Content uploaded by John Luciano Renne
Author content
All content in this area was uploaded by John Luciano Renne on Mar 24, 2015
Content may be subject to copyright.
29
Transportation Research Record: Journal of the Transportation Research Board,
No. 2464, Transportation Research Board of the National Academies, Washington,
D.C., 2014, pp. 29–37.
DOI: 10.3141/2464-04
This study provides a framework for a comprehensive, low-cost pedes-
trian safety analysis incorporating multiple data sources and analysis tech-
niques. Critically, the framework is flexible enough to provide meaningful
results and inform recommended safety interventions even if there are
gaps in data availability or completeness. The methodology includes an
evaluation of available crash records, an audit of current pedestrian facili-
ties, collection of pedestrian count data, and an assessment of relevant
contextual factors. Together, these elements provide a holistic view of
pedestrian safety and comfort, informing needed interventions. This
methodology was used to evaluate three pedestrian crash clusters in the
New Orleans, Louisiana, area and revealed serious deficiencies in the
pedestrian environment. The methodology uncovered critical systemic
data gaps that needed to be addressed to draw clearer relationships
between infrastructure and crash patterns. The methodology was also
found to be an effective tool for comparing and prioritizing proposed
investments and a means to demonstrate clearly to policy makers the need
for continued attention to the improvement of pedestrian safety in the
context of Complete Streets policy implementation.
Pedestrian crashes and fatalities are a persistent problem. Nationally,
pedestrians represent 12% of all traffic fatalities (about 4,000 deaths
per year), with an additional 59,000 pedestrian injuries occurring
annually (1). To address this problem, tools such as the Pedestrian
Bicycle Crash Analysis Tool have been created to help provide
solid evidence about crash types and potential countermeasures
that transportation policy makers can utilize to improve pedestrian
safety. Use of these tools, however, can be data intensive and expen-
sive. Financial and institutional costs can act as a significant barrier,
particularly for smaller organizations with limited resources. When
crash analyses are completed, they may not necessarily be linked
to broader planning or policy efforts, which limits their potential
for positive impact. This paper proposes a new framework piloted
in New Orleans, Louisiana, for an adaptable method of evaluating
pedestrian safety and the built environment that can facilitate
interdisciplinary dialogue among stakeholders and advance policy
initiatives. The framework is called Complete Streets.
After years of exceeding national average crash and fatality rates,
New Orleans was designated as an FHWA Pedestrian Safety Focus
City in 2011 (1); this designations resulted in a renewed focus on
addressing the issue of pedestrian safety at the state, regional, and
municipal levels. This research emerged from the need to understand
better the complex factors contributing to pedestrian crashes and to
provide a framework to identify, prioritize, and evaluate problem
areas. Municipalities and metropolitan planning organizations need
efficient, low-cost methods of understanding pedestrian safety issues
and implementing effective design, enforcement, and educational
countermeasures. Analyzing crash data is a common activity toward
that goal. However, such data are often limited, delayed, or incom-
plete, limiting the agency’s ability to evaluate crash circumstances
effectively and identify clear counter measures that could prevent
future crashes. Therefore, agencies must incorporate data from other
sources to understand holistically where, how, and why crashes are
occurring in their jurisdiction. An efficient, easy to use, and adaptable
methodology for extracting as much information as possible about
crash circumstances and active transportation behaviors in a given
area is needed to supplement (and in some cases compensate for)
imperfect data.
Since 2006, the Pedestrian Bicycle Resource Initiative (PBRI, a
joint project of the Merritt C. Becker, Jr. Transportation Institute at the
University of New Orleans and the New Orleans Regional Planning
Commission) has been engaged in the development of tools for evalu-
ating crash outcomes and improving safety for nonmotorized road
users. The initiative has included the following: evaluations of pedes-
trian and bicycle crash data, development of a process for identify-
ing statistically significant crash clusters, creation of a comprehensive
toolkit for auditing street segments and intersections for pedestrian
safety, and implementation of an annual citywide pedestrian and
bicycle count program. The purpose of this study was to synthesize
all of these activities into a flexible, replicable framework for con-
ducting a comprehensive pedestrian safety analysis, exploring crash
activity in a given area, and facilitating development of recommended
design, enforcement, and educational countermeasures to reduce crash
prevalence. The method could also be used to test the effectiveness of
design and engineering interventions in an ex ante and ex post fashion;
however, effectiveness tests were not done in this study.
This paper describes the methods employed for each component of
the proposed safety analysis framework and summarizes the findings
from the application of the methodology to three pedestrian crash
Development of Low-Cost Methodology
for Evaluating Pedestrian Safety
in Support of Complete Streets
Policy Implementation
Tara Tolford, John Renne, and Billy Fields
T. Tolford and J. Renne, University of New Orleans Transportation Institute,
University of New Orleans, 368 Milneburg Hall, 2000 Lakeshore Drive,
New Orleans, LA 70148. Alternate affiliation for J. Renne: Transport Studies
Unit, University of Oxford, United Kingdom. B. Fields, Department of Political
Science, Texas State University, 601 University Drive, Undergraduate Academic
Center 355, San Marcos, TX 78666-4684. Corresponding author: T. Tolford,
ttolford@uno.edu.
30 Transportation Research Record 2464
clusters in the New Orleans area. The applications included not only
the specific infrastructure deficiencies, user demands, and behavioral
concerns that appear to contribute heavily to crash incidence, but also
the broader systemic issues of pedestrian accommodation and the
collection and dissemination of data.
Finally, the study investigated the implications of the use of the
methodology as part of an evaluation framework for the implemen-
tation of Complete Streets. Complete Streets policies have recently
been adopted by Louisiana, New Orleans, and metropolitan plan-
ning organizations in New Orleans (2). However, adopting a Com-
plete Streets policy is the easy part, while implementing the policy
is a challenge (3). Developing low-cost and reliable methodologies
for evaluating the pedestrian realm is a key to providing communi-
ties with the necessary data to link pedestrian safety with the built
environment to create pedestrian-friendly neighborhoods.
LITERATURE REVIEW
The literature on each of the elements of the safety analysis meth-
odology (crash record analysis, count data collection, built environ-
ment evaluation, and contextual factors contributing to crashes) is
extensive. A variety of models and methods for evaluating pedes-
trian traffic conflicts and predicting collisions have been published
(4–6). Similarly, considerable research exists on the evaluation of
the built environment for pedestrian safety and comfort, in particu-
lar on the relationship between infrastructure improvements and
pedestrian behaviors (7–9). Although the efficacy of enforcement
programs aimed at improving pedestrian behavior has been dem-
onstrated (8, 9), researchers have found that in many cases design
interventions are an essential prerequisite to changing unsafe user
behaviors (7, 8). Because specific countermeasures that improve
pedestrian safety outcomes are well outlined in the literature, they
were discussed minimally in this research, although the finding that
infrastructure improvement should be prioritized was an underlying
assumption of the study.
Recent research has begun to develop models and new methodol-
ogies for estimating pedestrian volumes, another important facet of
understanding pedestrian safety and risk (10–12). Other researchers
have evaluated the impacts of crime and socioeconomic status on
pedestrian travel patterns and safety outcomes (13–15), opportuni-
ties for monitoring pedestrian volumes and behavior with new tech-
nologies (7), possible environmental justice implications stemming
from the link between crash rates and infrastructure deficiencies
(15), and the complex relationships between crashes and various
aspects of urban form (16, 17). Most of these methods have tended
to be highly data intensive or require specialized analytic skills not
often available to local communities.
Meanwhile, the institutional barriers, capacity limitations, and
data gaps that commonly plague pedestrian safety initiatives have
been well documented (5, 6, 13). Many of the data analysis tech-
niques explored in the literature, although scientifically robust, are not
feasible for local agencies to replicate. And collecting or accessing
complete, timely, and accurate data with precise geospatial informa-
tion and adequate contextual information is a challenge for many
regions (4, 5). As Miller identified, there needs to be a stronger con-
nection between the fields of planning and transportation safety in
support of pedestrian safety enhancement (6). In part, Miller sug-
gested, this can be achieved by developing safety analysis tools that
facilitate the integration of the technical skills represented in the
literature with other planning processes and encouraging dialogue
and collaboration around such analyses among planners, safety
professionals, and engineers. The safety analysis tools include, for
example, analyzing crash records, geo-locating crashes and defin-
ing crash clusters, estimating nonmotorized travel demand, iden-
tifying level of service or infrastructure deficiencies, identifying
countermeasures, and evaluating change.
The literature has provided relatively little guidance on how to
synthesize the analysis of crash data, evaluate the built environment
or facilities, and estimate user volume and contextual factors in a
comprehensive manner, given limited data and resources. Although
many of the advanced techniques and models outlined in the literature
can and should be incorporated into safety evaluations if resources
permit, practitioners also need to be able to generate meaningful data
efficiently for immediate planning and policy application.
Although working with qualitative or incomplete data introduces
a greater possibility for error, working with the available data also
makes it feasible to conduct analyses that may otherwise not be pos-
sible (6, 18, 19). Miller’s work recognized the challenges of work-
ing with incomplete data and incorporated additional data sources
(beyond crash records) in an effort to understand pedestrian safety
concerns; the work also addressed the opportunities for such analy-
ses to support regional policy initiatives (6). This research built on
previous literature and embraced the idea that a safety evaluation
can be a powerful tool to generate momentum for change and
to advance policy implementation. The study expanded the overall
scope of analysis by incorporating methods that address additional
facets of pedestrian safety and comfort and suggesting applications
for advancing Complete Streets policy initiatives.
METHODS AND DATA
The study addressed the needs of local and regional agencies to be
able to evaluate and prioritize areas with high pedestrian crash inci-
dence, as well as to explore in-depth the factors and circumstances
associated with those crashes in one or more specific corridors,
intersections, or nodes. The study method utilized geocoded crash
data to pinpoint clusters of crashes. The clusters were then evaluated
with a cost-effective pedestrian audit tool. The following tools and
methodology were used during each step of the analysis:
1. Identification and analysis of crash clusters,
2. Pedestrian sidewalk and intersection audits,
3. Pedestrian counts,
4. Area context,
5. Profile of fatal and severe crashes, and
6. Recommended interventions.
The analysis produced a concise summary of identified short-
comings in the pedestrian environment, estimated user demand,
and suggested countermeasures to improve safety for a given area.
The methodology can be used as a tool to advocate for change,
while providing benchmark metrics against which to evaluate future
progress. In addition, use of additional analysis tools beyond simply
evaluating crash statistics allows the identification of areas where
environmental conditions may discourage walking, resulting in few
crashes but decreased pedestrian connectivity.
Identification and Analysis of Crash Clusters
Crash data for incidents involving pedestrians between 2006 and
2010 in Orleans and Jefferson Parishes were obtained from the
Tolford, Renne, and Fields 31
Louisiana Department of Transportation and Development and geo-
coded with data fields for the primary street on which the crash
occurred and the intersecting street closest to the crash. Although
use of GPS technology to geo-locate crashes on site has become
increasingly prevalent in this region, GPS coordinates were not
available for the majority of crash records in the data set. Therefore,
crashes could only be coded to the nearest intersection, regardless of
where the crash actually occurred. The lack of specific crash location
data was an important factor in the evaluation of the built environment
factors in the crash vicinity.
Through previous research, relatively stable spatial pedestrian crash
patterns in the New Orleans region were identified, indicating possible
systemic deficiencies in the built environment that have created haz-
ardous conditions for users (20–23). Such deficiencies were unlikely
to be resolved without intervention, whether through increased edu-
cation about safe travel behavior, increased enforcement of existing
laws (or development of new laws), or design modifications. Many
of the same corridors, intersections, and crash clusters have been
identified year after year, marking areas that are particularly likely
to be in need of specific safety countermeasures.
Crash clusters were identified with the Spatial and Temporal
Analysis of Crime tool from CrimeStat (24), according to a method-
ology described and applied in the 2005 New Orleans Metropolitan
Bicycle and Pedestrian Plan (20). This tool facilitated the identifica-
tion of areas where a statistically significant number of crashes have
occurred (although the analysis could be completed for any corri-
dor or node of interest, without formal crash cluster identification).
Three of the crash clusters identified with this technique (Figure 1)
were selected by the New Orleans Regional Planning Commission at
this stage of the research for further analysis, based on the presence
of recent or planned infrastructure investment in the vicinity, overall
crash severity (each of the three clusters included one or more pedes-
trian fatalities), and potential value for concurrent planning efforts or
planned real estate development projects.
Next, crash records from the Department of Transportation and
Development crash database were retrieved for all pedestrian crashes
occurring within a quarter mile of each of the crash clusters. The
attributes evaluated included the total number of crashes per year;
crash severity; month, day of week, and hour of crashes; lighting,
road, and weather conditions; pedestrian age, sex, and race; primary
and secondary contributing factors; manner of collision; involve-
ment of drugs or alcohol; distracted or aggressive driving viola-
tions; hit and run status; pedestrian actions (e.g., playing in roadway,
crossing at intersection); and pedestrian condition (e.g., impaired,
distracted). In theory, these attributes would provide ample data to
evaluate the general circumstances and driver or pedestrian actions
that led to crash occurrence. However, for many of the attributes
evaluated, the records were incomplete for a substantial portion of
the crashes (up to 80% null values for some data fields). Therefore,
it was difficult to evaluate factors associated with crashes on the
FIGURE 1 Pedestrian crash clusters, Orleans and Jefferson Parishes (east bank core), 2006–2010 (expy 5 expressway;
hwy 5 highway). (Source: Data are from New Orleans Regional Planning Commission and the Louisiana Department of
Transportation and Development.)
32 Transportation Research Record 2464
basis of these records alone; thus it was not feasible to use multivariate
statistical analysis because of missing data. In addition, the codes
representing fields for primary and secondary contributing factors, as
well as manner of collision, lacked specificity regarding recognized
common crash types as identified by FHWA (25).
Some temporal, behavioral, and demographic trends could be
identified from the data. Notably, a high incidence of alcohol involve-
ment was noticed at two of the three crash clusters and pedestrian
actions were identified as a contributing factor for a large percent-
age of the crashes. Lighting, weather, and road conditions appear to
have had minimal involvement in the crashes, but, as observed by
Miller (6), where limited data exist, analysis of crash records that
rule out possible factors can also be helpful in narrowing the menu
of appropriate countermeasures. Although general patterns could be
identified, the data set was clearly inadequate for fully understanding
pedestrian safety concerns. Additional analysis tools were needed to
overcome the poor nature of the recordkeeping. Thus, improving the
quality and specificity of crash data to produce more precise descrip-
tions of crashes pertaining to nonmotorized users would significantly
enhance future data analysis efforts in this region.
Pedestrian Sidewalk and Intersection Audits
Based on the locations of crashes in each cluster, a set of street
segments and intersections was selected for a pedestrian field
audit. The instruments were developed by the PBRI for Auditing
Neighborhoods, Streets, and Intersections for Pedestrian Safety:
A Toolkit for Communities (26). The toolkit, which provides sur-
vey instruments to assess the quality and comfort of sidewalk and
pedestrian area segments on minor and major streets, signalized
intersections, and nonsignalized intersections, was intended to be
straightforward enough for use by community groups. The toolkit is
also detailed and robust enough to meet the needs of governmental
agencies.
Unlike many other pedestrian audit tools (e.g., FHWA’s), no data
are required that cannot be easily obtained by the auditor in the
field and no complex calculations are necessary. The audit evaluates
pedestrian attractors (including land use attractors as well as infra-
structure quality and availability) and detractors (including infra-
structure deficiencies, maintenance issues, and other factors known
to inhibit pedestrian access), based on best practices identified in
the transportation and public health fields (26). Possible audit scores
range from −4 (very poor) to +4 (very good) for each intersection or
one-block sidewalk segment. A value of zero or less indicates a fail-
ure to meet a basic benchmark for walkability; a score in the range
of 1 to 3 represents an area with pedestrian orientation but some
deficiencies. A score above 3 indicates a high-quality pedestrian
facility that is generally free from major defects or safety issues.
This tool is relatively quick to use and aggregates fine-grained quan-
titative and qualitative data into an overall score that can be easily
represented graphically, clearly illustrating overall, areawide defi-
ciencies as well as specific intersections or segments that represent
a critical gap in an otherwise pedestrian-friendly network.
Audits were conducted along major corridors in each crash clus-
ter, as well as along minor streets where crashes were known to have
occurred. Overall, the streets and intersections in the vicinity of the
pedestrian crash clusters scored poorly, largely because of lack of
dedicated, functional pedestrian signals or crosswalks; missing curb
ramps; sidewalk obstructions; and other deficiencies, all of which
were catalogued and photographed. A comparison of the audit scores
with crash locations and severity of crashes revealed mixed results.
Although some serious and fatal crashes occurred at or near inter-
sections that scored very poorly, other serious and fatal crashes
occurred at or near the highest-rated intersections, highlighting the
complexity of factors that contribute to crash occurrence. Because
of the small sample sizes of the crash clusters, statistical analysis of
any correlations between audit scores and crash incidence was not
conducted. The overall audit scores are shown in Table 1.
TABLE 1 Pedestrian Audit Scores at Target Crash Clusters
Number
of Audits
Conducted
Score (%)
Audit Component
Very Poor
(−4 to 0)
Poor
(>0 to 1)
Fair
(>1 to 2)
Good
(>2 to 3)
Very Good
(>3 to 4)
Median
Score
Crash Cluster 1
Signalized intersections 3 66.7 0.0 0.0 33.3 0.0 −0.50
Nonsignalized intersections 12 33.3 41.7 25.0 0.0 0.0 0.13
Sidewalk segments 32 12.5 37.5 25.0 15.6 9.4 0.50
Total 47 21.3 36.2 23.4 12.8 6.4 0.25
Crash Cluster 2
Signalized intersections 13 0.0 30.8 38.5 30.8 0.0 1.17
Nonsignalized intersections 8 12.5 25.0 37.5 25.0 0.0 2.00
Sidewalk segments 51 3.9 15.7 41.2 31.4 7.8 1.69
Total 72 4.2 19.4 40.3 30.6 5.6 1.67
Crash Cluster 3
Signalized intersections 3 66.7 0.0 33.3 0.0 0.0 −0.58
Nonsignalized intersections 4 75.0 0.0 25.0 0.0 0.0 0.00
Sidewalk segments 12 75.0 0.0 16.7 8.3 0.0 −0.25
Total 19 73.7 0.0 21.1 5.3 0.0 −0.25
Tolford, Renne, and Fields 33
Pedestrian Counts
Midblock screenline counts of the volume and characteristics of
pedestrians were conducted along two major streets in each crash
cluster. The data collection effort also included counting cyclists,
although this paper focused on the pedestrian aspects of the study. A
total of 8 h of count data was collected at each location over the course
of two weekdays, from 7:00 to 9:00 a.m. and from 4:00 to 6:00 p.m.,
in accordance with the recommendations of the National Bicycle and
Pedestrian Documentation Project (27). The methodology was devel-
oped to require a minimum amount of labor and training while
providing maximum data describing the demographic characteristics
and behaviors of pedestrians in a given area. In addition to record-
ing total user volumes during the count period, observers recorded
sex, race, and age of pedestrians, as well as travel orientation (i.e.,
sidewalk, street, or median).
Tot al v olu me s of o bs erv ed ped es tri an s we re e xtr ap ola te d to e sti-
mate daily, monthly, and yearly user volumes at each site. The adjust-
ment factors used to derive the estimates were from the National
Bicycle and Pedestrian Documentation Project (28). To perform the
extrapolation, all count observation periods for single sites were
separated into morning and afternoon counts. A mean count for each
time period was calculated. The averages were used to derive a daily
and weekly estimate for each time period, based on the time of day and
day of the week when the counts were observed. Additional adjust-
ment factors were calculated to derive weekly, monthly, and annual
estimates based on the time of year when the counts were collected
and depending on climate region.
As the counts were conducted in conjunction with a concurrent city-
wide count program, it was possible to compare counts in the crash
clusters with an average of the 26 count locations across the New
Orleans region. The comparison revealed that two of the three crash
clusters were in areas of unusually high pedestrian activity, while the
third, in a suburban area with minimal pedestrian infrastructure, had
little pedestrian traffic (see Table 2). However, the low user volumes
observed in the third area only served to reinforce the urgency of the
safety hazards identified, as the number of pedestrian crashes appeared
highly disproportionate to the number of pedestrians overall.
Area Context
Supplemental contextual information for the area surrounding each
crash cluster was also evaluated, including land use and zoning pat-
terns, urban design characteristics, transit access, and neighborhood
demographics, to develop a richer understanding of the crash-prone
areas in question. A mix of land uses, including residential, business,
and civic facilities, or urban site design characteristics (e.g., street
orientation, infrequent driveway cuts, and so forth) all pointed to an
environment that would draw pedestrians, even if the facilities pro-
vided were less than ideal. A Walk Score (as defined by the online
walkability model walkscore.com) for the centroid of each cluster
was noted, as an additional data point that has been found to be a
predictor of actual user demand (12). In addition, transit availability
has been shown to be a strong predictor of pedestrian presence (16).
Each of the three target crash clusters is an important transfer hub
between two or more transit lines and each demonstrated a diverse
mix of land uses, although the suburban site featured urban design
characteristics that were far less conducive to walking.
American Community Survey data were evaluated at the census
tract level to provide demographic context for the neighborhoods
surrounding each target crash cluster, including the percentage of the
population living below the poverty level, the percentage of work-
ers that commute to work by an active mode of transport, and the
percentage of households that lack access to a vehicle. For each of
the crash clusters, these details served to reinforce patterns observed
through manual counts and helped demonstrate (if applicable) the
clear and immediate need for pedestrian accommodation. Such data
could be more useful if they were collected across enough locations
to provide a statistically significant sample for quantitative analysis.
Profile of Fatal and Severe Crashes
The analysis revisited each of the crashes in each cluster that resulted
in fatal or severe injury and constructed a sketch of each incident.
The sketch served to highlight, where data existed, the specific char-
acteristics of each crash in relation to the qualitative data discussed
TABLE 2 Summary of Pedestrian Count Results at Target Crash Clusters
Cluster 1 Cluster 2 Cluster 3
Citywide Average
(Observed)Pedestrian Characteristic Count Site 1 Count Site 2 Count Site 1 Count Site 2 Count Site 1 Count Site 2
Pedestrians observed 468 492 692 485 45 68 317
Estimated daily traffic 1,731 1,652 2,490 1,620 163 209 928
Gender
Female (%) 36.8 34.8 24.1 34.0 44.4 30.9 40.6
Male (%) 63.3 65.2 75.9 66.0 55.6 69.1 59.4
Race
White (%) 29.7 13.4 3.8 32.8 11.1 27.9 58.4
Black (%) 65.7 79.7 94.8 61.0 86.7 70.6 36.1
Other (%) 4.7 6.9 1.5 6.2 1.1 1.5 5.5
Age group
Adult (%) 96.4 93.3 96.7 97.5 86.7 85.3 95.4
Youth (%) 3.6 6.7 3.3 2.5 13.3 14.7 4.6
Travel orientation
Street (%) 4.9 4.9 17.6 3.7 17.8 4.4 4.8
Sidewalk (%) 94.9 93.3 78.3 94.0 77.8 89.7 91.3
Median (%) 0.2 1.8 4.1 2.3 4.4 5.9 3.9
34 Transportation Research Record 2464
above and also to illustrate the fact that individual persons suffer the
consequences of the safety concerns identified.
In the analysis, several of the fatal and severe crashes relevant to
the crash clusters were associated with incomplete crash records,
inhibiting completion of the profiles. The rough summaries that
were completed indicated that the crash data records, as provided
by the state department of transportation, were lacking in sufficient
detail to relate accurately crash incidence to the specific deficiencies
detailed in the pedestrian audit. Access to original police reports
or the inclusion of additional attributes in the crash database (e.g.,
FHWA crash types, specific citations issued) would be needed to
evaluate crash causation more clearly and to highlight effectively
the stories of the individuals involved.
Recommended Interventions
Recommendations were developed for each of the three crash clusters
for education, enforcement, and design solutions to mitigate observed
or inferred pedestrian safety risks. The recommendations were based
on recommended best practices from AASHTO (29) and the Public
Rights-of-Way Access Advisory Committee (30).
Despite the frequently insufficient data provided in the crash
records, evaluation of the built environment surrounding the crashes
provided ample evidence of the design countermeasures that should
be implemented, as well as a means for prioritizing the measures
based on audit scores and crash locations. The additional study of the
context and demand characteristics of that area informed opportuni-
ties to improve safety through enforcement and educational efforts
targeting specific behaviors, although, as previous researchers have
observed (7, 8), some pedestrian behaviors will not improve until
infrastructure changes. For example, the study findings supported
the inference that many collisions may have occurred because of
lack of safe, convenient crossings across major arterial corridors.
Some crashes attributed to pedestrian actions may have indicated
areas where deficiencies in the provision of such crossings (e.g.,
long pedestrian delays, limited visibility, or obstructions) resulted
in risky behaviors.
In other cases, motorist violations, including impairment, distracted
driving, and aggressive driving, were to blame. In addition to educa-
tion and enforcement to deter these behaviors, design solutions can be
implemented to maximize awareness of the presence of pedestrians
along the corridors and to enhance pedestrian visibility physically.
Through a synthesis of qualitative research and quantitative data sets,
this research highlights infrastructure deficiencies and the behavioral
issues that were identified at each site and proposes various inter-
ventions that could help prevent the continued occurrence of crash
incidents.
DISCUSSION OF RESULTS
The study was aimed at understanding, through the evaluation
of several facets of safety concerns facing pedestrians, why crashes
occurred consistently in certain locations. The study was an explor-
atory effort to understand better some of the complexities impact-
ing safety outcomes through an evaluation of readily available data
sources as well as through direct observation of conditions and
behaviors in the built environment. The study was also intended to
present a comprehensive illustration of the problematic conditions
or behaviors that contribute to crash incidence and provide a basis
for discussing recommended countermeasures and generating sup-
port for those interventions. Ultimately the study was part of a larger
effort to affect policy development, implementation, and evaluation.
Like most regions, New Orleans has long provided accommo-
dations for pedestrians along roadways throughout the urbanized
area, but there is still much room for improvement. As anticipated,
the study illuminated many infrastructure deficiencies. For exam-
ple, many facilities would need to be retrofitted to comply with the
American Disabilities Act (ADA), including accommodations for
the hearing and visually impaired, because such accommodations
were virtually nonexistent. And many facilities needed upgrades to
outdated or nonfunctional equipment. In suburban parishes in the
New Orleans region, there were substantial gaps in pedestrian net-
works, creating a challenge and an opportunity to build high-quality,
new facilities in accordance with national best practices, making
a tremendous impact on overall connectivity. These are concerns,
no doubt, to which many urban and suburban areas across the nation
can relate.
Many accessibility issues were highlighted in the analysis of the
three sites, with important implications for the region’s efforts to
transition to ADA compliance. The study found rampant accessi-
bility issues in the area, with many curb ramps absent and many
sidewalk repairs needed. Sidewalk widths were below minimum
standards in some places. Removal or relocation of obsolete street
furniture, transit shelters, and utilities could help restore continuous
access. Consideration should be given to possible road diets, where
traffic volumes permit, to enhance pedestrian, transit, and bicycle
accessibility. Additional enforcement would be needed to maintain
sidewalk access for pedestrians, as parked cars and other obstruc-
tions were frequent, while educational campaigns targeting unsafe
behaviors could help reduce risk exposure for vulnerable users.
More broadly, the study revealed a need for systemic changes
in how pedestrian accommodation is provided and prioritized and
in how crash data pertaining to nonmotorized users are collected,
coded, and disseminated. The region has often failed to provide
adequate pedestrian facilities in areas with high user demand and in
areas with a poor safety record. The study provided a tool to advo-
cate for evidence-based improvements in how the region addresses
pedestrian safety overall.
The methodology of the analysis had some clear limitations. The
study was intended to be as comprehensive and robust as possible,
given readily available or easily collected data and a small research
budget. First, there were considerable limitations in the crash data
set that was used to evaluate crash locations and contributing fac-
tors. Importantly, there was limited use of statistical analysis (other
than the Statistical Analysis of Crime crash cluster tool) because
of the small sample sizes, which would have made it difficult to
generate significant results. Moreover, it is rare that planning agen-
cies would even want to adopt a tool that requires advanced sta-
tistics, given the limited staff resources at such agencies. A future
study that would include more nodes could result in a quantitative
model to test for correlations among variables. But this type of study
would be more of an academic exercise than a tool that planning
agencies would embrace. The accuracy and specificity of crash data
in the New Orleans region, as in many communities, are improv-
ing with advances in the use of GPS technology. Future analysis
efforts should therefore find it easier to retrieve and evaluate data.
However, it is important to note the importance of educating police
officers about the need to ensure consistent and complete informa-
tion about pedestrian crash incidents, which are often not taken as
seriously as vehicle-to-vehicle crashes.
Tolford, Renne, and Fields 35
Second, there were limitations inherent in the pedestrian audit
survey instruments. Although the tools were developed to reflect
compliance with AASHTO guidelines and other national best prac-
tices in pedestrian design, some important elements of pedestrian
safety and comfort were excluded. The study did not include the
presence or absence of pedestrian-scale lighting, because the audits
were assumed to have been conducted during daylight. And the
study did not include certain elements of ADA compliance, such
as the presence of detectable warnings or other aids for the visually
impaired and the grades and cross slopes of facilities. A recent study
in Atlanta developed a methodology, still in beta testing, that pro-
vided a low-cost computer application to collect some of these types
of data, which could be combined with the methodology presented
here for a more robust analysis (31). Noninfrastructure consider-
ations, including crime, blight, and vacancy, were also outside the
scope of the study. In addition, the pedestrian audit toolkit would
not be applicable to all roadway contexts, limiting replicability of
the study in some areas. Unique roadway geometries (e.g., round-
abouts) were not represented and the toolkit was not designed for
use on rural or very low volume roadways [although Gross et al.
(32) have developed a road safety audit for the latter circumstance].
Third, the assumptions underlying some of the tools used (e.g.,
the National Bicycle and Pedestrian Documentation Project’s
estimated daily traffic extrapolation and adjustment factor meth-
odology) have not been extensively evaluated for reliability in
this region, which is known to diverge somewhat from model travel
patterns applicable to other southern cities (33). Finally, evaluation
of demographic and contextual factors was limited to immediately
available data sources, and correlations between the variables and
crashes were not evaluated in-depth because of the time constraints
of the study.
Despite these limitations, the pedestrian safety evaluations pro-
duced according to this methodology provided the sponsoring agency
with sufficient evidence to guide planning efforts relating to trans-
portation infrastructure improvements. Future research could address
the limitations by incorporating additional data sets into the meth-
odology and developing models by which to evaluate correlations
among variables and determine crash prevalence and exposure rates.
In addition, a need exists to develop user-friendly tools for evaluat-
ing driver behavior and identifying appropriate countermeasures to
observed problems.
Each of the study’s three pilot crash clusters was selected in part
because of planned roadway or development projects occurring
within or adjacent to the study area. Although no interventions have
yet been made, as these projects progress it is essential that pedes-
trian infrastructure improvements are integrated into project plans
to restore or improve pedestrian access to land uses in the area, cre-
ate safer and more visible crossings to reduce crash incidence, and
enhance the economic revitalization of these corridors by creating
comfortable spaces for people to live, work, and play.
Better understanding the conditions present in a specific node,
neighborhood, city, or region that affect safety outcomes can help in
more effective prioritization of the use of limited resources for near
term interventions, as well as to plan holistically for programs and poli-
cies that will guide transportation planning in the long term. According
to AASHTO, local, regional, and state governments should consider
the following criteria in evaluating and ranking possible infrastructure
investments (29):
v Existing pedestrian volumes;
v Presence of major pedestrian generators;
v Speed of the roadway;
v Street classification;
v Crash data;
v School zones and catchment areas;
v Transit routes;
v Urban centers and neighborhood commercial areas;
v Low-income neighborhoods;
v Missing links in existing infrastructure networks;
v Priorities identified by residents, including requests to correct
identified problems;
v Diversity of activity types;
v Established ADA transition plan priorities and programs; and
v Planned roadway resurfacing projects.
Priority intervention areas should include areas that meet more
than one of these criteria. The methodological framework outlined
in this research facilitates the consideration of almost all of these
prioritization criteria (and could be expanded to address all the
factors listed) for evaluation of any district in need of pedestrian
infrastructure improvements.
As a result, this methodology could be useful to establish invest-
ment priorities for local areas, especially those looking to implement
Complete Streets policies. The adoption and implementation of a
Complete Streets approach creates an opportunity for—and often
demands—development of new processes for collecting and utiliz-
ing data. Analysis techniques such as those described in this research
could be used to (a) evaluate pedestrian conditions at the project
level and identify recommended improvements; (b) prioritize invest-
ments across a jurisdiction to ensure that resources are applied where
most needed; and (c) measure progress toward policy implementa-
tion, capturing changes in key metrics, including crash totals and
severity, built environment audit scores, and user volumes over time.
Jurisdictions with recently adopted Complete Streets policies should
consider incorporating multi-tool analysis frameworks to ensure a
coordinated, data-driven approach to policy implementation.
CONCLUSION
The basic framework described in this paper for evaluating pedes-
trian safety, prioritizing investment, and tracking change could be of
use to many local and regional agencies, consultants, and research-
ers. The specific recommendations for improvements at the target
areas in this study have not yet been implemented; therefore, no
ex post analysis is available at this time. However, education and
engineering interventions have been made in response to previous
PBRI studies, and this study has been used as a resource for the
development of the region’s Pedestrian Safety Action Plan. The
research proposed a way to delve pragmatically into crash analysis
and mitigation, particularly in areas where access to data is some-
what constrained or nonmotorized data collection programs have not
yet been institutionalized. The methodology developed in the study
should be especially valuable for areas that have adopted policies
that require such data for effective implementation.
The approach used in this study could be of particular benefit in
supporting Complete Streets policy implementation and evaluation.
Effective policy implementation should include coordination and
communication among stakeholders, including government agen-
cies, developers, advocates, engineers and planning professionals,
and the local community. A multifaceted approach to collecting and
interpreting data and highlighting safety concerns and opportunities
36 Transportation Research Record 2464
for change can be used to facilitate dialogue among stakeholders
and promote closer coordination on improving pedestrian safety
across agencies or jurisdictions. The approach can also guide pedes-
trian accommodation in a Complete Streets policy framework to
ensure that the needs of all users are being adequately considered
on all projects and that investments are made equitably and where
they will have the greatest impact.
Overall, development of a flexible, low-cost methodology for
conducting localized nonmotorized safety research will advance the
efforts of communities to improve safety outcomes, address acces-
sibility shortcomings, and implement new and innovative Complete
Streets policies.
ACKNOWLEDGMENTS
The authors thank the New Orleans Regional Planning Commission
and the Louisiana Department of Transportation and Development
for sponsoring this research. The authors also recognize the contri-
butions of Karen Parsons and Dan Jatres for the development of the
Pedestrian Bicycle Resource Initiative and many tools employed in
this study. The authors thank the student researchers who assisted
with data collection and analysis in conjunction with this research.
REFERENCES
1. FHWA. Pedestrian and Bicycle Safety. Undated. http://safety.fhwa.dot.
gov/ped_bike/. Accessed July 2, 2013.
2. Tolford, T. Complete Streets Policy Manual. Center for Planning Excel-
lence, Baton Rouge, La. 2012. http://connect.cpex.org/files/2012/08/
CSmanual_FINAL.pdf. Accessed July 2, 2013.
3. Schmitt, A. Passing a Law Is the Easy Part: The Challenge of Building
Complete Streets. Streetsblog USA, New York. http://dc.streetsblog.
org/2013/07/11/passing-a-law-is-the-easy-part-the-challenge-of-building-
complete-streets/. Accessed July 11, 2013.
4. Ismail, K., T. Sayed, and N. Saunier. Methodologies for Aggregating
Indicators of Traffic Conflict. In Transportation Research Record: Jour-
nal of the Transportation Research Board, No. 2237, Transportation
Research Board of the National Academies, Washington, D.C., 2011,
pp. 10–19.
5. Miller, J. S., N. J. Garber, and S. K. Korukonda. Understanding Causality of
Intersection Crashes: Case Study in Virginia. In Transportation Research
Record: Journal of the Transportation Research Board, No. 2236, Trans-
portation Research Board of the National Academies, Washington, D.C.,
2011, pp. 110–119.
6. Miller, J. S. Technical Details for Integrating Safety and Planning at
Level of Individual: Regional Case Study. In Transportation Research
Record: Journal of the Transportation Research Board, No. 2318, Tr an s-
portation Research Board of the National Academies, Washington, D.C.,
2012, pp. 35–44.
7. Zaki, M. H., T. Sayed, K. Ismail, and F. Alrukaibi. Use of Computer
Vis ion to I dent ify Ped est ria ns’ No nco nfo rmi ng B eha vior at U rba n In ter-
sections. In Transportation Research Record: Journal of the Transpor-
tation Research Board, No. 2279, Transportation Research Board of the
National Academies, Washington, D.C., 2012, pp. 54–64.
8. Savolainen, P. T., T. J. Gates, and T. K. Datta. Implementation of Targeted
Pedestrian Traffic Enforcement Programs in an Urban Environment. In
Transportation Research Record: Journal of the Transportation Research
Board, No. 2265, Transportation Research Board of the National Acad-
emies, Washington, D.C., 2011, pp. 137–145.
9. Cooper, J. F., R . J. Schneider, S. Ryan, and S. Co. Documenting Targeted
Behaviors Associated with Pedestrian Safety. In Transportation Research
Record: Journal of the Transportation Research Board, No. 2299, Tra ns-
portation Research Board of the National Academies, Washington, D.C.,
2012, pp. 1–10.
10. Schneider, R. J., T. Henry, M. F. Mitman, L. Stonehill, and J. Koehler.
Development and Application of a Pedestrian Volume Model in San
Francisco, California. In Transportation Research Record: Journal of
the Transportation Research Board, No. 2299, Transportation Research
Board of the National Academies, Washington, D.C., 2012, pp. 65–78.
11. Bricka, S., I. N. Sener, C. Dusza, N. Wood, and J. G. Hudson. Factors
Influencing Walking in Small Urban Region. In Transportation Research
Record: Journal of the Transportation Research Board, No. 2307, Tr an s-
portation Research Board of the National Academies, Washington, D.C.,
2012, pp. 52–59.
12. Weinberger, R., and M. N. Sweet. Integrating Walkability into Planning
Practice. In Transportation Research Record: Journal of the Transpor-
tation Research Board, No. 2322, Transportation Research Board of the
National Academies, Washington, D.C., 2012, pp. 20–30.
13. Shin, H.-S., G. Chen, and G. Holisko. Pedestrian Safety Programs in
Centers of Large Cities: Institutional Settings and Identified Barriers. In
Transportation Research Record: Journal of the Transportation Research
Board, No. 2264, Transportation Research Board of the National Acad-
emies, Washington, D.C., 2011, pp. 119–127.
14. Ferrell, C. E., and S. Mathur. Influences of Neighborhood Crime on Mode
Choice. In Tr ans port ati on Res earch R ecor d: Jour nal of th e Tr ans port a-
tion Research Board, No. 2320, Transportation Research Board of the
National Academies, Washington, D.C., 2012, pp. 55–63.
15. Kravetz, D., and R. B. Noland. Spatial Analysis of Income Disparities in
Pedestrian Safety in Northern New Jersey: Is There an Environmental
Justice Issue? In Transportation Research Record: Journal of the Trans-
portation Research Board, No. 2320, Transportation Research Board of
the National Academies, Washington, D.C., 2012, pp. 10–17.
16. Gladhill, K., and C. M. Monsere. Exploring Traffic Safety and Urban
Form in Portland, Oregon. In Transportation Research Record: Jour-
nal of the Transportation Research Board, No. 2318, Transportation
Research Board of the National Academies, Washington, D.C., 2012,
pp. 63–74.
17. Tal, G., and S. Handy. Measuring Nonmotorized Accessibility and Con-
nectivity in a Robust Pedestrian Network. In Transportation Research
Record: Journal of the Transportation Research Board, No. 2299,
Transportation Research Bo ard of the National Acad emies, Washington,
D.C., 2012, pp. 48–56.
18. McReynolds, R. L. Unsuccessful Implementation Efforts: Lessons
Learned. In Transportation Research Circular E-C070: Optimizing the
Dissemination and Implementation of Research Results: A Summary
of Workshop and Midyear Meeting Activities. Transportation Research
Board of the National Academies, Washington, D.C., 2004, pp. 14–18.
http://onlinepubs.trb.org/onlinepubs/circulars/ec070.pdf. Accessed
July 3, 2013.
19. Khisty, C. J., and S. Kikuchi. Urban Transportation Planning Educa-
tion Revisited: Reading the Dials and Steering the Ship. In Transporta-
tion Research Record: Journal of the Transportation Research Board,
No. 1848, Transportation Research Board of the National Academies,
Washington, D.C., 2003, pp. 57–63.
20. Regional Planning Commission for Orleans, Jefferson, Plaquemines,
Saint Bernard, and Saint Tammany Parishes. 2005 New Orleans Metro-
politan Bicycle and Pedestrian Plan. New Orleans, La., 2006. http://
transportation.uno.edu/phire-content/assets/files/PBRI-2005-New-
Orleans-Metropolitan-Bicycle-and-Pedestrian-Plan.pdf. Accessed
July 3, 2013.
21. University of New Orleans. Statistical Analysis of New Orleans 1999–
2006 Crash Data. Regional Planning Commission for Orleans, Jefferson,
Plaquemines, Saint Bernard, and Saint Tammany Parishes. New Orleans,
La., 2009. http://transportation.uno.edu/documents/bike-pedestrian/
PBRI-Statistical-Analysis-of-New-Orleans-1999-2006-Crash-Data.
pdf. Accessed July 3, 2013.
22. Fields, B., and T. Tolford. New Orleans Regional Pedestrian and Bicycle
Crash Report, 2006–2008. Regional Planning Commission for Orleans,
Jefferson, Plaquemines, Saint Bernard, and Saint Tammany Parishes. New
Orleans, La., 2011. http://www.norpc.org/assets/pdf-documents/studies-
and-plans/PBRI_CrashReport2006-2008_FINAL.pdf. Accessed July 3,
2013.
23. Tol fo rd, T. New Orleans Regional Pedestrian and Bicycle Crash Report,
2009–2010. Regional Planning Commission for Orleans, Jefferson,
Plaquemines, Saint Bernard, and Saint Tammany Parishes, New Orleans,
La., 2012. http://www.norpc.org/assets/pdf-documents/studies-and-
plans/PBRI_CrashReport_2009-2010_FINAL_7-2-12.pdf. Accessed
July 3, 2013.
24. Levine, N. CrimeStat III: A Spatial Statistics Program for the Analysis
of Crime Incident Locations. Ned Levine and Associates, Houston, Tex.,
Tolford, Renne, and Fields 37
and The National Institute of Justice., Washington, D.C., 2010. http://
www.icpsr.umich.edu/CrimeStat/. Accessed July 3, 2013.
25. University of North Carolina Highway Safety Research Center. Crash
Type Manual for Bicyclists. Publication FHWA RD-96-104. FHWA,
U.S. Department of Transportation, 1996. http://www.fhwa.dot.gov/
publications/research/safety/pedbike/96104/.
26. Renne, J. L. Auditing Neighborhoods, Streets, and Intersections for
Pedestrian Safety: A Toolkit for Communities. Regional Planning Com-
mission for Orleans, Jefferson, Plaquemines, Saint Bernard, and Saint
Tammany Parishes, New Orleans, La., 2009. http://transportation.uno.
edu/phire-content/assets/files/PBRI-Auditing-Neighborhoods-Streets-
and-Intersections-for-Pedestrian-Safety.pdf. Accessed July 3, 2013.
27. National Bicycle and Pedestrian Documentation Project. National Bicy-
cle and Pedestrian Documentation Project: Instructions. 2 0 10 . h t t p : / / b i k e
peddocumentation.org/index.php/download_file/-/view/16. Accessed
July 3, 2013.
28. National Bicycle and Pedestrian Documentation Project. Count Adjust-
ment Factors. 2009. http://bikepeddocumentation.org/index.php/
download_file/-/view/11. Accessed July 3, 2013.
29. Guide for the Planning, Design, and Operation of Pedestrian Facilities.
AASHTO, Washington, D.C., 2004.
30. Public Rights-of-Way Access Advisory Committee Subcommittee on
Tec hn ica l A ss ist an ce. Special Report: Accessible Public Rights-of-Way
Planning and Designing for Alterations. 20 07. htt p:/ /ww w.ac ces s-b oa rd.
gov/guidelines-and-standards/streets-sidewalks/public-rights-of-way/
guidance-and-research/accessible-public-rights-of-way-planning-and-
design-for-alterations. Accessed July 3, 2013.
31. Grossman, A., A. Frackelton, E. Palinginis, Y. Xu, V. Elango, F. Castrillon,
R. Sadana, and R. Guensler. Sidewalk Quality Analysis Towards More
Accessible Urban Environments. Presented at Joint Association of
European Schools of Planning/Association of Collegiate Schools of
Planning Congress, Dublin, Ireland, July 2013.
32. Gross, F., K. Eccles, and D. Nabors. Low-Volume Roads and Road
Safety Audits: Lessons Learned. In Transportation Research Record:
Journal of the Transpo rtat ion Research Board, No. 22 13, Transportation
Research Board of the National Academies, Washington, D.C., 2011,
pp. 37–45.
33. Fields, B. Active Transportation Measurement and Benchmarking
Development: New Orleans State of Active Transportation Report,
2010–2011. Gulf Coast Research Center for Evacuation and Transpor-
tation Resiliency, New Orleans, La., 2012. http://www.evaccenter.lsu.
edu/pub/11-05Part%201.pdf. Accessed July 3, 2013.
The opinions, findings, conclusions, or recommendations expressed in this paper
are those of the authors and do not necessarily reflect the views of the Louisiana
Department of Transportation and Development, the New Orleans Regional
Planning Commission, or any of the other agencies that provided financial or
material support.
The Pedestrians Committee peer-reviewed this paper.
