers reported being more likely to engage in risky driving behaviors and the pandemic may have led to changes in the profiles of those drivers engaging in risky driving behaviors during lockdown measures. These results have important implications for policies and can inform how to manage road safety during future lockdowns.

Full text

Since January 2020 Elsevier has created a COVID-19 resource centre with
free information in English and Mandarin on the novel coronavirus COVID-
19. The COVID-19 resource centre is hosted on Elsevier Connect, the
company's public news and information website.
Elsevier hereby grants permission to make all its COVID-19-related
research that is available on the COVID-19 resource centre - including this
research content - immediately available in PubMed Central and other
publicly funded repositories, such as the WHO COVID database with rights
for unrestricted research re-use and analyses in any form or by any means
with acknowledgement of the original source. These permissions are
granted for free by Elsevier for as long as the COVID-19 resource centre
remains active.

Accident Analysis and Prevention 160 (2021) 106324
Available online 6 August 2021
0001-4575/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
The impact of COVID-19 on road safety in Canada and the United States
W.G.M. Vanlaar
a
,
*
, H. Woods-Fry
a
, H. Barrett
a
, C. Lyon
a
, S. Brown
a
, C. Wicklund
b
, R.
D. Robertson
a
a
Trafc Injury Research Foundation, Canada
b
Trafc Injury Research Foundation USA, Inc., USA
ARTICLE INFO
Keywords:
Road safety
Driver behavior
COVID-19
Speeding
Distracted driving
Alcohol-impaired driving
Drugged driving
ABSTRACT
The COVID-19 pandemic has led to the implementation of unprecedented public health measures. The effect of
these lockdown measures on road safety remain to be fully understood, however preliminary data shows re-
ductions in trafc volume and increases in risky driving behaviors. The objective of the present study is to
compare self-reported risky driving behaviors (speeding, distracted driving, drinking and driving, and drugged
driving) during the pandemic in Canada and the U.S. to determine what differences exist between these two
countries. Data was collected using the Road Safety Monitor (RSM), an annual online public opinion survey that
investigates key road safety issues, administered to a representative sample of N =1,500 Canadian drivers and N
=1,501 U.S. drivers. Respondents were asked about the likelihood of engaging in risky driving during the
pandemic as compared to before COVID-19. Results show the majority of respondents indicated their behavior
did not change, and most positively, a small proportion reported they were less likely to engage in these risky
driving behaviors. However, notable proportions indicated they were more likely to engage in risky driving
behaviors during the pandemic, as compared to before COVID-19. Of those who indicated this, U.S. drivers had
signicantly higher percentages compared to their Canadian counterparts. Behaviors most often reported by this
sub-section of drivers who admit to being more likely to engage in risky driving during the pandemic were
speeding (7.6%) and drinking and driving (7.6%) in the U.S., and speeding (5.5%) and distracted driving (4.2%)
in Canada. Logistic regression results conrm that country was a signicant factor, as U.S. drivers had greater
odds of reporting they were more likely to engage in these risky driving behaviors, with the exception of
speeding. Age also had a signicant effect, as increasing age was associated with lower odds of reporting that
these risky driving behaviors were more likely during the pandemic. Conversely, sex did not have a signicant
effect. Overall, the current ndings suggest that a small proportion of drivers reported being more likely to
engage in risky driving behaviors and the pandemic may have led to changes in the proles of those drivers
engaging in risky driving behaviors during lockdown measures. These results have important implications for
policies and can inform how to manage road safety during future lockdowns.
1. Introduction
Since the World Health Organization (WHO)’s declaration of a
world-wide pandemic in March 2020 there has been a substantial
decrease in vehicle miles traveled (VMT) and trafc volumes (Wagner
et al., 2020; Shilling and Waetjen, 2020; Stavrinos et al., 2020; City of
Calgary, 2020; Harantov´
a et al., 2020). In the United States, the National
Center for Statistics and Analysis (NCSA) estimated the VMT during rst
half of 2020 was 16.6 % lower than the same period in 2019. An analysis
from the state of California showed a 30%–50% reduction in trafc
volume after the Governor’s stay-at-home order on March 21st,
compared to before the order (Shilling and Waetjen, 2020). In Canada,
an analysis of weekday trafc volume in Calgary, Alberta shows a 38%–
56% decrease beginning March 16th when a local state of emergency
was declared, compared to the beginning of March (City of Calgary,
2020). These reductions have largely been attributed to the shift towards
working from home and the closure of entertainment and leisure in-
dustries (Stavrinos et al., 2020; Harantov´
a et al., 2020).
It is hypothesized that the decrease in trafc volume was conducive
to an increase in speeding drivers. Wagner et al. (2020) suggests that
decreased trafc volume and congestion, particularly during the
beginning of the pandemic, coupled with the reduction in law
* Corresponding author.
E-mail address: wardv@tirf.ca (W.G.M. Vanlaar).
Contents lists available at ScienceDirect
Accident Analysis and Prevention
journal homepage: www.elsevier.com/locate/aap
https://doi.org/10.1016/j.aap.2021.106324
Received 8 January 2021; Received in revised form 3 June 2021; Accepted 28 July 2021

Accident Analysis and Prevention 160 (2021) 106324
2
enforcement, produced an environment with more opportunities for
speeding. In California, average speeding during peak hours increased
between 10 and 15 miles-per-hour over the period of March to May 2020
(Hughes et al., 2021). Similarly, in British-Columbia, Saanich police saw
a 700% increase in the number of cars impounded for excessive speeding
during March 2020 (Chan, 2020). In Alberta, Edmonton police have
recorded a 200% increase in drivers speeding more than 50 km/h over
the speed limit after the local state of emergency was declared on March
20th (Heidenreich, 2020). In Ontario, Toronto Police reported that from
March 15 to March 31, there was a 35% increase in speeding tickets and
an almost 200% increase in stunt driving (i.e., excessive speeding such
as driving at least 50 km/h over the speed limit in certain Canadian
jurisdictions) compared to the same period last year (City of Toronto,
April 2020).
In addition to speeding, preliminary data suggests an increase in
impaired driving. A study conducted by the National Highway Trafc
Safety Administration (NHTSA) showed an increase in alcohol-impaired
and drug-impaired driving during March-July 2020. Compared with
data collected from September 2019 to March 16, 2020, data from
March 17 to July 2020 revealed a 26.3% increase in seriously or fatally
injured drivers who tested positive for alcohol, and a 27.4% increase in
seriously or fatally injured drivers testing positive for at least one active
drug. In particular, active THC was more prevalent among fatally
injured drivers compared with alcohol (32.7% versus 28.3%), and
opioid use among drivers doubled (Thomas et al., 2020).
Preliminary data from the rst two quarters of 2020 also demon-
strates fewer motor vehicle collisions (both fatal and non-fatal) in the U.
S. and Canada (Carter, 2020; Katrakazas et al., 2020; Saladi´
e et al.,
2020; Shilling and Waetjen, 2020; Wagner et al., 2020). In the U.S., data
showed an average decrease in all trafc crashes during this period
ranging from 41% to 76% (Katrakazas et al., 2020; Saladi´
e et al., 2020;
Shilling and Waetjen, 2020). However, research also showed that while
crashes decreased, there was a 25% average increase in severe crashes,
possibly a result of increases in risky driver behaviors such as speeding
(Hughes et al., 2021). Further, compared to pre-COVID-19 baseline,
road crashes decreased by 50% and fatal road crashes decreased by 10%,
but casualty collisions increased by 6%, which was a result of the higher
proportion of single-vehicle collisions during this COVID-19 study
period (Carter, 2020). Similarly, an analysis of collision data in Florida,
New York, and Massachusetts showed a steady decline in the number of
vehicle collisions as the COVID-19 infection rate increased (Sutherland
et al., 2020). In Canada, Saskatoon police reported a 79.3% decrease in
persons injured in motor vehicle collisions in May 2020 compared to
May 2019 (Saskatoon Police Service, 2020), and Halifax police reported
a 63.6% decrease in collisions in April 2020 compared to April 2019
(Halifax Regional Municipality, 2020).
In summary, these preliminary data show the public safety measures
put in place to combat COVID-19 resulted in decreased road use (spe-
cically during peak weekday travel times) and suggest a rise in risky
road user behavior, such as speeding and impaired driving. Further,
some preliminary crash data show a decrease in both fatal and non-fatal
road crashes while others indicated there was a decrease in overall
crashes but an increase in more severe crashes. These ndings have
guided the present study to further investigate the preliminary reports of
increases to risky driving behavior during the pandemic to help facilitate
a greater understanding of the effects of the COVID-19 pandemic on
risky road behavior, in line with the research questions laid out in
Vingilis et al. (2020). Additionally, research examining the role of region
(Canada, United States) on risky driving behavior during the pandemic
can help determine if risky driving behaviors during the pandemic varies
between country. The role of age and sex in risky driving behavior
during the pandemic is also relevant, as this can help elucidate if certain
drivers are more likely to engage in risky driving behaviors during the
pandemic, and whether these drivers are representative of the typical
risky driver prole demonstrated throughout the literature. Therefore,
the primary objective of this research paper is to compare the
self-reported attitudes and risky driving behaviors in Canada and the U.
S. to determine what quantitative differences exist between these two
regions, and to further investigate the association between age, sex, with
self-declared risky road behaviors during the pandemic to determine if
certain sub-populations of drivers are more likely to engage in these
behaviors during the pandemic as compared to before COVID-19.
2. Methods
2.1. Data sources
Data on key road safety issues in Canada have been collected as part
of the Trafc Injury Research Foundation (TIRF) series of Road Safety
Monitor (RSM) surveys since 2002. The survey instrument contains a
core set of items that are asked each year to provide information on
trends in attitudes, opinions and behaviors of Canadian drivers. Sup-
plementary to this, questions probing into special, topical, and emerging
issues are included (e.g., automated vehicles, distracted driving, COVID-
19 and road safety).
Data on key road safety issues have also been collected by the Trafc
Injury Research Foundation USA, Inc. (TIRF USA) series of USA Road
Safety Monitor (USA RSM) surveys since 2015. Similarly, the survey
instrument contains a core set of items that are asked each year, and
emerging issues are probed with respect to attitudes, opinions and be-
haviors of American drivers.
Results presented here include data from the Canadian RSM and the
USA RSM with a focus on questions specically related to the effects of
the COVID-19 pandemic on road safety. These questions were asked for
the rst time in September 2020. The total number of daily COVID-19
cases in the month of September in Canada was 73 per 100,000 popu-
lation and 356 per 100,000 population in the U.S. (World Health Or-
ganization, 2021). Respondents were asked about various self-reported
risky driving behaviors including speeding, distracted driving, alcohol-
impaired driving, and drug-impaired driving. Respondents were asked
how likely they were to engage in these risky behaviors during the
COVID-19 pandemic, compared to their typical behavior before the
pandemic. As pandemic restrictions varied by country and state/prov-
ince, respondents were prompted to refer to the period of time when
COVID-19 restrictions were in fullest effect in their respective area. Two
subsequent reminders to specify this were given throughout this series of
questions.
The Canadian RSM was administered online to a national represen-
tative sample of 1,500 drivers aged 16 years and older who had driven in
the past 30 days and held a valid driver’s licence. The USA RSM was
administered online to a national representative sample of 1,501 drivers
aged 21 years or older who had driven in the past 30 days and held a
valid driver’s licence (Table 1). The survey required an average of
approximately 10 min to complete. Each sample was stratied by region
and weighted according to sex and age to ensure the results were
representative of the national population of their respective country.
Additional details on the overall RSM methodology can be found in a
study on trends in drugged driving (Robertson et al., 2017) and a study
on wildlife vehicle collisions (Vanlaar et al., 2019).
Table 1
Sample characteristics.
Canada U.S. Total
Mean age 46 46.9 N/A
Male (%) 748 (49.87) 751 (50.03) 1,499 (49.95)
Female (%) 750 (50) 750 (49.97) 1,500 (49.98)
Prefer not to say (%) 2 (0.13) 0 (0) 2 (0.07)
Total (%) 1,500 (100) 1,501 (100) 3,001 (100)
W.G.M. Vanlaar et al.

Accident Analysis and Prevention 160 (2021) 106324
3
2.2. Data validation
Data from both RSMs consist of self-reported data on driver attitudes
and behavior, which allows for the timely measurement of current road
safety issues from a large nationally representative sample. Objective
data is often used to validate ndings from self-reported data, as both
data have unique strengths and weaknesses (Tempelaar et al., 2020).
Therefore, correlations between RSM data on self-reported drinking and
driving behavior and the number of alcohol-related fatalities were
calculated for both Canada and the U.S. When assessing the strength of
these correlations, it must be considered that the data collection periods
for the fatality data and the RSM data do not completely overlap. To
illustrate, 2017 fatalities are those that occurred between January to
December 2017 whereas the 2017 RSM self-reported drinking and
driving data spanned October 2016 to September 2017 (because re-
spondents were asked in September about their behavior in the past 12
months).
To validate the Canadian RSM data, the correlation between data on
self-reported drinking and driving behaviors from the Canadian RSM
surveys and the number of alcohol-related fatalities from TIRF’s Na-
tional Fatality Database was calculated during the period of 2004–2017
(Fig. 1). The RSM question of interest investigated how many times in
the past year respondent had driven when they were likely over the legal
limit for alcohol. Alcohol-related fatalities from TIRF’s National Fatality
Database includes collected statistics from police reports, coroners and
medical examiners on all persons fatally injured in motor vehicle crashes
in all jurisdictions across Canada where at least one of the drivers in the
crash (either dying or surviving) was determined to have alcohol in their
system at the time of crash. A medium to strong, signicant correlation
coefcient (rho =0.66, p =0.01) was found between the number of
alcohol-related fatalities from TIRF’s National Fatality Database and
self-reported drinking and driving when probably over the legal limit
from TIRF’s RSM.
In the U.S., a small to medium, non-signicant correlation coefcient
(rho =0.67, p =0.34) was found between the number of alcohol-related
fatalities from the NHTSA Fatality Analysis Reporting System (FARS)
Trafc Safety Fact Sheet (National Center for Statistics and Analysis,
2020) and self-reported drinking and driving when probably over the
legal limit from TIRF’s USA RSM. Note that the dataset to calculate
correlations with the U.S data is smaller (i.e., only four years of data to
compare), hence less powerful than the Canadian data (i.e., 14 years of
data to compare).
2.3. Data analysis
Descriptive analyses were conducted to assess the impact of COVID-
19 on self-reported risky driving behaviors in each country. Self-
reported risky driving behaviors were asked on a scale from 1 (far less
likely) to 5 (far more likely). For the purpose of analyses, respondents
who indicated values from 1 to 3 were coded as less likely to engage in
the behavior, and respondents who reported values of 4 or 5 were coded
as more likely to engage in the behavior. Follow-up questions were
asked when respondents indicated a value of 4 or 5. A two-sample test of
proportions was used to evaluate whether there was a statistically sig-
nicant difference between the prevalence of self-reported risky driving
behaviors during the pandemic in Canada and the U.S.
For each risky road behavior (excessive speeding, distracted driving,
alcohol and driving, and drugs and driving) a binary logistic regression
model was used to study the association between explanatory variables
and the dependent variable. Demographics (country, sex, age), number
of trafc tickets received in the past 12 months, number of miles driven
in a typical month, and previous injury in a motor vehicle collision are
treated as the explanatory variables. Explanatory variables were chosen
based on the existing literature of risky driver proles (Fergusson et al.,
2003; Blows et al., 2005; Iversen, 2004). Specically, country as an
explanatory variable was chosen to determine if the effect of the
pandemic on road safety was different between Canada and the U.S.,
where local measures and government regulations have varied (Gupta
et al., 2021).
For each model, the dependent variable indicates if a behavior was
unlikely (0 =unlikely) or likely (1 =likely) during the COVID-19
pandemic, compared to typical behavior before the pandemic. Odds
ratios (OR) and 95% condence intervals (CI) were calculated for each
independent variable. Age was rescaled in 10-year increments, hence
the OR for age corresponds to the effect of a 10-year increase in age on
the dependent variable. All analyses were conducted with Stata/MP
14.1 for Windows 64-bit x86-64 (StataCorp, 2015). The ‘svy’ procedure
in Stata was used to weigh and stratify the data for both descriptive and
multivariate analyses.
3. Results
3.1. Descriptive analysis
3.1.1. Excessive speeding
Respondents were asked how likely they were to excessively exceed
the posted speed limit during COVID-19 as compared to before the
pandemic. A total of 5.5% of Canadians (CI: 4.3–7.1) admitted they were
796844
896848
786
706733
632
629
525487515
532
460
5.8
7.5
7.7
8.2
5.2
5.6
6.0
5.4
3.6
4.8
6.6
4.2
4.6
5.1
0
1
2
3
4
5
6
7
8
9
0
100
200
300
400
500
600
700
800
900
1000
20042005200620072008 2009 20102011201220132014201520162017
% of drivers who drove w hen thought to be
over legal limit
Number of alcohol-related fatalities
alcohol-related fat alities drove when t hey thought t hey were over the legal limit
Fig. 1. Percentage of alcohol-related fatalities and self-reported drinking and driving behavior in Canada.
W.G.M. Vanlaar et al.

Accident Analysis and Prevention 160 (2021) 106324
4
more likely to do so, as compared to before the pandemic, whereas 7.6%
of U.S. drivers (CI: 6.2–9.2) reported this (Table 2). The proportion of U.
S. drivers reporting this was signicantly higher than the proportion of
Canadian drivers (z = − 2.32, p =0.02). A sizable proportion of re-
spondents in both Canada and the U.S. reported there was no change in
their speeding behaviors during the pandemic as compared to before
COVID-19 (77.5% in Canada, CI: 74.9–79.8, versus 69.9% in U.S., CI:
66.8–72.1; z =4.73, p <0.001). Finally, some respondents reported
they were less likely to speed during the pandemic as compared to
before COVID-19, with 16.9% of respondents in Canada (CI: 14.9–19.3)
reporting this, compared to 22.8% in the U.S. (CI: 20.5–25.3) (z =
−4.05, p <0.001).
3.1.2. Distracted driving
Respondents were asked how likely they were to have been
distracted while driving during COVID-19 in comparison to before the
pandemic. In Canada, 4.2% of respondents (CI: 3.2–5.5) indicated they
were more likely to be distracted, as compared to before the pandemic,
whereas 6.8% of U.S. drivers admitted to this (CI: 5.5–8.4) (z = − 3.13, p
=0.002) (Table 3). Respondents who indicated they were more likely to
have been distracted while driving during COVID-19 in comparison to
before the pandemic were asked to specify the most frequent cause of
distraction. Over half (53.5%) of these respondents in Canada (CI:
39.1–67.3) and 44.7% in the U.S. (CI: 34.2–55.7) stated competing
thoughts unrelated to driving were the primary reason for their
distraction while driving. Largely, respondents indicated there was no
change, and that they were not more likely to be distracted during the
pandemic as compared to before COVID-19 (79.3% in Canada, CI:
76.9–81.6, versus 71.8% in U.S., CI: 69.2–72.3; z =4.78, p <0.001).
There were 16.5% of Canadians (CI: 14.4–18.8) who reported being less
likely to be distracted while driving during the pandemic as compared to
before COVID-19, and 21.4% of U.S. drivers also reported this
(CI:19.2–23.8) (z = − 3.42, p =0.001).
3.1.3. Alcohol and driving
Respondents were asked how likely they were to have driven a motor
vehicle within two hours of using alcohol during COVID-19 as compared
to before the pandemic. In Canada, 2.4% of drivers (CI: 1.7–3.5)
admitted they were more likely to drink and drive during the pandemic
as compared to before COVID-19 (Table 4). In the U.S., the percentage of
drivers admitting this was more than three times that of Canada (7.6%)
(CI: 6.2–9.3) (z = − 6.53, p <0.001). Three-quarters of Canadian drivers
(75.4%, CI: 72.8–77.9) indicated there was no change in their behavior
during the pandemic as compared to before COVID-19, and 68.7% of U.
S. drivers also reported this (CI: 65.9–71.3) (z =4.09, p <0.001).
Finally, similar percentages of Canadian and U.S. drivers reported being
less likely to drive within two hours of using alcohol during the
pandemic as compared to before COVID-19 (22.1%, CI: 19.8–24.7, and
23.7%, CI: 21.3–26.2, respectively; z = − 1.04, p =0.29).
3.1.4. Drugs and driving
Respondents were asked how likely they were to have driven a motor
vehicle within two hours of using drugs during COVID-19 as compared
to before the pandemic. There was a small percentage of Canadians who
reported they were more likely to take drugs and drive during the
pandemic, as compared to before COVID-19 (2.2%, CI: 1.4–3.3). A
signicantly larger percentage of U.S. drivers (6.2%, CI: 4.9–7.8) (z =
−5.46, p <0.001) reported this (Table 5). Respondents who indicated
they were more likely to consume drugs and drive during the pandemic
as compared to before COVID-19 were also asked to indicate the sub-
stance most frequently used before driving. Prescription drugs that may
affect driving were most frequently used before driving by 46% of Ca-
nadian respondents (CI: 24.8–68.8), and 39.8% of U.S. respondents (CI:
28.6–52.2). Marijuana was most frequently used by 33.5% of Canadian
respondents (CI: 15.7–57.7) versus 42.6% of U.S. respondents (CI:
31.6–54.4) before driving. Illegal drugs was most frequently used before
driving by 20.4% of Canadian respondents (CI: 7.7–44.2) and 17.6% of
U.S. respondents (CI: 10.4–28). A large proportion of both Canadian
(77.3%, CI: 74.8–79.7) and U.S. drivers (71.6%, CI: 68.5–73.7) reported
there was no change in the likelihood of having driven within two hours
of using drugs during the pandemic as compared to before (z =3.59, p <
0.001). A similar percentage of Canadian and U.S. drivers reported they
were less likely to drive under the inuence of drugs during the
pandemic as compared to before COVID-19 (20.6%, CI: 18.3–22.9, and
22.7%, CI: 20.3–25.1, respectively; z =1.40, p =0.163).
3.2. Multivariate analyses
3.2.1. Factors predicting self-reported excessive speeding
The logistic regression model for self-reported likelihood of excessive
speeding during the pandemic as compared to before COVID-19 is pre-
sented in (Table 6). No signicant difference in self-reported likelihood
of excessive speeding during the pandemic as compared to before
COVID-19 was observed between Canada and the U.S when controlling
for all other factors (OR =1.32, CI: 0.92–1.89). In terms of sex differ-
ences, females and males did not signicantly differ either (OR =0.86,
CI: 0.61–1.22). Age was a signicant factor in self-declared excessive
speeding during the pandemic as compared to before COVID-19. For
every ten-year increase in age, the odds of reporting being more likely to
excessively speed during the pandemic as compared to before decreased
by 25.6% (OR =0.74; CI: 0.66–0.83, p <0.001). The number of trafc
tickets received was also a signicant factor in this behavior. Compared
to respondents with less than two tickets, those with two or more tickets
had increased odds (207.3%) of reporting being more likely to exces-
sively speed during the pandemic as compared to before COVID-19 (OR
=3.07; CI: 1.66–5.68, p <0.001). Typical miles driven in a month (OR
Table 2
Comparison between Canada and the U.S. of the likelihood of self-reported
excessive speeding during the pandemic.
Canada U.S. P-value
More likely 5.5% 7.6% P <0.05
No change 77.5% 69.9% P <0.01
Less likely 16.9% 22.8% P <0.01
Table 3
Comparison between Canada and the U.S. of the likelihood of self-reported
distracted driving during the pandemic.
Canada U.S. P-value
More likely 4.2% 6.8% P <0.01
No change 79.3% 71.8% P <0.01
Less likely 16.5% 21.4% P <0.01
Table 4
Comparison between Canada and the U.S. of the likelihood of drinking and
driving during the pandemic.
Canada U.S. P-value
More likely 2.4% 7.6% P <0.01
No change 75.4% 68.7% P <0.01
Less likely 22.1% 23.7% n.s.
Table 5
Comparison between Canada and the U.S. of the likelihood of consuming drugs
and driving during the pandemic.
Canada U.S. P-value
More likely 2.2% 6.2% P <0.01
No change 77.3 71.6% P <0.01
Less likely 20.6% 22.7% n.s.
W.G.M. Vanlaar et al.

Accident Analysis and Prevention 160 (2021) 106324
5
=1.04, CI: 0.99–1.08) and previous injury in a motor vehicle collision
(OR =1.09, CI: 0.72–1.67) were not signicant factors in self-reported
excessive speeding during the pandemic as compared to before
COVID-19.
3.2.2. Factors predicting self-reported distracted driving
The logistic regression model for self-reported likelihood of
distracted driving during the pandemic as compared to before COVID-19
shows country was a signicant predictor of self-reported distracted
driving behavior during the pandemic as compared to before COVID-19
(OR =1.49; CI: 1.03–2.16, p =0.034) (Table 6). Respondents in the U.S.
were 49.2% more likely than Canadians to report they were distracted
while driving during COVID-19 as compared to before the pandemic. No
signicant difference in this self-reported behavior was found between
females and males (OR =1.15, CI.: 0.79–1.66). However, age was a
signicant predictor, as every ten-year increase in age corresponded to a
31.3% decrease in the odds of reporting distracted driving behavior
during the pandemic as compare to before (OR: 0.69; CI: 0.61–0.78, p <
0.001). The number of trafc tickets was also a signicant factor in this
behavior. Compared to respondents with less than two tickets, those
with two or more tickets were 350% more likely to report they were
distracted while driving during COVID-19 as compared to their typical
behaviour before the pandemic (OR: 4.50; CI: 2.49–8.14, p <0.001).
Typical miles driven in a month (OR: 0.98, CI: 0.93–1.03) and previous
injury in a motor vehicle collision (OR: 1.34, CI: 0.88–2.03) were not
signicant factors in self-reported distracted driving.
3.2.3. Factors predicting self-reported driving within two hours of
consuming alcohol
The logistic regression model for self-reported likelihood of driving
within two hours of consuming alcohol during the pandemic as
compared to before COVID-19 shows country was a signicant predictor
(OR: 3.34; CI: 2.11–5.28, p <0.001) (Table 6). American respondents
were 233.5% more likely than Canadians to report driving within two
hours of consuming alcohol during the pandemic as compared to before
COVID-19. Sex was not a signicant factor (OR: 0.78, CI: 0.53–1.15).
Age was a signicant predictor, as every ten-year increase in age cor-
responded to a 37.2% decrease in the odds of reporting this behavior as
more likely during the pandemic as compared to before COVID-19 (OR:
0.63; CI: 0.54–0.73, p <0.001). The number of trafc tickets was also a
signicant factor. Compared to respondents with less than two tickets,
those with two or more tickets were 138.2% more likely to report this
behavior during the pandemic as compared to before COVID-19 (OR:
2.38; CI: 1.27–4.48, p =0.007). Typical miles driven in a month (OR:
1.02, CI: 0.97–1.07) and previous injury in a motor vehicle collision
(OR: 1.09, CI: 0.67–1.78) were not signicant factors.
3.2.4. Factors predicting self-reported driving within two hours of
consuming drugs
The logistic regression model for self-reported likelihood of driving
within two hours of consuming drugs during the pandemic as compared
to before COVID-19 shows country was a signicant predictor (OR: 2.50;
CI: 1.46–4.28, p =0.001) (Table 6). Respondents in the U.S. were
150.4% more likely than Canadians to report taking drugs and driving
during the pandemic as compared to before COVID-19. Sex was not a
signicant factor (OR: 1.11, CI: 0.70–1.75). Age was a signicant pre-
dictor, as every ten-year increase in age corresponded to a 38% decrease
in the odds of reporting this behavior as more likely during the
pandemic as compared to before COVID-19 (OR: 0.62; CI: 0.52–0.74, p
<0.001). The number of trafc tickets was also a signicant factor.
Compared to respondents with less than two tickets, those with two or
more tickets were 598.5% more likely to report driving within two hours
of consuming drugs during the pandemic as compared to before COVID-
19 (OR: 6.99; CI: 3.74–13.05, p <0.001). Compared to respondents who
had no prior injury from a collision, those who reported being previously
injured in a motor vehicle collision were 86.6% more likely to report
driving within two hours of consuming drugs during the pandemic as
compared to before COVID-19 (OR: 1.87; CI: 1.11–3.14, p =0.019).
Typical miles driven in a month was not a signicant factor in this risky
driving behavior (OR: 0.95, CI: 0.89–1.02).
4. Discussion
The COVID-19 pandemic has affected countries across the globe in
various ways, however reduced trafc volumes due to lockdowns seem
common and have had a direct impact on road safety. With preliminary
data demonstrating increases in speeding and impaired driving, and
fewer but more severe crashes (Hughes et al., 2021; Carter, 2020;
Thomas et al., 2020; Wagner et al., 2020), a closer examination of self-
reported risky driving behaviors can provide greater insight into the
underpinnings of the effects of the pandemic on road users. The current
ndings demonstrated that overall, the majority of drivers in Canada
and the U.S. did not change their behavior as a result of the pandemic
and continued to behave as per usual when driving. A smaller proportion
of drivers indicated they were less likely to engage in risky driving be-
haviors during the pandemic than they were previous to the COVID-19
pandemic. Conversely, a small but notable proportion of drivers indi-
cated they were more likely to engage in risky driving behaviors during
the pandemic, as compared to before COVID-19. Despite these drivers
representing the minority of the current results, these small proportions
are concerning, because although the majority continue to behave as
usual, or have become more cautious while driving, this small propor-
tion of drivers still represents a sizable proportion of the driving popu-
lation that admit to being more likely to take risks during the pandemic.
For example, 5.5% of Canadians admitted they were more likely to
excessively exceed the posted speed limit during COVID-19 as compared
to before the pandemic, which translates to approximately 1.46 million
licensed drivers who admitted doing this behavior more during the
pandemic than before COVID-19.
When examining the results from the Canadian context, we observed
that excessive speeding had the highest percentage of respondents who
reported being more likely to engage in this risky driving behavior
Table 6
Logistic Regression Model for Self-Reported Risky Driving Behaviors During
COVID-19.
Excessive
Speeding
Distracted
Driving
Drinking and
driving
Drugged
Driving
Factors
(reference
category)
Dependent variable: self-declared behavior (0 =unlikely; 1 =
likely)
Odds Ratio (CI 95%)
Country
(Canada)
USA 1.32
(0.92–1.89)
1.49*
(1.03–2.16)
3.34**
(2.11–5.28)
2.50**
(1.46–4.28)
Sex (Female)
Male 0.86
(0.61–1.22)
1.15
(0.79–1.66)
0.78
(0.53–1.15)
1.11
(0.70–1.75)
Age 0.74**
(0.66–0.83)
0.69**
(0.61–0.78)
0.63**
(0.54–0.73)
0.62**
(0.52–0.74)
Trafc tickets
(<2 tickets)
greater than 2
tickets
3.07**
(1.66–5.68)
4.50**
(2.49–8.14)
2.38**
(1.27–4.48)
6.99**
(3.74–13.05)
Typical miles
driven
1.04
(0.99–1.08)
0.98
(0.93–1.03)
1.02
(0.97–1.07)
0.95
(0.89–1.02)
Injured in
motor
vehicle
collision
(No)
Yes 1.09
(0.72–1.67)
1.34
(0.88–2.03)
1.09
(0.67–1.78)
1.87*
(1.11–3.14)
Note. signicance levels are denoted by *p <0.05 **p <0.01.
W.G.M. Vanlaar et al.

Accident Analysis and Prevention 160 (2021) 106324
6
during the pandemic as compared to before COVID-19. This is supported
by other sources of information, such as enforcement data from early
2020 showing that speeding and stunt driving appear to have increased
when compared to data from the same time one year ago. To illustrate,
Edmonton police noted a 200% increase in drivers speeding more than
50 km/h over the speed limit (Heidenreich, 2020), Saanich police saw a
700% increase in the number of cars impounded for excessive speeding
(Chan, 2020), and Toronto Police reported a 35% increase in speeding
tickets and an almost 200% increase in stunt driving (22 tickets issued
for driving in excess of 50 km/h in March 2019 versus 65 tickets issued
in March 2020), and the Ontario Provincial Police saw a 40.1% increase
in the percentage of stunt driving offences compared to 2019 (City of
Toronto, April 2020; C. Goncalves, personal communication, June 19,
2020). After this, distracted driving had the second highest percentage
of respondents who reported being more likely to engage in this risky
driving behavior during the pandemic as compared to before COVID-19.
Finally, the percentage of respondents who reported being more likely to
drive after consuming alcohol, and drive after taking drugs during the
pandemic as compared to before COVID-19 were lower than that of
speeding and distracted driving.
Similarly in the U.S., excessive speeding had the highest percentage
of respondents who reported being more likely to engage in this risky
driving behavior during the pandemic as compared to before COVID-19.
However, unlike in Canada, drinking and driving also had one of the
highest percentages of respondents who reported being more likely to
engage in this risky driving behavior during the pandemic as compared
to before COVID-19. In support of these ndings, State data showed an
increase in speeding in the second quarter of 2020, when compared to
data from 2019, hypothesized to be due to lower trafc volumes and
reduced presence of law enforcement (Wagner et al., 2020). Georgia
State Police cited 140 drivers in only a two-week period for excessive
speeding of over 100 mph (160 km/h) (Wickert, 2020); California
Highway Patrol reported an 87% increase in citations for speeding in
excess of 100 mph (McGreevy, 2020); Chicago’s automated enforcement
cameras issued 14% more speeding citations (Wisniewski, 2020); and
the Virginia Department of Motor Vehicles found that although there
was a decrease in overall crashes, speed-related fatalities accounted for
50% of overall fatalities, compared to 42% in 2019 (Virginia DOT,
2020). In terms of drinking and driving, preliminary data from NHTSA
demonstrated a signicant increase of 26.3% in the prevalence of
alcohol detected in fatally injured road users (Thomas et al., 2020).
Further monitoring of alcohol-impaired driving is necessary, especially
since data on alcohol consumption habits demonstrate a different
pattern during the pandemic, with drivers consuming alcohol at home
alone, rather than in bars or restaurants (Wagner et al., 2020; Woods-Fry
et al., 2020). After this, distracted driving had the third highest per-
centage of respondents who reported being more likely to engage in this
risky driving behavior during the pandemic as compared to before
COVID-19. This is supported by other sources of information, initial
evidence from insurance applications suggests that mobile use while
driving is up by 38% in the U.S. (Whaley, 2020). Finally, the percentage
of respondents who reported being more likely to consume drugs and
drive was the lowest out of the other risky driving behaviors investi-
gated. Preliminary data about drugged driving during the pandemic
limited, however initial reports by NHTSA suggest there was a signi-
cant increase of 27.4% in the prevalence of drugs in seriously and fatally
injured drivers, as well as an increase in the proportion of drivers testing
positive for two or more categories of drugs (Thomas et al., 2020).
When comparing the percentages of Canadian and U.S. drivers who
report being more likely to engage in these risky driving behaviors
during the pandemic as compared to before COVID-19, a signicantly
higher percentage of U.S. drivers report being more likely to engage in
all risky driving behaviors examined. The most remarkable difference
between countries was in the percentage of drivers who reported being
more likely to drink and drive during the pandemic as compared to
before COVID-19, as well as those who reported being more likely to
take drugs and drive during the pandemic as compared to before COVID-
19. The proportion of drivers who reported these behaviors was over 4%
higher in the U.S., compared to Canada. Logistic regression results also
demonstrated that country was a signicant factor in the odds of re-
spondents reporting they were more likely to engage in distracted
driving, driving within two hours of consuming alcohol, and driving
within two hours of taking drugs during the pandemic as compared to
before COVID-19. Not surprisingly, respondent country had the largest
effect on the likelihood of engaging in drinking and driving during the
pandemic, as U.S. drivers were 234% more likely than Canadian drivers
to report engaging in this risky behavior during the pandemic as
compared to before COVID-19.
Other driver characteristics that had signicant effects on these risky
driving behaviors were age and number of trafc tickets in the past 12
months. Specically, every ten-year increase in age was associated with
lower odds of reporting these risky driving behaviors during the
pandemic as compared to before COVID-19. Moreover, compared to
those who had fewer than two tickets, respondents with two or more
trafc tickets in the past 12 months were signicantly more likely to
report engaging in these risky driving behaviors during the pandemic as
compared to before COVID-19. Sex did not have a signicant effect on
any of these risky driving behaviors, suggesting that males and females
were equally likely to report engaging in these behaviors during the
pandemic as compared to before COVID-19. These results are particu-
larly interesting, as existing road safety data suggests males are typically
more likely to engage in risky driving behaviors such as alcohol-
impaired driving, drug-impaired driving, and distracted driving
(Brown et al., 2020; Lyon et al., 2020; Woods-Fry et al., 2020). Unlike
age, it appears the pandemic may have eroded the protective effect sex
can have on females’ propensity for risk taking; how exactly, or why,
this happened, however, is not clear from our data.
The current ndings, complemented by additional sources of pre-
liminary data form a more comprehensive picture that points toward an
increase in excessive speeding in Canada, and an increase in excessive
speeding and drinking and driving in the U.S. These ndings about the
change in self-reported risky driving behaviors in Canadian and U.S.
drivers during COVID-19 restrictions also complement crash and fatality
data as it becomes available. However, more research is needed to
determine the full impact of the COVID-19 pandemic on road safety. As
lockdowns continue and the pandemic restrictions evolve, the long-term
effects of the pandemic on driver behavior must be examined.
The current study possesses certain limitations, primarily regarding
the way COVID-19 restrictions were dened based on the interpretation
of the respondent. When asked to compare their behavior during the
pandemic to their typical behavior before the pandemic, respondents
were instructed to refer to their behavior during the pandemic at the
time when COVID-19 restrictions were in fullest effect in their respective
area. However, it is recognized that the period of time when restrictions
and stay-at-home orders were in place varied between countries, as well
as between states/provinces. While bias due to this variation is not
entirely impossible, we believe the instruction provided to respondents
will have led to a high degree of standardization in responses between
respondents. Furthermore, although the number of COVID-19 cases per
100,000 population in Canada and the U.S. were noted at the time of the
data collection, we acknowledge that there is signicant variability in
the number of COVID-19 cases per region during the data collection
period and results could have been impacted by this. Finally, while self-
report data can indeed suffer from certain biases (such as memory bias
or social desirability bias), our RSM data were validated by correlating
them with fatal crash data. This validation exercise demonstrated a
rather high degree of validity, suggesting low or no bias in our data. This
is consistent with other RSM studies we conducted where similar
external validation demonstrated good validity (e.g., Robertson et al.,
2017; Vanlaar et al., 2012; Vanlaar et al., 2019).
W.G.M. Vanlaar et al.

Accident Analysis and Prevention 160 (2021) 106324
7
5. Conclusions
The impact of COVID-19 on road safety must be carefully monitored
going forward. Despite a large proportion of road users in both Canada
and the U.S. who indicated they did not change their behavior or were
somewhat more cautious, there remains a sizable proportion of drivers
who engaged in risky behaviors on the road more often during the
pandemic compared to their regular behaviors had restrictions not been
in place. These ndings offer a unique and informative perspective
about the change in drivers’ self-reported behavior during the pandemic
and help guide further investigations into risky driving behavior during
subsequent COVID-19 lockdowns and future pandemics.
These ndings suggest that during the current pandemic and possible
future resurgences of COVID-19 or other pandemics, targeted enforce-
ment strategies must focus on speeding and impaired driving in partic-
ular to help reduce and prevent these behaviors. Further, these ndings
suggest the common prole of driver characteristics typically observed
when examining these road safety issues seems to have been somewhat
altered during the pandemic, and it appears the pandemic may have
eroded the protective effect sex can have on females’ propensity for risk
taking. As such, driving safety campaigns during the current pandemic
and future resurgences must be tailored accordingly to ensure effective
public messaging about risky driving behaviors. Future research is
necessary to gain an in-depth understanding of how the pandemic has
affected road safety overall, and to examine the effects of gradual re-
openings and subsequent restrictions on driver behavior to determine
how long-term pandemic restrictions have affected road user behavior.
CRediT authorship contribution statement
W.G.M. Vanlaar: Conceptualization, Data curation, Formal anal-
ysis, Funding acquisition, Investigation, Methodology, Project admin-
istration, Validation, Resources, Visualization, Writing - original draft,
Writing - review & editing. H. Woods-Fry: Data curation, Formal
analysis, Investigation, Visualization, Writing - original draft, Writing -
review & editing. H. Barrett: Writing - original draft, Writing - review &
editing. C. Lyon: Investigation, Resources, Writing - review & editing. S.
Brown: Resources, Writing - review & editing. C. Wicklund: Resources,
Writing - review & editing. R.D. Robertson: Conceptualization, Fund-
ing acquisition, Methodology, Project administration, Supervision, Re-
sources, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgements
The data from TIRF’s National Fatality Database used for the corre-
lation analyses in this research have been collected with nancial sup-
port from the Canadian Council of Motor Transport Administrators
(CCMTA), Transport Canada, the Public Health Agency of Canada
(PHAC) and Desjardins. The survey data from TIRF’s RSM used for the
analyses in this research have been collected with nancial support from
Desjardins, Beer Canada and Labatt. The survey data from TIRF USA’s
RSM have been collected with nancial support from the Anheuser-
Busch Foundation. The analyses of the survey data in this research
have been made possible with funding from Desjardins, Labatt and
Anheuser-Busch Foundation.
References
Blows, S., Ameratunga, S., Ivers, R.Q., Lo, S.K., Norton, R., 2005. Risky driving habits
and motor vehicle driver injury. Accid. Anal. Prev. 37 (4), 619–624.
Brown, S., Vanlaar, W.G.M., Robertson, R.D., 2020. Marijuana Use Among Drivers in
Canada, 2000–2017: TIRF’S National Fatality Database. Trafc Injury Research
Foundation, Ottawa, ON.
Carter, D., 2020. Effects of COVID-19 Shutdown on Crashes and Travel in NC. North
Carolina, Department of Transportation. Transportation Research Board (TRB)
Webinar (2020).
Chan, A., April 2020. Saanich sees spike in speeders as roads empty due to COVID-19.
CTV News. https://vancouverisland.ctvnews.ca/saanichsees-spike-in-speeders-as-
roads-empty-due-tocovid-19-1.4886785.
City of Calgary. (April, 2020). Changing travel trends during COVID-19. Calgary:
Alberta.
City of Toronto. (April, 2020). City of Toronto Urges Drivers to Obey Rules of the Road.
Toronto: Ontario.
Fergusson, D., Swain-Campbell, N., Horwood, J., 2003. Risky driving behaviour in young
people: prevalence, personal characteristics and trafc accidents. Aust. N. Z. J.
Public Health 27 (3), 337–342.
Gupta, M., Pawar, N.M., Velaga, N.R., 2021. Impact of lockdown and change in mobility
patterns on road fatalities during COVID-19 pandemic. Transp. Lett. 1–14.
Halifax Regional Municipality (2020). Trafc Collisions (accessed May 28, 2021).
Harantov´
a, V., H´
ajnik, A., Kalaˇ
sov´
a, A., 2020. Comparison of the Flow Rate and Speed of
Vehicles on a Representative Road Section before and after the Implementation of
Measures in Connection with COVID-19. Sustainability 12 (17), 7216.
Heidenreich, P., April 2020. Recent uptick in speeding on Edmonton roads concerns
mayor and city ofcial. Global News. https://globalnews.ca/news/6739220/edmon
ton-drivers-speedingcoronavirus-covid-19-iveson/.
Hughes, J.E., Kafne, D., Kafne, L. Decline in trafc congestion increased accident
severity in the wake of COVID-19.
Iversen, Hilde, 2004. Risk-taking attitudes and risky driving behaviour. Transp. Res. Part
F: Trafc Psychol. Behav. 7 (3), 135–150.
Katrakazas, Christos, Michelaraki, Eva, Sekadakis, Marios, Yannis, George, 2020.
A descriptive analysis of the effect of the COVID-19 pandemic on driving behavior
and road safety. Transp. Res. Interdiscip. Perspect. 7, 100186. https://doi.org/
10.1016/j.trip.2020.100186.
Lyon, C., Vanlaar, W.G.M., Robertson, R.D., 2020. Road Safety Monitor 2019: Distracted
Driving Attitudes and Practices, 2004–2019. Trafc Injury Research Foundation,
Ottawa, ON.
McGreevy, P. (2020, April 22). Tickets for speeding in excess of 100 mph surge 87% amid
coronavirus shutdown, CHP says. Los Angeles Times. www.latimes.com/california/
story/2020-04-22/tickets-for-drivers-speeding-more-than-100-mph-surges-87-amid-
california-shutdown-chp-says.
National Center for Statistics and Analysis (2020). Overview of motor vehicle crashes in
2019. (Trafc Safety Facts Research Note. Report No. DOT HS 813 060). National
Highway Trafc Safety Administration.
Robertson, Robyn D., Mainegra Hing, Marisela, Pashley, Charlotte R., Brown, Steve W.,
Vanlaar, Ward G.M., 2017. Prevalence and trends of drugged driving in Canada.
Accid. Anal. Prev. 99, 236–241.
Saladi´
e, `
O., Bustamante, E., Guti´
errez, A., 2020. COVID-19 lockdown and reduction of
trafc accidents in Tarragona province, Spain. Transp. Res. Interdiscip. Perspect. 8,
100218.
Saskatoon Police Service (2020). Crime Stats Reports (accessed May 28, 2021).
Shilling, F., & Waetjen, D. (2020). Special Report (Update): Impact of COVID19
Mitigation on Numbers and Costs of California Trafc Crashes.
StataCorp, 2015. Stata Statistical Software: Release 14. StataCorp LP, College Station,
TX.
Stavrinos, Despina, McManus, Benjamin, Mrug, Sylvie, He, Harry, Gresham, Bria,
Albright, M. Grace, Svancara, Austin M., Whittington, Caroline, Underhill, Andrea,
White, David M., 2020. Adolescent driving behavior before and during restrictions
related to COVID-19. Accid. Anal. Prev. 144, 105686. https://doi.org/10.1016/j.
aap.2020.105686.
Sutherland, Mason, McKenney, Mark, Elkbuli, Adel, 2020. Vehicle related injury patterns
during the COVID-19 pandemic: What has changed? Am. J. Emerg. Med. 38 (9),
1710–1714.
Tempelaar, Dirk, Rienties, Bart, Nguyen, Quan, Kovanovic, Vitomir, 2020. Subjective
data, objective data and the role of bias in predictive modelling: Lessons from a
dispositional learning analytics application. PLoS ONE 15 (6), e0233977. https://
doi.org/10.1371/journal.pone.023397710.1371/journal.pone.0233977.
g00110.1371/journal.pone.0233977.g00210.1371/journal.pone.0233977.
g00310.1371/journal.pone.0233977.t00110.1371/journal.pone.0233977.
t00210.1371/journal.pone.0233977.t00310.1371/journal.pone.0233977.
t00410.1371/journal.pone.0233977.t00510.1371/journal.pone.0233977.
t00610.1371/journal.pone.0233977.t00710.1371/journal.pone.0233977.
t00810.1371/journal.pone.0233977.t00910.1371/journal.pone.0233977.
t01010.1371/journal.pone.0233977.t01110.1371/journal.pone.0233977.
t01210.1371/journal.pone.0233977.t01310.1371/journal.pone.0233977.
t01410.1371/journal.pone.0233977.t01510.1371/journal.pone.0233977.
t01610.1371/journal.pone.0233977.t01710.1371/journal.pone.0233977.
t01810.1371/journal.pone.0233977.t01910.1371/journal.pone.0233977.
s00110.1371/journal.pone.0233977.s00210.1371/journal.pone.0233977.
s00310.1371/journal.pone.0233977.s00410.1371/journal.pone.0233977.
r00110.1371/journal.pone.0233977.r00210.1371/journal.pone.0233977.
r00310.1371/journal.pone.0233977.r004.
Thomas, F. D., Berning, A., Darrah, J., Graham, L., Blomberg, R., Griggs, C., Crandall, M.,
Schulman, C., Kozar, R., Neavyn, M., Cunningham, K., Ehsani, J., Fell, J., Whitehill,
J., Babu, K., Lai, J., and Rayner, M. (2020, October). Drug and alcohol prevalence in
seriously and fatally injured road users before and during the COVID-19 public
W.G.M. Vanlaar et al.

Accident Analysis and Prevention 160 (2021) 106324
8
health emergency (Report No. DOT HS 813 018). National Highway Trafc Safety
Administration.
Vanlaar, Ward G.M., Barrett, Hannah, Hing, Marisela Mainegra, Brown, Steve W.,
Robertson, Robyn D., 2019. Canadian wildlife-vehicle collisions: an examination of
knowledge and behavior for collision prevention. J. Saf. Res. 68, 181–186.
Vanlaar, Ward, Robertson, Robyn, Marcoux, Kyla, Mayhew, Daniel, Brown, Steve,
Boase, Paul, 2012. Trends in alcohol-impaired driving in Canada. Accid. Anal. Prev.
48, 297–302.
Vingilis, E, et al., 2020. Coronavirus disease 2019: What could be the effects on Road
safety? Accident, Analysis and Prevention 144 (105687). https://doi.org/10.1016/j.
aap.2020.105687.
Virginia Department of Transportation. (2020, June 19). Volume down on Virginia’s
roadways, while unbelted and speed-related rashes and fatalities increase. www.
virginiadot.org/newsroom/statewide/2020/volume-down-on-virginia-roadways%
C2%A0while-unbelted-and-speed-related-crashes-and-fatalities-increase%C2%A06-
19-2020.asp.
Wagner, E., Atkins, R., Berning, A., Robbins, A., Watson, C., & Anderle, J. (2020,
October). Examination of the trafc safety environment during the second quarter of
2020: Special report (Report No. DOT HS 813 011). National Highway Trafc Safety
Administration.
Whaley, K. (2020, October 8) ’Zooming’ and driving: A new concern during the COVID-
19 pandemic. ABC News. https://abc13.com/distracted-driving-deaths-zoom-while-
take-a-work-meeting-research/6861205/.
Wickert, D. (2020, April 20). Trafc is light in Georgia, but more drivers are topping 100
mph. Atlanta Journal Constitution. www.ajc.com/news/state–regional-govt–
politics/trafc-light-georgia-but-more-drivers-are-topping-100-mph/
v2xAa1gWkciNt2QBeAe8HJ/.
Wisniewski, M. (2020, May 1). With fewer people on the roads, crashes are down – but
some drivers see lack of trafc as excuse to speed. Chicago Tribune. www.
chicagotribune.com/coronavirus/ct-coronavirus-speeding-up-crashes-down-
20200501-vgr3yinpibh3xkmc45dmvxycwa-story.html.
Woods-Fry, H., Vanlaar, W.G.M., Wicklund, C., & Robertson, R.D. (2020) Alcohol-
Impaired Driving & COVID-19 in the United States: Results from the 2020 TIRF USA
Road Safety Monitor. Trafc Injury Research Foundation USA, Inc.
World Health Organization (2021). WHO Coronavirus (COVID-19) Dashboard: COVID-
19 Explorer. Retrieved from https://covid19.who.int/table on June 3 2021.
W.G.M. Vanlaar et al.