Ghost Town: A review of DC area bike fatalities

Part 1 in a series

Since 1987, over 100 DC area cyclists have died in motor vehicle crashes. This map shows where they were.¹ And there's just one intersection in the region which had two separate fatal crashes. 

In the above map, red pins show crashes in an intersection, yellow in the roadway, black in a crosswalk, blue on the shoulder of the roadway, orange on a sidewalk, green in a bike lane, or white where the location was not available.²

These fatalities have occurred in every jurisdiction, on busy highways and quiet neighborhood streets, and on every part of the roadway from sidewalks to traffic lanes.

The real "Intersection of Doom" is at Gaithersburg's edge

The intersection of Lee Highway and North Lynn Street, where drivers make a right turn across cyclists' path coming off the Mount Vernon Trail, gets much coverage as the "Intersection of Doom." But fortunately, I found no actual bicycle fatalities there.

Nor were there any where the Mt. Vernon Trail connections cross the George Washington Parkway, another harrowing experience for cyclists and a big problem spot that needs fixing. But there was one location where two separate fatal bike crashes occurred.

In 1997, a driver hit 15-year-old Alexis Smith on her bicycle in the crosswalk as she crossed the ramp from Great Seneca Highway to Sam Eig Highway, just west of the end of I-370 in Montgomery County. Then in 2009, another driver hit and killed Codi Alexander, 16 at the same spot. However, Montgomery County wasn't the place with the most fatal bike crashes.

Prince George's has the most deaths by far

Of the seven jurisdictions I looked at, Prince George's had the most fatalities, with 36. Here is the full list:

Jurisdiction	Number
Prince George's	36
District of Columbia	25
Montgomery	21
Fairfax	19
Alexandria	2
Falls Church	2
Arlington	2

Some of the variation might be explained by population and square mileage, but Prince George's County is neither the largest nor the most populous. And comparisons get more complicated because DC's surge of daytime population means that considering its resident population understates the amount of exposure cyclists have there.

Most fatal crashes happen at intersections

If we combine fatalities listed as in the intersection and in the crosswalk, it shows that more than half of all fatal crashes happen at intersections. (Some crashes listed as on sidewalks or in bike lanes also may be at intersections.)¹

Type	Number
Intersection	37
Roadway	36
Crosswalk	20
Shoulder	7
Bike lane	3
Sidewalk	2
Unknown	1

Where this data comes from

I assembled this list and map from two main sources: media reports and the National Highway Traffic Safety Administration's (NHTSA) Fatality Analysis Reporting System (FARS). Most media reports are newspaper accounts available on highbeam, which is why they only go back to 1987. These accounts are usually very accurate and reliable. I only flagged one possible error during the review.

However, these are not particularly comprehensive. Only 53% of all fatal bike crashes get reported in newspapers, and usually as only one story about the fatal crash itself. Occasionally a reporter will follow up with a second item once authorities release the victim's name. For a particularly sensational story, there may coverage all the way through a trial and sentencing. Most media accounts, however, just end with a line saying something like, "Police are continuing to investigate the incident."

The NHTSA FARS data, on the other hand, is significantly more comprehensive but riddled with a vast array of errors. It also only goes back to 1994. Some of the errors come from problems with the forms themselves, while people filling them out introduce others.

These errors ranged from trivial cases, such as mislabeling a female fatality as male, to nonsensical cases where a bike fatality was coded as "Safety Belt Used Improperly," to the outright misleading case where a cyclist was mislabeled as a pedestrian. But 98% of the fatalities with media accounts also appeared in FARS.

Still, FARS data under-counts total bike fatalities because it does not include crashes on driveways or parking lots or crashes that don't involve a motor vehicle. I identified 15 such fatalities. In addition, the United States Park Police apparently doesn't submit FARS forms to the NHTSA, as crashes they investigated don't appear. Nor do bike deaths that arise from medical conditions such as heat stroke or from murder (except in the one case where the murder weapon was a car). So while the FARS data is more comprehensive, it is not complete.

The map above includes every bike fatality identified except for one that had an unworkable location description.³ 

The original version of this post failed to count one Arlington fatality. Part of the reason for this was that the two Arlington County fatalites both occurred on the same day, May 8th, but in different years. 

Crossposted at GreaterGreaterWashington

1 Prior to 2001, all FARS location data is reported in the form of road designator and milemarker. Starting in 2001 GPS coordinates are also included. Maryland and Virginia report the road designator by an official route number that may not be well known (Cherrywood Lane in PG County, for example, is MU 40). DC, God bless them, reports the actual name of the street as used on streetsigns. As a result, locations for Maryland and Virginia FARS-only derived data prior to 2001 is approximate.

² Roadway fatalities are those that occur in travel lanes away from intersections.

³ That one was reported to FARS as being on county road 0123 in Montgomery County, but there is no 123 in Montgomery County. There is one in Prince George's County. So either the county was coded wrong (33 instead of 31, for example) or the road was.

Comments

[Michael H]

General comment: I don't see a link to your Twitter feed anywhere on the site. I think that would be useful.

[Greenbelt]

Nice work. Most of the fatalities in my home area seem to be on-shoulder or on-road hit from behinds. That's pretty depressing.

[DE]

I am far from certain on this because I don't get over there that much, but I did work in Lanham for a while, and I'd say that there seem to be more roads in PG county that have high speed limits and cycling combined. In Arlington and DC, the speed limits tend to be lower on the types of roads that cyclists use. Fairfax of course has its share of roads with high speed limits as well.

[Nancy]

re Wolfgang Jakobsberg's collision (Seven Locks Rd near Bradley Blvd in Potomac, MD): The driver claimed Mr. Jakobsberg swerved into his lane. The problem is that the shoulder Mr. Jakobsberg was riding on disappears right at the area of the collision. You can see it in this Google streetview: https://www.google.com/maps/@39.0137783,-77.1602661,3a,75y,199.45h,67.62t/data=!3m4!1e1!3m2!1svSuQP_VvnHFvE02f5pZAJw!2e0

I can't believe after all these years that nothing has been done about this conflict point (and there are still no sidewalks for kids to walk to the elementary school).

Anyway, I would put the onus on the driver for attempting to pass the cyclist without leaving sufficient room. (It happened in the early afternoon, so lighting was not an issue.)

[T]

Depressing to view and interesting at the same time.

The one closest to me involves a little girl on a street I know is entirely residential. Running a stop sign and getting killed should not have happened there. The speed limit is maybe 20mph and it's just a tough place to speed. Perhaps a wrong time/wrong place/crappy outcome scenario, but still, wow.

[jeffb]

@Nancy

I used to ride this section as part of my daily commute. I found it safer to cross the road to the opposite side and ride on the fairly wide shoulder facing traffic.

Not ideal for sure. And one time I had an SUV driver pull out of the post office (it's now a temple i believe) and upon seeing me he deliberately drove his car up the shoulder right at me forcing me to bail off the bike and into the grass.

Fun days in suburbia :)

[D]

very sad. thanks for putting this together though.

[Nancy]

@Jeffb--I used to ride that as my daily commute, too. I had to turn left (east) from Seven Locks onto Bradley to get home, so I got used to taking the lane well before the intersection and before the shoulder disappeared.

Going the other direction past the post office one day, a driver drove his car the wrong way on the shoulder right towards me in an attempt to get into the parking lot. He didn't understand why I went ballistic. (He actually followed me back down Seven Locks and asked me. Weird, weird guy.)