Accident Analysis and Prevention 106 (2017) 181–190

Contents lists available at ScienceDirect

Accident Analysis and Prevention

journal homepage: www.elsevier.com/locate/aap

Evaluating the Safety In Numbers effect for pedestrians at urban T

intersections
⁎
Brendan Murphya, , David M. Levinsonb,c, Andrew Owena
a
University of Minnesota, Center for Transportation Studies, United States
b
University of Minnesota, Department of Civil, Environmental, and Geo-Engineering, United States
c
University of Sydney, School of Civil Engineering, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Assessment of collision risk between pedestrians and automobiles offers a powerful and informative tool in urban
Pedestrians planning applications, and can be leveraged to inform proper placement of improvements and treatment projects
Safety to improve pedestrian safety. Such assessment can be performed using existing datasets of crashes, pedestrian
Collisions counts, and automobile traffic flows to identify intersections or corridors characterized by elevated collision risks
Urban planning
to pedestrians. The Safety In Numbers phenomenon, which refers to the observable effect that pedestrian safety
is positively correlated with increased pedestrian traffic in a given area (i.e. that the individual per-pedestrian
risk of a collision decreases with additional pedestrians), is a readily observed phenomenon that has been studied
previously, though its directional causality is not yet known. A sample of 488 intersections in Minneapolis were
analyzed, and statistically-significant log-linear relationships between pedestrian traffic flows and the per-pe-
destrian crash risk were found, indicating the Safety In Numbers effect. Potential planning applications of this
analysis framework towards improving pedestrian safety in urban environments are discussed.

1. Introduction Minneapolis has a sufficiently rich dataset of pedestrian counts avail-

Assessment of collision risk between pedestrians and automobiles
offers a powerful and informative tool in urban planning regimes, and
can be leveraged to inform proper placement of improvements and
treatment projects to improve pedestrian safety. Existing, available
datasets of crashes, pedestrian counts, and automobile traffic flows can
be combined to identify intersections, corridors, or other urban areas
that exhibit elevated collision risks to pedestrians. As the availability of
count data gradually increases due to automation techniques in
counting and crash detection, the process of leveraging these data to
determine areas of cities in need of intermodal conflict mitigation will
become easier and more problem areas can be identified (Greene-
Roesel et al., 2008; Bu et al., 2007). Many cities and urban areas do not
have access to good-quality count data for traffic other than vehicles,
and must instead use estimation techniques to model active transport
flow levels, a technique commonly used in planning applications in
Europe and Asia, but not yet in the United States (Raford and Ragland,
2004). Example methodologies for modeling pedestrian and bicycle
non-motorized traffic levels are readily available (Liu and Griswold,
2009; Do et al., 2013; Raford and Ragland, 2005; Tabeshian and Kattan,
2014; Schneider et al., 2013; Hankey et al., 2012). However, the city of
able to allow for meaningful pedestrian safety analysis without im-
plementing complex traffic estimation models.
Safety levels associated with transportation in cities continue to be a
problem, with 1.24 million road users being killed in on-road accidents
in 2010 globally, and another 20–50 million injured (World Health
Organization, 2013). Further, a full 22% of traffic deaths worldwide are
pedestrians, which is quite a high figure considering the transportation
mode of walking harbors little danger unto itself; additionally, only 79
countries worldwide have implemented policies to physically separate
vulnerable road users (pedestrians, cyclists, and motorcyclists) from
autos (World Health Organization, 2013). Non-motorized transporta-
tion modes tend to be some degree of unsafe in most average developed
urban areas, except where specific programs and treatments have been
employed to address the safety concerns, such as in Copenhagen,
Denmark (Jensen, 2008).
This investigation aims to evaluate whether the Safety In Numbers
phenomenon is observable in originally collected pedestrian and crash
data in the Midwestern, U.S. city of Minneapolis, Minnesota. Safety In
Numbers (SIN) refers to the phenomenon that pedestrian safety is po-
sitively correlated with increased pedestrian traffic in a given area, e.g.
that the per-pedestrian risk of injurious interaction with motorized

⁎
Corresponding author.
E-mail addresses: murph677@umn.edu (B. Murphy), dlevinson@umn.edu (D.M. Levinson), aowen@umn.edu (A. Owen).

http://dx.doi.org/10.1016/j.aap.2017.06.004
Received 12 August 2015; Received in revised form 16 May 2017; Accepted 6 June 2017
Available online 15 June 2017
0001-4575/ © 2017 Elsevier Ltd. All rights reserved.
B. Murphy et al. Accident Analysis and Prevention 106 (2017) 181–190

vehicles decreases as a function of the increasing flow of pedestrian
traffic. SIN is well-supported by pedestrian crash data across a number
of studies in various urban environments and reviews (Jacobsen, 2003;
Leden, 2002; Bhatia and Wier, 2011). The concept has seen relatively
widespread adoption in urban planning schools of thought, though its
temporal causality is not clear-cut (Bhatia and Wier, 2011), and it is
commonly discussed only in the context of pedestrian risk depending on
pedestrian flow levels. The United States Department of Transportation
(USDOT) Strategic Plan for Fiscal Years 2012–2016 aims to reduce non-
vehicle-occupant fatalities to 0.15 per 100 million vehicle-miles-tra-
veled (VMT) by 2016. However, such a goal does not account for risk
dependence on pedestrian flow levels, and thus the federal guidelines
ignore the SIN effect.
By necessity, data informing placement of improvements and pro-
jects for walking and bicycling safety, such as pedestrian bump-outs
(Wolshon and Wahl, 1999) and traffic-calming measures (Campbell
et al., 2004), must be sufficiently granular; travel behavior studies are
typically performed at the Transportation Analysis Zone (TAZ) level,
which is insufficiently fine-grained to allow for analysis of the shorter-
distance travel modes of bicycling and walking. Schneider et al. (2004)
provides a relevant framework for building a comprehensive pedestrian
risk assessment model, with a granular focus on a specific university
campus and a model which included factors of pedestrian flow, vehicle
flow, and an environmental factor (crosswalk length). Wier et al. (2009)
provide precedent for area-level modeling of pedestrian risk in-
corporating zoning and land use characteristics. These levels of detail
correlate well with the realities of implementation of pedestrian safety
investments, which occur not on the city-wide level, but within specific
intersections and road segments.
Pedestrian traffic counts, Average Annual Daily Traffic (AADT), and
crash data from the city of Minneapolis are used to build a model of
crash frequencies at the intersection level as a function of modal traffic
inputs. This model determines whether the SIN effect is observable
within the available datasets for both pedestrians and cars, as well as
determine specific spatial locations within Minneapolis where pedes-
trians experience elevated levels of risk of automobile crashes, relative
to intersections elsewhere in the city. The ability to identify specific
unsafe locations based upon aggregated count and crash data offers an
additional tool for city planners to implement in multimodal and pe-
destrian-specific planning.
morning peak (7–9 AM), midday (11 AM–1 PM), and evening peak
(4–6 PM); these were summed together to produce a 6-h count, re-
presenting pedestrian traffic when a higher number of cars are on the
road. The counts took place sporadically across the 14-year window,
and most intersections were only counted once or twice due to the
rotating schedule on which counts occurred. An “annual average daily
6-h count” was obtained by averaging the intersection-counts over the
number of years for which that intersection was counted. AADT mea-
surements were associated with street links, not to intersections. To
associate AADT numbers with intersections, the AADT for each unique
street at an intersection (typically two) was summed together. Crashes
were tabulated to include both non-fatal and fatal crashes. Finally,
manual geocoding of the three datasets (TMC, AADT, and crash counts)
to the intersection spatial layer took place, to allow spatial analysis.
At a given intersection, “pedestrian risk” is defined as the number of
crashes, between cars and pedestrians, that occurred across the 14-year
period, divided by the 6-h count of pedestrians; this gives a metric to
assess the risk of being hit by a car that an individual pedestrian may
experience at such an intersection. “Car risk” is defined as the number
of crashes that occurred across the analysis period, divided by AADT,
which gives a metric to assess the risk of being involved in a car-pe-
destrian collision which a driver may experience. How these risks vary
from intersections with low to high traffic flows determines whether
pedestrians or cars experience SIN.
The sample of intersections to use within the analysis was created
by identifying intersections with both nonzero pedestrian counts and
nonzero AADT data; there were 448 such intersections, and a summary
of the sample data can be found in Table 2 in Section 3. A map of the
intersections included in the analysis is shown in Fig. 1. Of note, the 6-h
pedestrian counts comprised a full 84.29% of the 12-h count totals,
averaged across the 448 included intersections. Of the 448 in the
Table 1
Hypotheses for signs of variable exponents in per-pedestrian and per-car crash risk
models.
Model 6-h peds (bp) AADT (bc)
Single-variable, pedestrian risk − n/a
Single-variable, car risk n/a −
Multivariable, pedestrian risk − +
Multivariable, car risk + −

2. Methodology

The existence of the SIN effect was examined within collected data Table 2
for the city of Minneapolis, at the intersection level. This framework Dataset summary statistics.
was chosen over other possible areas of analysis, such as mid-block or a
Description Value
link-based framework, due to intersections being the predominant lo-
cation where pedestrians interact with cars. Turning movement counts Intersections with pedestrian counts & AADT 448
(TMCs) for the years 2000–2013 provided pedestrian count data at the Minimum 6-h pedestrians 1
Maximum 6-h pedestrians 14,793
intersection level; AADT measurements from 2000–2013 provided ve-
Average 6-h pedestrians (standard deviation) 832.96 (1843.72)
hicle traffic flow levels on street links; traffic crash records from 6-h peds average percent of 12-h count 82.49%
2000–2013 yielded crash data with location-specific metadata to allow Minimum AADT 252
geocoding; an ESRI shapefile of intersection geolocations in Maximum AADT 40,623
Minneapolis provided by the city allowed for spatial analysis and Average AADT (standard deviation) 8893.33 (5613.00)
Total crashes at sampled intersections 1192 (1180 injuries, 12
geocoding. Datasets pertaining to built environment variables, such as
deaths)
speed limits, number of lanes, and intersection geometry are not in- Minimum crashes at sampled intersections 0
cluded in this analysis. Maximum crashes at sampled intersections 27
A few steps of data processing occurred prior to building models of Average total crashes at sampled intersections 2.66 (3.86)
(standard deviation)
crash counts, and crashes per pedestrian. The TMC data identified in-
dependent pedestrians passing through an intersection, defined by their Note: Summary statistics for datasets used in pedestrian safety analysis: pedestrian
directional heading, across 12-h counts (6 AM to 6 PM); to account for turning movements between 2000 and 2013, AADT between 2000 and 2013, and ag-
this, pedestrian counts for each direction (e.g. northbound, eastbound) gregate crash reports between 2000 and 2013, for the city of Minneapolis.
were summed together to yield a total count. The time windows ex-
tracted from the count data were the three peak periods of the day:

182
B. Murphy et al. Accident Analysis and Prevention 106 (2017) 181–190

Fig. 1. Locations of intersections in Minneapolis with both

pedestrian counts and AADT data.

sample, 105 intersections had 6-h counts which constituted 100% of the Such a log-linear model form is commonly used within crash mod-
12-h count totals. eling frameworks involving vehicular flows; Abdel-Aty et al. (1998)
The model form used is as follows: give a log-linear framework for modeling crash frequency using de-
mographic and environmental variables, and Lee et al. (2003) give a
b
R = C *Qp p *Qcbc (1) model of crash exposure in terms of vehicular flows and environmental
variables.
where R is the risk factor (either crashes per pedestrian, or crashes per
Single-variable models including only pedestrian traffic or vehicle
car), Qp is 6-h pedestrian traffic flow, Qc is auto traffic flow (AADT), and
traffic, as well as a model with both traffic modes, are included in the
C is a constant. This model form allows for log-linear regression ana-
analysis. Table 1 outlines the hypotheses for the signs of the exponents
lysis, since
in both single-variable and multivariable models of pedestrian and car
log(R) = bp *log(Qp) + bc *log(Qc ) + log(C ) (2) risk factors.
In general, it is hypothesized that increased traffic of a mode has a

183
B. Murphy et al. Accident Analysis and Prevention 106 (2017) 181–190

Histogram of pedestrian counts negative effect on rate of crashes per vehicle or user of that mode (e.g.
increased pedestrians correlate with lower per-pedestrian crash risk);
300

and, that increased traffic of a mode has a positive effect on rate of

crashes per vehicle or user of the other mode (e.g. increased car traffic
250

correlates with higher per-pedestrian crash risk). Thus, the SIN effect is
predicted for both pedestrian and auto modes, and it is predicted that
200

safety for a mode decreases with an increase in traffic of the other

Frequency

modes.
150
100

3. Data analysis
50

Table 2 lists summary statistics for the datasets used in the safety
analysis: automobile-pedestrian crash totals between 2000 and 2013;
0

pedestrian turning movement counts (TMC) on an average day between

0 3000 6000 9000 12000 15000
2000 and 2013; and automobile AADT figures between 2000 and 2013.
Pedestrian counts
The ”average total crashes at sampled intersections” figure represents
Fig. 2. Histogram of pedestrian count data. the average number of crashes that occurred at an intersection across
the entire 14-year analysis period.
As mentioned in Section 2, a total of 448 intersections were iden-
tified as having both nonzero pedestrian counts and AADT data. The
Histogram of AADT average 6-h pedestrian count across these intersections was 832.96, and
the average AADT was 8893.33. The number of crashes counted at
individual intersections, across the 14-year window, ranged from 0 to
80

27, with an average of 2.66. A series of figures shows visual re-

presentations of the datasets included in the analysis. Figs. 2 to 4 show
60

histograms of the pedestrian counts, AADT data, and crash counts, re-
spectively. Figs. 5 to 7 show maps of the pedestrian counts, AADT data,
Frequency

and crash counts, respectively.

40

As shown in the histograms in Figs. 2 to 4, intersections trend to-

wards having low pedestrian counts, middling AADT, and few crashes.
Pedestrian counts and crash counts show decreasing exponential dis-
20

tributions, while AADT values show a more unimodular distribution.

Fig. 5 gives a spatial representation of pedestrian counts, showing sig-
nificant activity within the downtown core, as well as areas east of the
0

Mississippi River, corresponding to the University of Minnesota

0 10000 20000 30000 40000
campus. Fig. 6 shows the spatial AADT distribution across the city;
AADT AADT is more uniformly distributed across a broader area than the
Fig. 3. Histogram of AADT data. pedestrian activity. Finally, Fig. 7 shows the spatial distribution of crash
counts, with the most occurring along busy corridors to the south and
immediate northwest of the downtown core.
Figs. 8 and 9 show log–log scatter plots of the relationships between
crashes per pedestrian and pedestrian traffic, and crashes per car and
car traffic, respectively. These plots suggest a negative relationship
Histogram of crash counts between crashes per pedestrian and pedestrian traffic, and between
crashes per car and car traffic, and give a view of the dependent vari-
ables for the single-variable and multivariable log-linear regression
analysis. The parallel linear patterns in these plots correspond to iso-
200

clines of intersections with the same crash counts.

Fig. 10 shows a map of crash totals at intersections, divided by the
150

number of pedestrians counted across the three 2-h peak periods in a

Frequency

day. This view of the data gives spatial representation to intersections,

or clusters of intersections, characterized by different levels of risk as-
100

sociated with crossing the intersection. Focus area A corresponds to the

north-south corridor of Penn Avenue in Minneapolis; focus area B is the
50

University of Minnesota campus; focus area C is the downtown Central

Business District (CBD); focus area D constitutes the area surrounding
the east-west corridor of Lake Street and the north-south corridor of
0

Lyndale Avenue. All four of these focus areas show a change in pe-
0 5 10 15 20 25
destrian risk, from considering only raw crash counts (Fig. 7) to ac-
Crashes
counting for pedestrian counts (Fig. 10). Areas A and D show elevated
Fig. 4. Histogram of crash count data. pedestrian risk when accounting for pedestrian counts, and areas B and
C show lower levels of per-pedestrian crash risk.

184
B. Murphy et al.
4. Regression results
Table 3 shows the log-linear regression results for the single-vari-
able models, and Table 4 shows the log-linear regression results for the
multivariate models. For the single-variable log-linear model for SIN for
pedestrians, the coefficient of log(6-hpeds) was found to be −0.035 and
strongly significant. This is also the exponent bp in Eq. (1), indicating a
negative exponential relationship between pedestrian traffic levels and
the per-pedestrian risk of a crash. For the single-variable log-linear
model for SIN for automobiles, the coefficient of log(AADT) was found
to be −0.0003 and strongly significant. This is the exponent bc in Eq.
(1), suggesting a negative exponential relationship between auto traffic
185
Accident Analysis and Prevention 106 (2017) 181–190
Fig. 5. Average annual 6-h pedestrian counts. Both dot size
and color scale correlate with pedestrian count levels. (For
interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)
levels and the per-automobile risk of hitting a pedestrian. However, the
effect is two orders of magnitude less for automobiles than it is for
pedestrians. Additionally, an R2 of 0.213 was observed for the single-
variable pedestrian risk model, while an R2 of only 0.098 was observed
for the single-variable car risk model.
For the multivariable log-linear model describing per-pedestrian
crash risk, the exponent bp was again found to be negative (−0.036)
and strongly significant, indicating a negative relationship between
increasing pedestrian traffic and per-pedestrian risk of a crash. The
exponent bc was found to be positive and weakly significant, indicating
that increased automobile traffic has a positive relationship with the
per-pedestrian risk of a crash at intersections.
B. Murphy et al.
Similar relationships were found in the multivariable log-linear
model describing per-car crash risk. Exponent bp was positive and
strongly significant, indicating a positive relationship between pedes-
trian traffic and per-car risk of hitting a pedestrian; exponent bc was
negative and strongly significant, indicating a negative relationship
between auto traffic and per-car risk of hitting a pedestrian. Again, the
model for per-pedestrian crash risk showed a higher R2 (0.219) than
that of the per-car crash risk model (0.117).
In both single-variable and multivariable models, the SIN effect
appeared to be stronger for pedestrians than for cars, indicated by the
coefficient for log(6-hpeds) in the crashes-per-pedestrian models being
larger in magnitude than the coefficient for log(AADT) in the crashes-
186
Accident Analysis and Prevention 106 (2017) 181–190
Fig. 6. Average annual daily car traffic (AADT). Both dot size
and color scale correlate with AADT levels. (For interpreta-
tion of the references to color in this figure legend, the reader
is referred to the web version of this article.)
per-car models. This is visible in Fig. 8, in which the relationship for
pedestrians appears to be more coherent than that shown in Fig. 9 for
cars.
5. Discussion
The SIN effect was observed in the analysis of the available pedes-
trian, auto, and crash data at 448 sampled intersections in Minneapolis,
in using a log-linear model form. The log-linear form was used in lieu of
negative binomial estimators, since the per-user crash rates appeared to
follow negative exponential decay, and were also not integer count
values. At intersections characterized by higher levels of pedestrian
B. Murphy et al.
traffic, lower per-pedestrian rates of crashes involving automobiles and
pedestrians were observed; at intersections characterized by higher
levels of auto traffic, lower per-car rates of crashes involving auto-
mobiles and pedestrians were observed. The SIN effect for cars was
found to be a few orders of magnitude weaker than the SIN effect for
pedestrians; SIN pertaining to pedestrian safety is a well-documented
phenomenon (see Jacobsen, 2003; Leden, 2002; Bhatia and Wier,
2011). The precise reasons behind this effect are not definitively
187
Accident Analysis and Prevention 106 (2017) 181–190
Fig. 7. Crash counts at sample intersections across the ana-
lysis time window of 2000–2013. Both dot size and color
scale correlate with crash count levels. (For interpretation of
the references to color in this figure legend, the reader is
referred to the web version of this article.)
known; however, the aforementioned studies have hypothesized psy-
chological effects on drivers, in that when driving in environments
characterized by greater average levels of pedestrians, drivers may tend
to act with more caution. And while these various relationships were
statistically observed within the traffic data, it is important to note that
the causal directionality of the SIN effect cannot be inferred directly.
The per-pedestrian crash rate was found to increase with increasing
automobile traffic, and the per-car crash rate was found to increase
B. Murphy et al.
Crashes per pedestrian vs. average 6−hour pedestrians
Crashes (2000−2013) per pedestrian
5e−01
5e−02
5e−03
5e−04
1 10 100
Pedestrians (6−hour average)
Fig. 8. Scatter plot of 14-year crashes (2000–2013) per pedestrian (6-h average daily) vs.
6-h pedestrian counts, log–log scale.
Crashes per car vs. AADT
5e−03
Crashes (2000−2013) per car
1e−03
2e−04
5e−05
500 1000 2000 5000
AADT (2000−2013)
Fig. 9. Scatter plot of 14-year crashes per car (AADT) vs. AADT, log–log scale.
with increasing pedestrian traffic, as shown in Table 4. These effects,
along with the SIN phenomenon observed for both pedestrians and cars,
were consistent with the hypotheses outlined in Table 1. Holding pe-
destrians constant and increasing car traffic increases the risk of a pe-
destrian being hit by a car by simple probability of interaction, and the
same holds true for increasing pedestrian traffic for a given AADT
value. However, the SIN effect was stronger and more coherent for
pedestrians than for cars, indicated by their disparate R2 values. This
suggests that there may be more factors relevant to the number of auto-
pedestrian crashes per vehicle than just the intersection's AADT, such as
intersection geometry, speed limit, number of lanes, or other environ-
mental variables. These secondary factors were not included in this
analysis, due to the Safety In Numbers effect strictly relating to the
modal traffic levels and number of crashes at spatial locations. Inter-
sections with similar traffic patterns may indeed be disparate in their
relative safety levels due to built environment characteristics, and basic
SIN analysis can assist in identifying these locations which warrant
more detailed monitoring and description.
A small amount of additional explanatory power was yielded when
considering opposite-mode traffic levels in a per-user crash count
model, e.g. including automobile traffic in the model for per-pedestrian
crashes yielded an increase of 1.39% in R2, and including pedestrian
traffic in the model for per-car crashes yielded an increase of 19.39% in
R2 (see Tables 3 and 4). However, the low overall R2 values indicate
Accident Analysis and Prevention 106 (2017) 181–190
that there is a ceiling of explanatory power available to a model which
only includes metrics of modal traffic volumes. This limitation is pur-
posefully imposed by focusing the models on only the variables im-
mediately relevant to calculating per-user safety factors – traffic vo-
lumes and crash counts.
An ongoing challenge with pedestrian safety analysis dependent on
count and crash data is the issue of data quality and availability. Data
practices vary from city to city and state to state, with implications to
investigations intending to aggregate safety data for cross-jurisdiction
comparison. Additionally, a large amount of city data collection per-
taining to street utilization is still performed manually, and such pro-
cesses are error-prone and inconsistent between jurisdictions. This
study used an available, fairly robust dataset provided by the City of
Minneapolis, covering pedestrian counts, AADT, and crash data for 14
years from 2000 to 2013. Some cities, such as Boston, do not have
1000 10000
robust pedestrian and bicycle counting programs throughout the city;
others, such as Philadelphia, may have varying data release and non-
disclosure agreements between Metropolitan Planning Organizations
(MPOs), cities, and police departments; still other cities may have in-
consistent data tracking and release practices. The collection and pro-
cessing of pedestrian and bicycle spatial safety data on an aggregate
scale becomes exceedingly difficult. Better standards of practice in data
collection, management, and distribution are needed.
Additionally, count data tend to vary widely due to heterogeneity of
density, destination availability, and other factors in urban areas. The
lowest pedestrian count in the sample of intersections included in the
analysis was 1 over the 6-h count period. The Twin Cities of
Minneapolis and St. Paul are not quite among the most walkable cities
in North America, and rank 16 and 28 in terms of Walk Score, re-
spectively (Score, 2016). There will be some areas and intersections
within the city where people rarely walk, relative to areas with more
pedestrians. Due to these heterogeneities of urban density in popula-
tion, jobs, and other economic or life activities, it would be expected
that some intersections have very few pedestrians. The effect of this
10000
reality on the model would lead to decreasing the model's ability to
accurately predict large numbers of pedestrians, due to the considerable
count variance.
This analysis took place over a 14-year time period, during which
there may have been changes to the on-street pedestrian network facil-
ities which influenced both pedestrian and auto traffic, as well as crash
frequency. For instance, construction on the Metro Transit Green Line
light rail system, which runs through the University of Minnesota campus
(area B in Fig. 10), began in 2010. Auto and pedestrian traffic patterns in
the area were greatly altered, with part of the area being permanently
closed to car traffic, and many pedestrian improvements being im-
plemented with the light rail line. This study does not account for area-
specific network changes, as the focus is city-wide aggregate pedestrian
safety; more detailed investigation into network changes in specific areas,
and resulting traffic and accident patterns, would be needed.
Visualizing unsafe intersections, or groups of intersections, within
an urban area is an important angle of analysis to undertake with the
types of datasets used in this investigation. Problematic areas within the
city environment become readily apparent; when multiple intersections
with relatively high pedestrian injury risk-burden lie in the same cor-
ridor, such as Lake Street in Minneapolis (see area D in Fig. 10), a
discussion of pedestrian safety and the surrounding built environment
should occur. The entire Lake Street corridor stands out as an area with
elevated pedestrian risk burdens, given the number of pedestrians
walking there, compared to the relatively walk-friendly downtown
district (area C) and the University of Minnesota campus (area B). Si-
milarly, the Penn Avenue corridor in North Minneapolis (area A) shows
a series of intersections with elevated pedestrian risk.
Most of the corridors displaying elevated pedestrian risks in Fig. 10
may be classified as urban arterials, with more than one lane of traffic
188
B. Murphy et al.
in each direction. That such roadways would be unsafe for pedestrians
may seem obvious, but a visualization tool can be more powerful in
informing planners the scope and extent of dangerous streets within an
urban area than simple crash counts alone.
6. Conclusion
Through the pedestrian risk-burden analysis framework outlined in
this study, it is possible to observe the Safety In Numbers effect at the
individual intersection level, for both pedestrians and cars, as well as
identify intersections with a disproportionately high rate of crashes for
189
Accident Analysis and Prevention 106 (2017) 181–190
Fig. 10. Total crashes (2000–2013) per pedestrian (6-h daily
average) at intersections in Minneapolis. Letter labels in-
dicate focus areas for discussion.
its level of pedestrian activity. Pedestrians were found to be at a lower
risk of being hit by a car at intersections with higher pedestrian traffic,
and individual cars were found to be at a lower risk of hitting pedes-
trians at intersections with more car traffic. The causality of the SIN
effect is not understood, and more research should be conducted to
understand its causes, but it is still a justification for improving the
walkability and pedestrian safety of urban environments. Assessing the
per-pedestrian crash rates at spatial locations, as opposed to crash
counts alone, allows practitioners and planners to more readily identify
target areas where improvements to pedestrian infrastructure may be
warranted.
B. Murphy et al. Accident Analysis and Prevention 106 (2017) 181–190

Table 3 on traffic accident involvement using log-linear models. Accid. Anal. Prev. 30 (6),
Single-variable log-linear regression results. 851–861. http://dx.doi.org/10.1016/S0001-4575(98)00038-4.
Bhatia, R., Wier, M., 2011. “Safety in Numbers” re-examined: can we make valid or
practical inferences from available evidence? Accid. Anal. Prev. 43 (1), 235–240.
log(crashes/ped) log(crashes/car)
Bu, F., Green-Roesel, R., Diogenes, M.C., Ragland, D.R., 2007. Estimating Pedestrian
Accident Exposure: Automated Pedestrian Counting Devices Report. Tech. Rep.
log(6-hpeds) −0.035***
University of California, Berkeley. https://escholarship.org/uc/item/0p27154n.
(0.003) Campbell, B.J., Zegeer, C.V., Huang, H.H., Cynecki, M.J., 2004. A Review of Pedestrian
log(AADT) −0.0003*** Safety Research in The United States and Abroad. Tech. Rep. University of North
(0.00004) Carolina, Chapel Hill.
Constant 0.223*** 0.003*** Do, M.T., Grembek, O., Ragland, D., Chan, C.-Y., 2013. Weighting integration by block
(0.019) (0.0004) heterogeneity to evaluate pedestrian activity. TRB 92nd Annual Meeting. 14 pp.,
schémas, tabl., ill., bibliogr.
Observations 448 448
Greene-Roesel, R., Diogenes, M.C., Ragland, D.R., Lindau, L.A., 2008. Effectiveness of a
R2 0.213 0.098 commercially available automated pedestrian counting device in urban environ-
Adjusted R2 0.211 0.096 ments: comparison with manual counts. TRB 2008 Annual Meeting. pp. 1–16.
Residual std. error (df = 446) 0.109 0.001 Hankey, S., Lindsey, G., Wang, X., Borah, J., Hoff, K., Utecht, B., Xu, Z., 2012. Estimating
F statistic (df = 1; 446) 120.392*** 48.536*** use of non-motorized infrastructure: models of bicycle and pedestrian traffic in
Minneapolis, MN. Landsc. Urban Plan. 107 (3), 307–316. http://dx.doi.org/10.1016/
*** j.landurbplan.2012.06.005.
Note: (standard error); *p < 0.1; **p < 0.05; p < 0.01.
Jacobsen, P.L., 2003. Safety in numbers: more walkers and bicyclists, safer walking and
bicycling. Inj. Prev. 9 (3), 205–209.
Table 4 Jensen, S.U., 2008. Safety effects of blue cycle crossings: a before–after study. Accid.
Multivariable log-linear regression results. Anal. Prev. 40 (2), 742–750. http://dx.doi.org/10.1016/j.aap.2007.09.016.
Leden, L., 2002. Pedestrian risk decrease with pedestrian flow. A case study based on data
log(crashes/ped) log(crashes/car) from signalized intersections in Hamilton, Ontario. Accid. Anal. Prev. 34 (4),
457–464.
log(6-hpeds) −0.036 ***
0.0001*** Lee, C., Hellinga, B., Saccomanno, F., 2003. Real-time crash prediction model for the
(0.003) (0.00002) application to crash prevention in freeway traffic. Transportation Research Board
log(AADT) 0.014* −0.0003*** 82nd Annual Meeting. http://dx.doi.org/10.3141/1840-08.
(0.007) (0.00004) Liu, X., Griswold, J., 2009. Pedestrian volume modeling: a case study of San Francisco.
Constant 0.102 0.003*** Yearb. Assoc. Pacif. Coast Geogr. 71 (1), 164–181. http://dx.doi.org/10.1353/pcg.0.
0030.
(0.067) (0.0004)
Raford, N., Ragland, D., 2004. Space Syntax: An Innovative Pedestrian Volume Modeling
Observations 448 448 Tool for Pedestrian Safety. http://escholarship.org/uc/item/50m064zp.
R2 0.219 0.117 Raford, N., Ragland, D.R., 2005. Pedestrian volume modeling for traffic safety and ex-
Adjusted R2 0.215 0.113 posure analysis: the case of Boston, Massachusetts. Transportation Research Board
Residual std. error (df = 445) 0.109 0.001 85th Annual Meeting.
F statistic (df = 2; 445) 62.284*** 29.462*** Schneider, R.J., Ryznar, R.M., Khattak, A.J., 2004. An accident waiting to happen: a
spatial approach to proactive pedestrian planning. Accid. Anal. Prev. 36 (2),
*
*** 193–211. http://dx.doi.org/10.1016/S0001-4575(02)00149-5.
Note: (standard error). p < 0.1; **p < 0.05; p < 0.01. Schneider, R.J., Henry, T., Mitman, M.F., Stonehill, L., Koehler, J., 2013. Development
and Application of the San Francisco Pedestrian Intersection Volume Model. Safe
All associated datasets and appropriate metadata used within this Transportation Research & Education Center (2).
Score, W., 2016. Walk Score. https://www.walkscore.com/cities-and-neighborhoods/.
study have been uploaded to the Data Repository for the University of Tabeshian, M., Kattan, L., 2014. Modeling nonmotorized travel demand at intersections
Minnesota (DRUM). based on traffic counts and GIS data in Calgary, Canada. Transp. Res. Rec.: J. Transp.
Res. Board. http://dx.doi.org/10.3141/2430-05. 17 pp.
Wier, M., Weintraub, J., Humphreys, E.H., Seto, E., Bhatia, R., 2009. An area-level model
Acknowledgements
of vehicle-pedestrian injury collisions with implications for land use and transpor-
tation planning. Accid. Anal. Prev. 41 (1), 137–145. http://dx.doi.org/10.1016/j.
We thank the University of Minnesota Center for Transportation aap.2008.10.001. http://www.sciencedirect.com/science/article/pii/
S0001457508001930.
Studies, and Roadway Safety Institute, for logistical and financial sup-
Wolshon, B., Wahl, J., 1999. Novi's main street: neotraditional neighborhood planning
port of this research. and design. J. Urban Plan. Dev. 125 (1), 2–16. http://dx.doi.org/10.1061/(ASCE)
0733-9488(1999)125:1(2.
References World Health Organization, 2013. Global status report on road safety. Inj. Prev. 318. ,
http://injuryprevention.bmj.com/content/15/4/286.short. http://www.who.int/
violence_injury_prevention/road_safety_status/2013/en/index.html.
Abdel-Aty, M.A., Chen, C.L., Schott, J.R., 1998. An assessment of the effect of driver age

190