2 percent of rear-end crashes involve a slowly-moving lead vehicle

Rear-end crashes are, by far, the most common motor vehicle crash. Looking at all MV-MV (that is, motor vehicle crashes excluding single-vehicle crashes), a whopping 47% were classified as rear-end, using 2012 Arizona data. That’s almost 50,000 rear-end collisions a year, just in Arizona!

The explanations generally run to generic excuses: the driver of the striking vehicle was driving too fast for conditions, or “distraction” — which are more-or-less true by definition.

The driver of the pickup in the pickup pictured was seriously injured Mar 25, 2016 when he or she slammed into the rear-end of a reportedly stopped school bus (none of the bus occupants were seriously injured).

Digging into the problem, one of the Human Factors / Naturalistic driving studies:

Lee, S. E., Llaneras E., Klauer S., Sudweeks J. (2007) Analyses of Rear-End Crashes and Near-Crashes in the 100-Car Naturalistic Driving Study to Support Rear-Signaling Countermeasure Development, DOT HS 810 846

Of the 7,024 observed rear-end events, 45 percent involved a decelerating lead vehicle, 38 percent involved a stopped lead vehicle, 2 percent involved a slower moving lead vehicle, and 15 percent occurred under various other situations. Crashes were predominately characterized by situations in which the lead vehicle was stopped, whereas near-crashes and incidents were more evenly distributed across instances of both stopped and decelerating lead vehicles. The majority of rear-end crash events (59%, or 16 out of 27) involved a stopped lead vehicle, while 22 percent (6 out of 27) occurred under conditions of a decelerating lead vehicle.

i.e. few rear-end collisions involve “slow” vehicles. This study is specifically only motor vehicle; but this perhaps offers a clue as to why very few Bike-MV crashes are a motorist striking the rear-end of a bicyclist — drivers are quite good at not avoiding hitting slowly moving lead vehicle, perhaps simply because they have additional reaction time.

What about MV-Bike Rear-end collisions?

Rear-end collisions are only about 3% of all Bike-MV crashes in Arizona (and that includes incidents where the bicyclist strikes the rear-end of a MV). The incidence of fatal Bike-MV, though, is much higher, at about 25% of all overall bicyclist traffic fatalities nationally, though some like to claim the figure is higher. (although state data is available, the number of bicyclist fatalities in Arizona is generally too small to draw any meaningful conclusions from so everything below refers to US national data).

Breaking down the motorist-overtaking crash group between urban and rural, and daylight vs. not-daylight is important, as it reveals what counter-measures might be expected to help.

As expected, the urban/daylight (the largest category, by the way) has only relatively fewer, only 14%, motorist overtaking; versus rural/not-daylight at 42%. The 14% (together with the 3% figure, mentioned above) suggests things like Bike Lanes, protected Bike lanes, cycle tracks are unlikely to change the safety picture much.  The much higher non-daylight fractions suggest that lighting could be an especially important countermeasure.

BIKECGP	urban/rural	Daylight/not	2010	2011	2013	average
Mot Overtaking	Rural	Daylight	29	43	44	39
Mot Overtaking	Rural	DarkDawnDusk	33	35	41	36
Mot Overtaking	Urban	Daylight	30	30	37	32
Mot Overtaking	Urban	DarkDawnDusk	62	56	65	61
ALL	Rural	Daylight	103	122	141	122
ALL	Rural	DarkDawnDusk	75	89	98	87
ALL	Urban	Daylight	229	237	243	236
ALL	Urban	DarkDawnDusk	211	229	261	234
As percent	Rural	Daylight	28%	35%	31%	32%
	Rural	DarkDawnDusk	44%	39%	42%	42%
	Urban	Daylight	13%	13%	15%	14%
	Urban	DarkDawnDusk	29%	24%	25%	26%

minutia: the table is stored at FARS.xls. PBCAT data existed only for 2010, 2011, and 2013. Unknown rural/urban’s were assigned to rural. Unknown lighting were assigned to not daylight. These adjustments were very small single digits (mostly 1’s, a 2, and and two 3’s). See the spreadsheet for adjustments. UPDATE: in late 2016 the fed’s released FARS data with pbcat crash typing for years 2014 and 2015; I don’t have it in a nice table as above but the overall composite for 2014+2015 is 27.5% (422/1,547)  motorist-overtaking type fatal collisions.

It’s not in the table but the overall (all lighting conditions) Urban overtaking percentage is 20%. [ 32+61 / 236 + 234 ]
Rural overall is 35%. [39 + 36 / 122 +87 ].
Compare to newer (2014-16) presented at  pedbikeinfo.org webinars on motorist-overtaking rates of 24 and 38.5% respectively.

SELECT eBIKECGP, sUrbanRural, sLighting ,count(1) "count of overtaking by U/R & Lighting", i.Year FROM 
((( 2011_incident as i JOIN 2011_person AS p_bike ON (i.ST_CASE = p_bike.ST_CASE AND p_bike.eINJ_SEV LIKE ('Fatal%') AND p_bike.ePER_TYP IN ('Bicyclist', 'Other Cyclist') )) JOIN 2011_vehicle as vehicle ON (i.ST_CASE = vehicle.ST_CASE AND vehicle.VEH_NO=1 )) JOIN 2011_PBType AS pbtype ON (i.ST_CASE = pbtype.ST_CASE AND p_bike.PER_NO=pbtype.PER_NO))
WHERE eBIKECGP LIKE ('Motorist Over%') GROUP BY 1,2,3;

SELECT sUrbanRural, sLighting ,count(1), i.Year FROM ((( 2011_incident as i JOIN 2011_person AS p_bike ON (i.ST_CASE = p_bike.ST_CASE AND p_bike.eINJ_SEV LIKE ('Fatal%') AND p_bike.ePER_TYP IN ('Bicyclist', 'Other Cyclist') )) JOIN 2011_vehicle as vehicle ON (i.ST_CASE = vehicle.ST_CASE AND vehicle.VEH_NO=1 )) JOIN 2011_PBType AS pbtype ON (i.ST_CASE = pbtype.ST_CASE AND p_bike.PER_NO=pbtype.PER_NO)) GROUP BY 1,2;

SELECT count(1), i.Year "Count of Motorist Overtaking fatalities" FROM ((( 2011_incident as i JOIN 2011_person AS p_bike ON (i.ST_CASE = p_bike.ST_CASE AND p_bike.eINJ_SEV LIKE ('Fatal%') AND p_bike.ePER_TYP IN ('Bicyclist', 'Other Cyclist') )) JOIN 2011_vehicle as vehicle ON (i.ST_CASE = vehicle.ST_CASE AND vehicle.VEH_NO=1 )) JOIN 2011_PBType AS pbtype ON (i.ST_CASE = pbtype.ST_CASE AND p_bike.PER_NO=pbtype.PER_NO)) WHERE eBIKECGP LIKE ('Motorist Overtak%');

SELECT sUrbanRural,count(1) "count", i.Year FROM 
((( 2011_incident as i JOIN 2011_person AS p_bike ON (i.ST_CASE = p_bike.ST_CASE AND p_bike.eINJ_SEV LIKE ('Fatal%') AND p_bike.ePER_TYP IN ('Bicyclist', 'Other Cyclist') )) JOIN 2011_vehicle as vehicle ON (i.ST_CASE = vehicle.ST_CASE AND vehicle.VEH_NO=1 )) JOIN 2011_PBType AS pbtype ON (i.ST_CASE = pbtype.ST_CASE AND p_bike.PER_NO=pbtype.PER_NO))
GROUP BY 1;
SELECT count(1), i.Year "Count of all bicyclist fatalities" FROM ((( 2011_incident as i JOIN 2011_person AS p_bike ON (i.ST_CASE = p_bike.ST_CASE AND p_bike.eINJ_SEV LIKE ('Fatal%') AND p_bike.ePER_TYP IN ('Bicyclist', 'Other Cyclist') )) JOIN 2011_vehicle as vehicle ON (i.ST_CASE = vehicle.ST_CASE AND vehicle.VEH_NO=1 )) JOIN 2011_PBType AS pbtype ON (i.ST_CASE = pbtype.ST_CASE AND p_bike.PER_NO=pbtype.PER_NO));

(note to self: for some reason i mix up Solomon’s Curve and Smeed’s Law, and i also sometimes think it’s Sneed’s law, which although it related to traffic safety doesn’t have to do with the issue at hand. I also sometimes think it’s Sneed’s Law, which is completely incorrect — but now i can google for it! Smeed’s Law, named after R. J. Smeed… is an empirical rule relating traffic fatalities to traffic congestion as measured by the proxy of motor vehicle registrations and country population

Solomon U-shaped Curve ( #solomon )

Normally cited by those wishing slow-moving vehicles should never use any roadway, even though it is perfectly legal, is the Solomon u-shaped curve, the claim is typically something to the effect of it’s safest when all traffic travels at the same speed; ergo a vehicle must never go slower than speed of traffic and must not be on the roadway in the first place. A corollary to this line of thinking is the belief that posted speed limits are really minimum speed limits.

A USDOT paper, Synthesis of Safety Research Related to Speed and Speed Management 1998 has a good backgrounder on the theoretical basis of these ideas:

In a landmark study of speed and crashes involving 10,000 drivers on 600 miles (970 kilometers) of rural highways, Solomon (1964) found a relationship between vehicle speed and crash incidence that is illustrated by a U–shaped curve. Crash rates were lowest for travel speeds near the mean speed of traffic, and increased with greater deviations above and below the mean.

however:

Excluding these (vehicles entering or vehicles slowing to leave the roadway) crashes from the analysis greatly attenuated the factors that created the U–shaped curve characteristic of the earlier studies. Without vehicles slowing to turn, or turning across traffic, the investigators found the risk of traveling much slower than average was much less pronounced. Crash risk was greatest for vehicles traveling more than two standard deviation above the mean speed.

It’s also wasn’t clear to me how or if any of this has to do with urban or suburban streets — where by definition, traffic starts and stops for a variety of necessary reasons.

There was an article published Jan 2015 in fiverthirtyeight discussing these concepts as well as some background of NYC’s move to make the default city-wide speed limit 25mph, down from 30mph.

Traffic Safety and Human Behavior
There’s a updated 2017 edition of Traffic Safety and Human Behavior: Second Edition, edited by David Shinar where, of course, the Solomons’ curve issue is dealt with; from the chapter Speed and Safety:

… Solomon (1964) found that drivers who drove either significantly above or below the prevailing average traffic speed were more likely to have crashes than drivers who drove at speeds close to the average. However, most crash-involved slow drivers were turning at the time of the crash; and when turning vehicles were removed from the analysis only those driving at speeds significantly above the traffic speed remained over-invovled in crashes (Fildes and Lee, 1993)

He later goes into a finely detailed review of Solomon (rural only; data from the 1950’s, etc) as well as reviewing a plethora of research published in the decades since that attempts to correct Solomon’s short-comings.

Shintar also notes that speed, when self-reported by (presumably speeding) drivers involved in a crash tend to be below actual speeds because of “Stannard’s Law” which states “drivers tend to explain their traffic accidents by reporting circumstances of lowest culpability compatible with credibility (Aronoff, 1971)”

NTSB: Reducing Speeding-Related Crashes Involving Passenger Vehicles

In July 2017, the NTSB (National Transportation Safety Board) released a major study: SS1701 Reducing Speeding-Related Crashes Involving Passenger Vehicles. In discussing Solomon 1964, it echo’s Shintar “These studies generally provided consistent evidence that driving faster than the surrounding traffic increased crash involvement rates; the evidence was less conclusive with respect to driving slower than the surrounding traffic (Aarts and van Schagen 2006)”. The report goes on to question the concept of setting speed limits at the 85th percentile on roads that are not freeways.

Speed and Crash Risk

Speed and Crash Risk, irtad (part of OECD) issued a 2018 report on the issue echoing the same sentiments about mis-use of Solomon’s work.

National Motorist’s Association

The NMA, or those identifying with it, commonly spread propaganda based on mis-stating Solomon’s 1964 work; see e.g. links in this rebuttal (to this piece):
The National Motorists Association’s drive toward alternative facts,

Metro Orlando

Some slides and discussion of Mighk Wilson’s presentation and analysis of crash types vs. severity; Pragmatic Bicycle Drivers (group membership required):

…Summarizing: advocates are using exurban and rural overtaking fatalities to justify urban bikeways that aggravate turning and crossing conflicts on lower-speed roads where very few serious overtaking crashes are happening.

A couple other points: it’s erroneous to make this into a simplistic dichotomy of “overtaking versus right hook.” Drive-outs and left crosses are also relevant. Also, comparisons between “bike lane and no bike lane” are effectively comparisons between edge riding with a stripe plus a couple more feet of lane width, and edge riding without them. We don’t have data to compare bike lane versus lane control.

Using only fatalities as the primary rationale for design is a bad idea; it should be broadened out to incapacitating (or serious) injuries. Fatalities are much more a function of speed, lighting, intoxication, and motor vehicle type than of bikeway design.

.