Advanced Search

Please click here to take a brief survey

Safety on the roads: transit vs. car
Jeremy Faludi, 24 Jun 05

Cascadia Scorecard has a great article on the relative safety of different modes of transportation. They say that in the US, light rail is about 80 times safer to use than driving; buses are ten times safer; driving downtown is safer than driving in the burbs; and unfortunately bicycling is ten times more dangerous than driving. The good news, however, is that this last fact is due to bad design/policy: in Germany cycling fatalities have plummeted to a fraction of what they were 30 years ago.

How do bike-friendly places get that way? According to an article in Pittsburgh's Post-Gazette profiling bike-friendly cities, you need "good facilities for bicycling, an urban design oriented to people and not automobiles, traffic restrictions in residential neighborhoods, stricter enforcement of traffic regulations, better traffic education for motorists and nonmotorists, and restrictions on automobile use."

Bookmark and Share


Safety arguments are tough to make because often times statistics have very little to do with personal risk and they ignore the individuality inherent in risk assessment.

Risk in movement usually boils down to two things - physics and skill. The reason that planes are the "safest" mode (or one of the safest) is that you have a variety of actors with relatively high skill (pilots, air traffic controllers, professional mechanics, as well as strict regulatory oversight) driving a large number of passengers in relatively wide open space (both vertically and horizontally). The only drawback is that risk isn't a smooth continuum - a major failure usually leads to a complete loss.

Pedestrians, bicyclists, and motorcyclists all have higher risks than automobiles because of the physics involved with exposure (skin versus metal), as well as huge mass/momentum differences.

You can't just write off the higher danger of bicycle transport relative to automobiles just because of "bad design/policy" -- the same could be said of cars. We still have relatively weak penalties for drunk driving in this country, and consequently we still get about 40% of auto fatalities having alcohol consumption as a factor. Speed limits are also a factor of policy, as is road design, safety features, and most importantly, driver skill requirements (or lack of).

If people think cars are dangerous, I want someone to find an example of a professional race car driver being killed in an auto accident on a public road by another driver, and where the race car driver wasn't driving recklessly. I keep thinking there must be an example of it, but I can't find it.

The point is that skill plays a major part in safety.

And getting back to the psychology of safety, and specific risk to an individual, that's going to vary based upon whether people fear large-scale risk (eg, sarin attacks on subways), individual assaults within a vehicle, and so on.

As long as all modes share the same horizontal plane, and we continue to have a huge number of motor vehicles driven by low-skill drivers, then it's just going to be a bloody mess all the way around.

Posted by: Joseph Willemssen on 24 Jun 05

I'm getting a little science fictional here but, maybe the safety issues Joseph points out can be used as an argument for the automation of vehicle piloting.

If robots piloted our cars, there'd be no drunkeness, no road rage, no drowsiness, no distraction and vehicle maintenence would improve due to better driving habits or at least ones without human variation.

Of course this wouldn't really make things safer for the cyclists.

Posted by: Mr. Farlops on 24 Jun 05

I'm getting a little science fictional here but, maybe the safety issues Joseph points out can be used as an argument for the automation of vehicle piloting.

If robots piloted our cars, there'd be no drunkeness, no road rage, no drowsiness, no distraction and vehicle maintenence would improve due to better driving habits or at least ones without human variation.

Of course this wouldn't really make things safer for the cyclists.

This harkens back to our discussion about light rail, buses, cars, PRT, etc. A huge advantage of automated systems is that there can be much greater throughput and safety compared to automobiles.

The Feds have been funding research for at least a decade with automobiles that would allow them to drive almost bumper to bumper at 100 mph or more. If you take that kind of technology and apply it to microrail, along with smart routing, "lattice pie" intersections, one ways, and the like, then you'd see much greater flow and speed and far greater safety.

Think of the effect of one person's carelessness on traffic in a metro area if it results in a simple accident on the interstate. Add in people's propensity to tailgate, switch lanes too much, own high profile vehicles limiting visibility for other vehicles, and so on, and it just is a very poor way to get people to move around. That doesn't even factor in the slow reaction time of human beings and our tendency not to look far enough ahead and flow properly with other drivers.

Posted by: Joseph Willemssen on 24 Jun 05

As a Heath, Safety, and Environment consultant, I have been following the safety issue associated with Light Rail for a couple of years now. It is important to understand the "total" safety issue when discussing Light Rail safety. Indeed, light rail transit system can be a very safe when it is isolated from street traffic. When you look at the total statistics from all light rail systems and merge the data, you can get a false understanding about light rail safety. When light rail systems share an on-grade right-of-way with street traffic (trucks, buses, automobiles, cyclists, pedestrians, emergency vehicles, etc.) the "total" safety of the system can be very poor. Sure, the passengers of the train car may not get injured, but when the light rail car hits someone else, risk of serious injury is great.

So, when I read Jeremy Faludi's post, I found it to be misleading. In going to his "Cascadia Scorecard" source, I found the full text of what they said. It is as follows:

"If you want to be safer, take transit.  Measured per passenger-mile, transit buses and commuter rail are about as safe as you can hope for; buses seem to be more than 10 times safer than cars for their occupants, while commuter rail is about 80 times safer.  (But both are more dangerous than cars for other occupants of the roads.)"

Why Jeremy Faludi did not give a more complete "safety" account, I do not know. However, I do know that when the advocates for a light rail system in Phoenix, AZ tried to persuade the voters to approve a tax to pay for their new light rail system, they avoided the "total safety" issue because the system they wanted was going to be on the street level. Such systems are inherently unsafe because a risk situation will exist every time a light rail car passes through an on-grade intersection with cross-streets. Collisions will happen, people will be killed. The Phoenix light rail planners knew this but they wanted the ambiance of a "streetcar" mass transit system in the downtown "urban" renewal project. So, when the light rail planners in Phoenix talked about safety, they too combined safety figures for all light rail systems and then reported on the safety of the passengers in the light rail cars. Total street traffic system safety was not a primary concern when picking a mass transit system design.  Such intentional deceit is often a sign of government corruption.

Posted by: Gary Green on 25 Jun 05

Gary Green has also ignored a very telling point: that automobiles, buses, bicycles, trucks, and motorcycles all register very high danger points at intersections, and when making turns and lane changes. Collisions, injuries, and deaths resulting from these situations far and away exceed not only the absolute number attributable to light rail, even on surface street lines (FTA and DOT statistics bear this out immutably and inargueably) but, moreover, the injury/death rate per passenger-mile on LRVs is far better than the rate attributable to any automobile/motorcycle/bicycle conveyance. Should we not, then, act to reduce this horrific situation by reducing the number of vehicle miles operated by these forms of transport in favour of demonstrably safer public transport (bus, rail, and air)? Any of these are less subject to poor decisions made by badly-trained, poorly-skilled vehicle operators that end in grevious results, most times involving innocent parties. Moreover, most incidents involving commercial vehicles, whether passenger or freight, are initiated by actions of those same badly-trained, poorly skilled, and often distracted or impaired operators rather than the commercial operators (also a matter of public record).

Tom Fairbairn
Minneapolis, MN

Posted by: Tom Fairbairn on 26 Jun 05

So, Gary, you're right that I edited, but I don't think it was misleading; it was just the one-sentence summary. When linking to stories from other sources, there's no sense in rewriting the whole thing, and Cascadia Scorecard did the research, so they should get the readers.

Joseph, you're right that skill plays a big part in accident avoidance, but let's be honest, the average driver must be assumed to have very low skill. It's a lowest-common-denominator thing, and it's also a probability thing--every time you walk out the door, you're rolling the dice to some extent, and your transportation modality basically determines what kind of dice you're rolling.

Saying it's not a design problem but a skill problem is implying that driver-licensing should be a much stricter system, requiring extensive training and constant re-certification, perhaps in-car breathilizers or reflex-testers that must be passed before the car will turn on. How far down that path do you want to go? How far do you think you have to go to solve the problem entirely?

I think it's an interesting idea, but even when people are experts, they still make stupid decisions when they're distracted or rushed or tired. The most reliable way to ensure safety is by changing the physics of the situation, as you described it, which means bike lanes separated by a curb (or, better, pedestrian & bike-only paths), reduced traffic speeds, etc., which are all urban design and policy tools.

Posted by: Jeremy Faludi on 26 Jun 05

Where in the argument do we ask: who is the person who is creating the danger, relative to some system of rules?

We then segregate the mishaps into those where the injured party created the danger versus those mishaps where some other party created the danger.

Then we can say more accurately what modes of travel are more dangerous.

Posted by: Allan Jayne on 27 Jun 05

" Saying it's not a design problem but a skill problem"

Whoa - I didn't say that. I said that it's both, and that risk is individual-specific in many cases, so it's inaccurate to apply general safety statistics to individuals.

Posted by: Joseph Willemssen on 27 Jun 05



MESSAGE (optional):

Search Worldchanging

Worldchanging Newsletter Get good news for a change —
Click here to sign up!


Website Design by Eben Design | Logo Design by Egg Hosting | Hosted by Amazon AWS | Problems with the site? Send email to tech /at/
Architecture for Humanity - all rights reserved except where otherwise indicated.

Find_us_on_facebook_badge.gif twitter-logo.jpg