High Injury Network gives most holistic portrayal of safety trends on LA’s streets

Did you see Estey & Bomberger’s study that claimed to identify what they saw as the “most dangerous” intersections throughout California?

Yeah, so did we.

Source: Estey & Bomberger "Most Dangerous Intersections in California [New Study]"

Source: Estey & Bomberger’s “Most Dangerous Intersections in California [New Study]”

Looking specifically at crashes that occurred at intersections, a personal injury law firm in San Diego published alarming (although, not surprising) research in collaboration with 1point21 Interactiv.

The study received a lot of media attention!

One of the outlets that picked up this story was LAist. They highlighted the fact that “of the 444 most dangerous intersections in California, a shocking 221 of those intersections are in the City of Los Angeles.”

ABC10 noted that “the No.1 [most] dangerous intersection of Devonshire St. and Reseda Bl. in Los Angeles had a total score of 147 with 24 crashes and 41 injuries.”

Screenshot of CurbedLA's article about the Estey & Bomberger study

And, Curbed LA described how the firm “found [Los Angeles] is home to three of the most dangerous intersections in California—two of which are in the same neighborhood (Northridge).”

We agree that streets in Los Angeles have a lot of collisions. In fact, our Vision Zero team of experts is focused specifically on making our 7,500 miles of roadway safer! But, our approach to defining safety is quite different than how Estey & Bomberger examined safety in their study.

We would like to share with you how our process paints a more accurate picture of which streets in Los Angeles need safety improvements the most.

Vision Zero’s Analysis

We have been conducting our own analysis that focuses on severe injuries and deaths of our most vulnerable road users, starting with people walking and bicycling. Why did we start there?

In our collision data, we found that there is an over-representation of people walking & biking. They account for roughly 15% of all collisions, but approximately 50% of all deaths. If you think about it, this makes a lot of sense given that people walking and biking are exposed to the elements, not shrouded by 4,000 lbs of metal and safety features like airbags and seat belts.

Vision Zero is an injury reduction strategy, not a collision reduction strategy. So, in order to achieve our goals as quickly as possible, we must focus on areas that will deliver the biggest return on investment, such as locations that have a higher concentration of deaths and serious injuries for the most vulnerable population.

Our analysis identified a network of streets, called the High Injury Network (HIN), that has a higher incidence of severe and fatal collisions. The HIN is comprised of 386 corridors that represent 6% of Los Angeles’ street miles. 65% of all deaths and severe injuries involving people walking and biking occur on these 6% of streets.

We also give more weight to counts of Killed or Serious Injuries (commonly referred to in the transportation safety field as “KSI”) among people walking or biking, so deaths or serious injuries at all intersections are multiplied by three, while vehicle-vehicle deaths or serious injuries do not receive a multiplying factor. For example, if an intersection contains one fatal pedestrian collision, two severe bicycle injuries, and one fatal vehicle-vehicle, the score would be 10 (3 for the pedestrian, 6 for the two bicycles, and 1 for the vehicle-vehicle). 

The Estey & Bomberger study data overlaps with some portions of the High Injury Network, but because of their methodology, there are segments in South LA and Northeast LA that don’t show up at all in their data. This is significantly contradictory to our findings that areas in South LA, Northeast LA, and neighborhoods with poor health outcomes have a disproportionate amount of severe and fatal injuries from collisions. In fact, nearly half of the HIN falls within our most vulnerable communities.

Why does our ranking look different?

First, we use different data.

Estey & Bomberger used 2015 collision data from the Statewide Integrated Traffic Records System (SWITRS). By using data from only one year, their intersection score is reflective of a snapshot in time.

However, we analyzed 5-years of SWITRS data from 2009-2013 (the most recently available data at the time of our analysis), which gives us an opportunity to see trends over time. Using only one-year of collision information provides incomplete information – there are often many factors that go into a collision, ranging from construction, to weather, to individual behavior. In order to get a better snapshot of locations that have systemic issues and demand engineering, enforcement, and education efforts, we undertook a 5-year longitudinal analysis to identify patterns.  We recognize that we need to update our analysis with data from 2014, 2015, and 2016, much of which is now (provisionally) available on SWITRS. This is a key strategy in our forthcoming Vision Zero Action Plan. However, our five-year analysis is still effective at identifying locations with significant safety needs.

On top of SWITRS, we also incorporate data from the LA County Department of Public Health to provide information about social equity and health outcomes.

Second, we use a different formula for project prioritization.

Estey & Bomberger assigned values to each crash, injury, and fatality to calculate a score to rate the danger, using this formula:

Crash (x1) + Injury (x3) + Fatality (x10) = Intersection Score

You can see that under their model, all injuries are weighted the same. This means that the formula does not distinguish between a fractured finger and a severe injury, like a broken skull. As a result, the Estey & Bomberger study winds up giving more weight to car crashes than we do. This is because their data includes minor car collisions in which no one was severely injured. Since vehicles are getting safer every year thanks to enhanced safety features, minor collisions tend to result in minor injuries.

Our formula is more complex. We worked with community members and technical experts to develop  the following three location-based priorities to include in an intersection score methodology:


Location Priority
Safety Question
How many severe or fatal injury (KSI) collisions have taken place at the intersection?
Has the intersection had KSI collisions that involve older adults or children?
Social Equity
Is the intersection in a community that has been traditionally underinvested in or underserved?

These three priorities were identified as more important than all others in determining locations that deserve attention. They help us emphasize important aspects of a community that might not show up in a simple collision analysis. For instance, it is important for us to prioritize the most vulnerable road users, so we emphasize collisions that involve children, senior citizens, people walking, and people biking.

Then, we put all of this data into our HIN intersection formula:

Fatality (x1.5)* + Severe Injury** + Child or Senior*** + Target Community**** = Intersection Score

*weighted higher for severity

**raw value

***0 or 1 if a child or senior was present

****0 or 1 is the location was in a target community


Intersections with high scores highlight patterns, and these patterns show us which roadways and neighborhoods in Los Angeles warrant the most attention.

Similarly, we identify which thoroughfares have the highest numbers of fatalities and severe injuries to analyze corridors. One reason that it is important to analyze corridors in addition to intersections is that speed is a contributing factor to many serious crashes. Speed itself must be addressed not just at a single physical point, but along a stretch of roadway where that speed is attained. Intersections are where conflicting movements occur, so they certainly can be part of the solution. However, if speeds are reduced, people can crash into each other with much less collateral damage.

Addressing the Cause

It’s important for us to incorporate data about the cause of a collision so that we can identify tailored solutions for each corridor and intersection. For this reason, our analysis identifies site-specific countermeasures, such as highly visible crosswalks and protected bike lanes, that can be implemented to prevent future collisions.

We have found that examining the safety of a corridor allows us to look at factors that provide a holistic context for why collisions occur, including:

  • Proximity to freeways
  • Number of vehicle lanes
  • Presence of a bike lane
  • Speed limit
  • Average vehicle speed


Our analysis has resulted in 12 statistically significant “collision profiles” for pedestrian and bicycle collisions that result in death or serious injury, which help refine our project development process.

Collision Profile Name
% of Total KSI Collisions on HIN
Lack of crossing infrastructure
-Pedestrian crossing not at crosswalk or walking on shoulder
-Pedestrian collisions with motor vehicles
46.7% (of ped KSI collisions)
Lack of bicycle facility
-No bike path or bike lane present
-Bicycle collisions with motor vehicles
72.6% (of bike KSI collisions)
Intersecting arterials
-Intersection made up on major or secondary highways or reclassified highways
-Pedestrian/bicycle collisions with motor vehicles
-Driver did not stay after KSI collision
-Pedestrian/bicycle collisions with motor vehicles
-Driver or pedestrian/bicyclist had been under the influence of alcohol or drugs
-Pedestrian/bicycle collisions with motor vehicles
Complex intersections
-Intersection has 5 or more legs, including divided highways
-Pedestrian/bicycle collisions with motor vehicles
Left turns at signals
-Vehicles making left turns Most KSIs occur at locations with no protected-only left turn signals
-Pedestrian/bicycle collisions with motor vehicles
-Speed is primary factor in collisions
-Pedestrian/bicycle collisions with motor vehicles
Children near schools
-School with ¼ mile Most KSIs occur between 6am-9am and 12pm-6pm
-Pedestrian/bicyclist age 5-17 collisions with motor vehicles
Right turns at signals
-Vehicles making right turns
-Pedestrian/bicycle collisions with motor vehicles
Red light running
-Vehicles proceeding straight or making left turns
-Pedestrian/bicycle collisions with motor vehicles
Freeway ramps
-KSI at ramp intersection
-Pedestrian/bicycle collisions with motor vehicles


Vision Zero Action Plan

As you can see, many deaths or serious injuries occur in places that lack pedestrian and bicycle infrastructure. Behaviors such as drinking and driving, speeding, and running red lights are also frequent causes of collisions. Now that we have a greater understanding of why these collisions occur, we are ramping up efforts to tackle this issue head on by creating awareness and developing needed infrastructure as part of our Vision Zero Action Plan!

All of this analysis will be covered in detail in our forthcoming Vision Zero Action Plan, which will also have key strategies for engineering, enforcement, education, and evaluation, all with a commitment to equity and engagement. Here’s a sneak-peak of the Vision Zero Action Plan, which will be centered around 4 outcomes:

1. Create Safe Streets for All

  • Install life-saving improvements on the priority corridors and intersections along the High-Injury Network, such as optimizing 400 traffic signals and redesigning at least 12 miles of city streets every year to accommodate safe transportation for all
  • Update 100 percent of the expired speed surveys on the priority corridors by the end of 2017


2. Develop a Culture of Safety

  • Invest at least $2 million in a comprehensive education campaign that addresses top collision factors, such as speeding and insobriety


3. Adopt New Policy and Legislation to Strengthen Safety

  • Develop a state legislative strategy that strengthens laws related to moving violations that contribute to fatal and severe injury collisions, such as speed deterrents or increased penalties for distracted driving


4. Respond to Relevant Data

  • Include Vision Zero principles in the 2017 LAPD Traffic Plan
  • Update the High-Injury Network with 2014–2016 data when it becomes available


To measure our success, we will use the following benchmarks in comparison to 2016:

  • 20% reduction in traffic deaths by 2017
  • 50% reduction in traffic deaths by 2020
  • 100% reduction in traffic deaths by 2025


And finally, starting in 2017, we will roll out a citywide education campaign that will focus on building awareness and changing the behaviors that lead to the city’s most severe and deadly collisions.

Once our Action Plan is released, we will share it with you on our blog and social media! Follow us at @VisionZeroLA and share your stories using #VisionZeroLA.

Thanks for reading!