Category: safety

It used to be that SUVs were both more deadly to others, because of something dubbed poor “crash compatibility”, and not particularly safe (or perhaps i should say: not as safe as they could have been) for their own occupants due to a propensity to roll over; see this 2005 IIHS study that looked at 1999-2002 model years. It was a bit of a lose-lose proposition.

The latest version of looking at the risk of dying in any particular car, which covers model year 2005-2008, shows a marked decrease in SUV rollover deaths, presumably due to design changes in SUVs the most prominent being stability control “Recently calculated driver death rates for 2005-08 models show that drivers of SUVs are among the least likely to die in a crash. That change is due largely to ESC (Electronic Stability Control)”

Who is your Crash Partner?

Those studies look only at the risk of death to the driver of any particular vehicle — without regard to any other factors of the collision. It has long been known that SUVs pose a higher risk to others, because of their rigid frame design, which is also rides higher; in a collision with a car, particularly a t-bone, the rigid frame tends to slice into the car, disproportionately killing the car occupants. Happily, design changes made to SUVs have helped the sit

uation, to the point where similar weight vehicles, whether they are SUVs or cars, have similar risk of death.

“Whether you’re in an SUV or just sharing the road with one,” Nolan says, “recent improvements to these vehicles are making you safer.”

The results don’t contradict the basic physics of crashes. Size and weight are still key, and a small, lightweight vehicle is going to fare worse than a big, heavy vehicle in a crash. In general, SUVs and pickups are heavier than cars, so in that sense different types of vehicles always will be mismatched. But the study shows that, beyond weight, differences in vehicle styles don’t have to be a safety problem.

— IIHS, Effort to make SUVs, pickups less deadly to car occupants in crashes is paying off, news release 9/28/2011

 Pickups remain problematic, though even they have shown improvement.

What if your Crash Partner is a Pedestrian?

None of the above addresses this topic. Other studies have shown SUVs/Light Trucks are significantly more dangerous to pedestrians compared with automobiles: “Analysis of these three databases has clearly demonstrated that pedestrians have a substantially greater likelihood of dying when struck by an LTV (light truck or van) than when struck by a car.”  The fatality and injury risk of light truck impacts with pedestrians in the United States, Devon E. Lefler, Hampton C. Gabler, Accident Analysis and Prevention, v.36, pp. 295-304, Elsevier (2004)  (see also an earlier paper/version from the same authors sounded the alarm  did anyone pay attention, or even care?  The Emerging threat of Light Truck Impacts with Pedestrians is basically the same article)

Similar study published in 2005 Injury Prevention: United States pedestrian fatality rates by vehicle type by L J Paulozzi of the CDC, using 2002 FARS data “Compared with cars, the RR (relative risk) of killing a pedestrian per vehicle mile was 1.45 (95% CI 1.37 to 1.55) for light trucks… The greatest impact on overall US pedestrian mortality will result from reducing the risk from the light truck category”. This methodology is very straightforward, it takes the FARS data and segregates it by bodystyle (the paper does not state exactly how that was done; it looks easy, see below); and computes the RR (relative risk) based on Table VM-1, which is in Section V of FHWA Highway Statistics 2002. UNFORTUNATELY, the fhwa stopped reporting VM-1 in that way. Commencing with 2007 they no longer differentiate between passenger cars and LTVs; inexplicably they now differentiate by wheelbase, thus that data is useless for this purpose; so i guess we’ll never know how many more pedestrians are killed by LTVs (SUVs, pickups, etc). There are some footnotes to VM-1, saying methodology changes due to motorcycle reporting that do not seem to explain this change.

However, the data is all available in any Traffic Safety Facts Annual Report, e.g. here is 2011 and 2012 (search the library for newer ones). It is in Tables 7, 8, 9, 10 for Passenger cars, Light trucks, heavy trucks, and motorcycles. It lists VMT, and registration data; the only thing left to do is to extract from FARS the quantity of non-occupant fatalities split by those 4 vehicle types. I’ve already added a “synthetic” field to my FARS mysql data called sMODEL, it is based on the FARS field MODEL.

Here is a newer meta-study, that i would guess references the Paulozzi study and has very similar result, that i need to look up from Traffic Inj Prev. 2010 Feb;11(1):48-56. doi: 10.1080/15389580903390623.Do light truck vehicles (LTV) impose greater risk of pedestrian injury than passenger cars? A meta-analysis and systematic review. ” the risk of fatal injury in pedestrian collisions with LTVs compared to conventional cars was odds ratio 1.54, 95 percent confidence interval 1.15-1.93″

—

(given the dramatic change in the mix of the US vehicle (higher percentage of light trucks) fleet since whenever the cross and fisher data came from (mid 70s)…. it would be interesting to know if anything could be shown more statistically in, say, 2005.

There’s something called the “household” fleet, see exhibit 1 of the NHTS (Nat. Household Trans Survey)… mixture changed from 80/20 (automobiles/light trucks) to 50/50(!) from 1977 to 2008

There’s also this from the Cross and Fisher data (mid 1970’s or so):

TYPE OF MOTOR VEHICLE DRIVEN BY MOTORISTS IN THE FATAL AND NON-FATAL SAMPLES

[…]

Table 12 shows that trucks are involved in a proportionately greater number of fatal accidents (19%) than non-fatal accidents (9.4%). More than 80% of the trucks were pickups or vans; the remainder were larger types of trucks. These data suggest that the likelihood of fatal injuries increases as a function of the size of the vehicle. For instance, dividing the proportion of fatal cases by the proportion of non-fatal cases yields a ratio of .9 for passenger cars, 1.9 for pickups and vans, and 3.2 for larger types of trucks. However, because of the small number of cases involving a truck, these data can only be considered suggestive.

[Updated Jan 2015: There are more recent summaries posted at  phoenix.gov/streets/safety-topicss; direct links (unfortunately, many of these links now are 404 or link to the wrong report): bike (2010) ; ped(2012) ; all traffic(2012); bike (2013) (see section below for 2013 numbers) also an interesting MV collision rate study spanning 2006-2010.  I did a quick glance at the bike and numbers are very similar to the 2007 which is detailed below… including the tendency of Phoenix PD to mis-characterize collisions at driveways and crosswalks as the fault of the bicyclist ]
[Update 2023: I see a bicycle summary for year 2020 , see below, which was released in 2022; ]

Phoenix, and many other entities issue a report, usually called something like a Bicyclist Collision Summary. For some background, complaints, and links to others, see Understanding Collision Summaries.

At hand, I have the most recent, 2007, report from the City of Phoenix, which can be still be found here ( Local copy in case that link breaks) Continue reading “Listening to Phoenix’s Bicycle Collision Summary”

Here are some city documents:

• Detailed UPDATED drawing.
• Detailed original drawings.
• Presentation discussing the Diet.

The Diet

The diet part of the plan seems like a slam dunk… Normally any road diet is opposed becasue of fears that the lane removal will increase automobile congestion. In this particular case, that isn’t possible because of the unusual circumatance that this 1-mile stretch of 3 through lanes in each direction, is bounded on both ends by 2 through lanes. I.e. both north of Bethany Home Road, and south of Camelback Road is already only two lanes.

The Buffered Bike Lane

The “problem” then became what to do with “extra” space? A generously wide bike lane, including gutter is only 6′ wide, and the diet meant that 12′ of space had to be filled (in both directions). The answer came in the form of placing a 6′ buffer between the bike lane and the rightmost traffic lane. A.k.a a Buffered Bike Lane, see e.g. nacto.org.

I am somewhat skeptical of placing space between cyclists and overtaking vehicles. While this is presented as an un-alloyed good thing by many facilities advocates, it clearly has safety drawbacks which usually go unmentioned. here is a more balanced view, as presented in the Feb 2010 (the latest) Draft AASHTO Guide, p.78 (my emphasis):

Striped buffers may be used to provide increased separation between a bike lane and another adjacent lane that may present conflicts, such as a parking lane with high‐turnover or a higher speed travel lane. The benefits of additional lateral separation should be weighed against the disadvantages; a buffer between the bike lane and the adjacent motor vehicle travel lanes places cyclists further from the normal sight lines of motorists, who are primarily looking for vehicles in the normal travel lanes, and buffers between the travel lane and bike lane reduce the natural “sweeping” effect of passing motor vehicles, potentially requiring more frequent maintenance.

 That all being said, I objected to the original design which called for the outer buffer stripe to gradually arc into the intersection. This seemed to me to be a recipe for extra right-hooks. City staff readily agreed to my and Gene’s suggestion to end the buffer ahead of each intersection, and then a bit of dashed line; which is incidentally, as shown in the NACTO guide as recommended. (so thanks to Kerry Wilcoxon, and Joe Perez).

This should make the buffer “not bad” at intersections, yet doesn’t do anything for the many driveways. In other words, it should be no worse than a standard bike lane at intersections, but I fear it will raise risks at driveways relative to bike lane. So anyway, I’m reserving my judgement on the whole buffered bike lane thing. The hope is that it will encourage/entice cyclists off the sidewalks, where most collisions occur. However that doesn’t help those of us who are already legally using the roadway, and in fact may well be putting us in more danger.

The re-striping

The striping project apparently happened on schedule 4AM Saturday morning 1/7/2012, there are some pics on P4’s Facebook page (f.b. login required to view). TBAG has listed a ride to visit the new work on 1/8/2010.