Here’s How Much Safer Transit Is Compared to Driving

Traffic fatality risk by transportation mode. Image: Journal of Public Transportation
Traffic fatality risk by transportation mode. Image: Journal of Public Transportation

Keep this in mind the next time a high-profile train crash generates more press coverage than a year’s worth of car wrecks: Despite the media sensationalism and overwrought regulatory responses that follow such events, transit is already a lot safer than driving.

Looking at traffic fatalities per mile traveled in the U.S., analyst Todd Litman found that riding commuter or intercity rail is about 20 times safer than driving; riding metro or light rail is about 30 times safer; and riding the bus is about 60 times safer. Factoring in pedestrians and cyclists killed in crashes with vehicles, the effect is smaller but still dramatic: the fatality rate associated with car travel is more than twice as high as the rate associate with transit. Litman’s study was recently published in the Journal of Public Transportation [PDF].

Litman notes that most transit travel involves some walking or biking, which carry a relatively high risk of traffic injury. But those risks are mostly offset by the health benefits of physical activity. Living in a place with good transit has safety benefits as well: Litman cites research showing that cities with higher transit ridership rates tend to have lower per-capita traffic fatality rates.

Using FBI data, Litman also busts the myth that transit is linked to high levels of crime. While direct comparison is difficult because transit riders and drivers are susceptible to different types of crimes (transit riders are more likely to encounter assault and property theft, while drivers see more to road rage incidents, vehicular assault, and auto theft), Litman shows that on balance, people riding transit are less likely to be victimized than car drivers, passengers, and owners.

While much transit service operates in low-income communities with relatively high crime rates, Litman writes, relatively few crimes occur on transit property. In fact, when you normalize for exposure, owning a car and making driving trips is riskier than riding transit, Litman finds. Litman says transit facilities tend to have low crime risk because of there are so many other people around keeping an eye on things: employees, passengers, and bypassers.

“The greatest risks occur when passengers walk and wait in isolated areas, but these risks are no greater than what motorists encounter walking to and from isolated parking lots,” he writes.

Furthermore, the types of property thefts transit riders usually encounter — a stolen wallet or phone — generally incur much less expense than having your car stolen or vandalized. The average car theft costs about $6,000, according to Litman.

In media coverage and in transit agencies’ own public messaging campaigns, transit is often linked to the threat of terrorism, but internationally, Litman notes, about 360 times more people are killed in auto collisions than in incidents of terrorism.

Litman concludes that transit agencies should make the safety of bus and rail travel more of a key selling point, instead of broadcasting messages like the “If you see something, say something” campaign that end up contributing to a heightened sense of risk.

  • Adam Herstein

    From that graph, it seems that commuter rail is almost as dangerous as driving. Also, none of the numbers (e.g. riding metro or light rail is about 30 times safer) match the graph. In my example, the graph shows 2 vs 8, or 4 times.

    How did you come to these conclusions based on the graph?

  • Bolwerk

    It probably is, to people outside the vehicle. Some, maybe most, of that is presumably suicides and grade crossing accidents though.

  • Adam Herstein

    If you’re evaluating the safety of a transportation mode, both internal and external users need to be considered. If a portion of the deaths is suicides, then that should be reflected in the graph.

    Claiming that buses are safe because riders are safe, while people are getting run over by buses is not a valid claim. It’s like claiming guns are safe because the shooter is safe, while 90% of deaths are by people the gun is pointed at.

  • BBnet3000

    Litman concludes that transit agencies should make the safety of bus and rail travel more of a key selling point, instead of broadcasting messages like the “If you see something, say something” campaign that end up contributing to a heightened sense of risk.

    Definitely the important message here, and probably understated by using mileage rather than per trip.

  • As Bolwerk mentioned, and as I can confirm, most commuter rail deaths are suicides by people standing on the tracks. They clearly ARE including those in the above graph, but blaming trains for that is obviously silly. The fact that they’re appealing ways to kill yourself in this country doesn’t have a whole lot of policy relevance, anymore than it should mean we need to stop building bridges because people might jump off of them (another relatively popular suicide method).

  • Bolwerk

    I think you’re overreacting to the safety “threat” here. You can’t make safety issues zero, but these are deaths per billion passenger-miles. These are also fatalities, no other casualties like maiming or expensive medical care or economic cost (probably all highest for the passenger car mode).

    And I have my doubts any measure discrete risk. Trucks and buses are dangerous to pedestrians making right turns. Heavy rail transit keeps outsiders off tracks, so most exposure is in stations. Drivers seem likely to kill themselves, and trucks probably have the added dangerous of crushing cars.

    Then there is a question of methodology. Does a train hitting a car count against the train mode, the car mode, or both? Or does it depend on who is at fault (if the train can be at fault)?

  • davistrain

    “Safety” as a selling point, whether for automobiles or transit vehicles ranks rather far down the list of desirable features. Convenience, comfort, travel time and to some extent, snob appeal are more important to the general public.

  • davistrain

    And usually when people are thinking about “safety”, they are looking at it from the “user” point of view. The old “I’d rather be a hammer than a nail” idea.

  • Chris

    Are grade crossing incidents counted towards the commuter rail numbers, or the cars that violated the right of way? Still seems odd that “Commuter Rail” is so much worse than heavy rail.

    What is “Commuter Rail” anyway? Commuter Trains tend to be Heavy Rail in US terminology. Or is “Commuter Rail” used for Light Rail / Street Cars / Subways here?

  • Bolwerk

    It probably is grade crossing accidents. Counted against only one mode or not, a car-on-train collision is proportionately going to impact rail numbers more. Car passenger-mileage must so overwhelm train mileage that it’s easy for something that’s scarcely a statistical blip for automobiles to be a huge effect for commuter rail.

    They seem to use “heavy rail” as shorthand for what is properly called rapid transit – the big, grade separated electric urban rail transit with frequent service (e.g., Chicago El, DC Metro, NYC Subway). Commuter rail is “heavy rail” too, but it’s the suburban rail service on mainline rail track. They don’t seem to count light rail/streetcars/trams at all.

    Commuter rail is exposed to grade crossings, while “heavy rail” rapid transit is not – this probably explains the difference.

  • Dan

    Maybe heavy rail is intercity rail (Amtrak) and commuter rail is for suburban regional commuter rail. If so, Amtrak could be lower because intercity trains run more miles in rural areas with fewer potential conflicts with passenger vehicles.

    Seems like any incident that involves a train is counted under this category. In reality, most commuter rail fatalities are caused by suicides, or by drivers that try to cross when they shouldn’t. If driver error causes the crash those fatalities should be counted as passenger car fatalities.

  • Bolwerk

    That occurred to me too, but then they seem to lump Amtrak and commuter rail together in another statistic (“Table 2”). They also lump rapid transit and light rail in together on the same table.

    Near as I can tell, they don’t define exactly what they mean by any of those terms, and AFAIK they’re all a bit vague. And the categories are a bit arbitrary, since light rail arguably has more in common with urban transit buses in terms of danger and exposure than it does with heavy rail. Light rail could be fully grade-separated, but rapid transit is by definition grade-separated.

    It makes the point that transit is safer than cars well enough, but it doesn’t do much else.

  • Horseswaggled

    The simple railroad safety formula

    100 million divided by total train miles. That number times number the
    railroad murdered during the same period. Pretty much shows the
    railroad idiots outrunning their sight stopping distance are at fault.

    State – All States County – All
    January To July, 2014
    Total fatalities: 488
    Total train miles: 441,055,279
    +++++++++++++++++++++
    SELECTION: Railroad – Amtrak [ATK ]
    Total fatalities: 69
    Total train miles: 23,669,936
    Amtrak kills 292 people in 100 million miles NOT 1.
    ++++++++++++++++++++
    BNSF Rwy Co. [BNSF]
    Total fatalities: 53
    Total train miles: 108,586,539
    BNSF kills 49 people in 100 million miles NOT 1.
    +++++++++++++++++++
    Buffalo & Pittsburgh RR, Inc. [BPRR]
    Total fatalities: 1
    Total train miles: 266,196
    Buffalo & Pittsburgh RR kills 376 people in 100 million miles NOT 1.
    ++++++++++++++++++++
    Caltrain Commuter Railroad Company [PCMZ]
    Total fatalities: 5
    Total train miles: 824,636
    Caltrain Commuter Railroad Company kills 606 people in 100 million
    miles NOT 1.
    ++++++++++++++++++++
    Canadian National – North America [CN ]
    Total fatalities: 17
    Total train miles: 12,090,848
    Canadian National – North America [CN ] kills 141 people in 100
    million miles NOT 1.
    ++++++++++++++++++++
    Canadian Pacific Rwy Co. [CP]
    Total fatalities: 3
    Total train miles: 6,947,821
    Canadian Pacific Rwy Co. [CP] kills 43 people in 100 million miles NOT
    1
    +++++++++++++++++++++
    CSX Transportation [CSX ]
    Total fatalities: 58
    Total train miles: 58,168,604
    CSX Transportation [CSX ] kills 100 people in 100 million miles NOT 1.
    +++++++++++++++++++++
    Kansas City Southern Rwy Co. [KCS
    Total fatalities: 6
    Total train miles: 6,655,815
    Kansas City Southern Rwy Co. [KCS] kills 90 people in 100 million
    miles
    NOT 1.
    ++++++++++++++++++++
    MARC Train Service [MACZ]
    Total fatalities: 1
    Total train miles: 767,058
    MARC Train Service [MACZ] kills 130 people in 100 million miles NOT 1.
    +++++++++++++++++++
    Massachusetts Bay Transit Authority [MBTA]
    Total fatalities: 5
    Total train miles: 2,485,094
    Massachusetts Bay Transit Authority [MBTA] kills 201 people in 100
    million miles NOT 1.
    +++++++++++++++++++
    Norfolk Southern Corp. [NS ]
    Total fatalities: 73
    Total train miles: 55,592,180
    Norfolk Southern Corp. [NS ] kills 131 people in 100 million miles NOT
    1.
    ++++++++++++++++++
    Northeast IL Regional Commuter Rail Corp. [NIRC]
    Total fatalities: 12
    Total train miles: 4,359,210
    Northeast IL Regional Commuter Rail Corp. [NIRC] kills 275 people in
    100 million miles NOT 1.
    +++++++++++++++++++
    Pan Am Rwys/Guilford System [GRS]
    Total fatalities: 1
    Total train miles: 1,065,586
    Pan Am Rwys/Guilford System [GRS] kills 94 people in 100 million miles
    NOT 1.
    ++++++++++++++++++++
    South Florida Regional Transit Authority [SFRV]
    Total fatalities: 3
    Total train miles: 664,799
    South Florida Regional Transit Authority [SFRV] kills 451 people in
    100
    million miles NOT 1
    ++++++++++++++++++++
    Trinity Rwy Express [TREX]
    Total fatalities: 2
    Total train miles: 264,364
    Trinity Rwy Express [TREX] kills 757 people in 100 million miles NOT
    1.
    ++++++++++++++++++++
    Union Pacific RR Co. [UP ]
    Total fatalities: 106
    Total train miles: 101,183,089
    Union Pacific RR Co. [UP ] kills 105 people in 100 million miles NOT
    1.
    +++++++++++++++++++
    Virginia Rwy Express [VREX]
    Total fatalities: 1
    Total train miles: 214,212
    Virginia Rwy Express [VREX] kills 467 people in 100 million miles NOT
    1.++++++++++++++++++++

  • Heavy rail usually means the type of trains that run in subways like in NYC and on els like in Chicago. I don’t know if it’s a standard, but they tend to be powered by a third rail while light rail tends to have the overhead catenary system.

    Commuter rail is longer-distances, also “heavy” typically I believe but used for travel from the suburbs into the city, for the most part. It’s often called regional rail.

    As far as I can tell, this graph doesn’t include light rail, which is what many new rail systems use (examples include Seattle’s line, all of Los Angeles’ lines except the two subway routes, Denver, Salt Lake City, etc.).

  • Jacob Lynn

    You seem confused — the claim under discussion is deaths per PASSENGER-mile, not deaths per mile. There are many people on board the typical train, who would otherwise likely be in many individual vehicles.

  • Jacob Lynn

    The sentence is describing relative danger to the user of the transit mode (that is, it’s referring to the blue parts of the bar chart). The yellow bars describe danger to people outside the vehicle (pedestrians, etc). Agreed that suicide deaths should be more clearly delineated, and that urban buses are more dangerous to other users than they ought to be.

  • Kenny Easwaran

    I’m guessing that the difference between “commuter rail” and “heavy rail” is that “heavy rail” is about subways and els, while commuter rail is FRA-compliant rail on freight tracks. (That leaves light rail and streetcars entirely uncounted though.)

    At any rate, for the fatalities to others, it’s not completely clear whether the right metric is per passenger mile, per service hour, per route mile, per crossing, or what. If commuter rails run mostly empty, then “per passenger mile” is an overcount.

    Furthermore, this graph understates some things, because it only counts fatalities, and not all the social and personal and economic costs involved with all the measures taken to mitigate fatalities.

  • ahwr

    Passenger miles wouldn’t overcount an empty train, seat miles would.

  • Joe R.

    Unless the crossing gates are broken, any car-train collision shouldn’t count against trains. Same thing with suicides or people killed trespassing on the tracks.

  • ahwr

    What’s the reasoning there for non suicides? If a pedestrian was crossing outside of a crosswalk or against the light and gets hit should it count “against” cars? What if a a driver speeds around a curve ignoring posted warnings and smashes into a tree? What if a pedestrian tries to run across the street, either trips or misjudges the timing and gets hit by a light rail vehicle? Does that count as “against” surface rail?

  • Joe R.

    Generally safety statistics for any mode of transport only include deaths/injuries for people using that mode. That’s why I think it’s peculiar that grade crossing incidents are counted against trains, especially when there’s literally nothing a train engineer can do to avoid a motor vehicle crossing in front of them (or a person attempting suicide)..

    There may be some gray area when we’re talking about different modes which legally share the same space. In that case, we might attribute pedestrian or cyclist deaths caused by driver error to motor vehicles. There is generally no such gray area with rail vehicles other than light rail running on surface streets. By definition, people and motor vehicles aren’t allowed on railroad tracks except at designated crossings, and then only when the gates are up. As such, unless the gates are malfunctioning, any deaths outside the train should be counted 100% against either motor vehicles or pedestrians. Same thing with pedestrian deaths on limited access highways. Those shouldn’t count against motor vehicles.

    And by a similar line of reasoning 100% of pedestrian deaths on sidewalks caused by curb jumping vehicles can be counted against motor vehicles as motor vehicles are not legally allowed to drive on sidewalks (except at driveways where they must yield to pedestrians).

  • ahwr

    Generally safety statistics for any mode of transport only include deaths/injuries for people using that mode.

    Before responding to this it would help if we are on the same page, so what’s the point of the statistics? What are they being used to show, or to judge?

  • Joe R.

    You do know that most of those fatalities are people in motor vehicles where the driver choose to drive around crossing gates? As such, I don’t even know why they’re counted against trains when they should really be counted against motor vehicles. Here is more accurate information:

    http://pedestrianobservations.wordpress.com/2011/06/02/comparative-rail-safety/

    Note that in the US there were 159 deaths over 20 years. That’s why I thought your statistics looked suspect. They include grade crossing incidents which unfairly increase the numbers.

  • Joe R.

    Safety statistics show the likelihood of death or injury when using a particular mode of transport. That’s very useful information to someone who might want to know what is the safest way to travel. They’re not meant to be all inclusive. If you think about it, motor vehicles indirectly kill more people than they kill via pollution and other negative externalities. And other modes of transport have their own negative externalities as well. You can also keep statistics on these things to some extent, but they’re completely separate from the general safety statistics. Remember in multimodal collisions in order to attribute which mode to count against you first need to see who was at fault. Often this is impossible to determine, so statistics based on this would be highly suspect. It’s much more clear cut doing it the way it’s been done-if someone dies inside a vehicle, that counts against that particular type of vehicle. If the vehicle kills or injures people outside the vehicle, that doesn’t count.

  • Joe R.

    Given the magnitude of the numbers, it’s pretty apparent they’re counting grade crossing incidents against commuter rail for reasons I can’t even begin to fathom. Those should indeed be counted against the motor vehicles which violated the right-of-way. Apparently they have an agenda which benefits by making trains look much worse than they really are.

  • Horseswaggled

    Who cares. Fact is passenger trains with no brakes are more out of control than a drunk 4x over the limit. Well over 1/2 of grade crossing fatalities are pedestrians.

  • Joe R.

    Passenger trains have brakes.

    If someone decides to drive or walk across the tracks when the gates are down, or the lights are flashing, how exactly is that the fault of the train which hits them?

    Trains are very safe as a way to travel. They would also be very safe to those outside them if people just wouldn’t go where they don’t belong. It’s not like a train can come barreling down a sidewalk full of people without any notice. When someone is hit by a train, it’s because they made the choice to put themselves in harm’s way. It’s very clear to anyone with the cognitive function of a preschooler what railroad tracks look like, what they are used for, and why people don’t belong on them.

    What exactly is your point of posting those loaded statistics other than to make trains look much worse than they really are?

  • ahwr

    Safety statistics show the likelihood of death or injury when using a particular mode of transport.

    So pedestrian deaths don’t count against any mode then? If you look at how these statistics are kept, even how the data is presented in the chart above there are two categories for each mode. “Users” – a passenger on a train or bus, a truck driver, a car driver etc…and “others” – someone in a car hit by a train or truck, a pedestrian hit by a tram or car etc…

    They’re not meant to be all inclusive.

    These statistics ignore emissions and other negative externalities that are harder to deal with, but they are meant to include all fatalities that result from collisions. Whether between a car and a pedestrian, a tram and a biker, a CR train and an SUV etc…

    Remember in multimodal collisions in order to attribute which mode to count against you first need to see who was at fault.

    No. The headline figures being thrown out here and elsewhere do not make a distinction.

    That’s very useful information to someone who might want to know what is the safest way to travel.

    A substantial share of MV fatalities are drunk drivers, or passengers of drunk drivers. Another huge chunk are single vehicle crashes where the driver wasn’t paying attention and flew off the road. Lumping all MV fatalities together makes it look more dangerous than it is for someone who pays attention while driving a well maintained vehicle.

    Is this about control? A train on tracks has a given set of behaviors, others are expected to work around that, so assuming the train operator doesn’t deviate from what’s expected and there are no mechanical failures pedestrians and drivers can safely navigate around the train in a way others can’t with cars, buses, or trucks? True, but the numbers thrown around in this article aren’t fine grained enough to tell you that. It would be hard to put something like that together because in a train or on a bus you have little control, you are mostly* dependent on others for your safety. In a car you play a much larger role, so the capability of the individual greatly affects the safety of the mode. You bike a lot, you know how to keep yourself safe cycling right? A headline number for bike fatalities per mile wouldn’t be all that applicable to you would it?

    * Heavy rail fatalities generally include stuff like a rider falls down the stairs in a station and hits their head and dies, or slips onto the tracks just as a train pulls into the station – so you do have some control as the user there I guess.

    If you think about it, motor vehicles indirectly kill more people than they kill [via crashes] via pollution and other negative externalities

    For how long will that remain true, assuming it still is? Emissions have been cut substantially and can be cut further, even continuing to use ICEs. Emissions from a PZEV fleet would lead to how many deaths?

  • ahwr

    Well over 1/2 of grade crossing fatalities are pedestrians.

    Are you still talking about mainline rail, as in your first post?

    http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/national_transportation_statistics/html/table_02_39.html

    In 2012 the grade crossing fatality split was 175 motor vehicles and 97 non motor vehicles.

  • Joe R.

    The premise of this article is simply to show transit is safer than driving even if you take a devil’s advocate position and include those hit by the vehicle in the statistics. It’s not meant to advocate changing the way safety statistics are kept.

    No. The headline figures being thrown out here and elsewhere do not make a distinction.

    Which is exactly why they’re worthless as anything other than talking points. You certainly wouldn’t use these numbers to make policy.

    Lumping all MV fatalities together makes it look more dangerous than it is for someone who pays attention while driving a well maintained vehicle.

    Except there’s that element of random chance. The world’s best driver can still die if a drunk driver decides to do something like drive up a highway entrance or exit ramp in the wrong direction.

    You bike a lot, you know how to keep yourself safe cycling right? A headline number for bike fatalities per mile wouldn’t be all that applicable to you would it?

    In a general sense, no it wouldn’t, but I’m under no illusion that I’m 100% safe while riding even doing everything right. There’s still that random drunk driver who might hit me from behind at high speed who I’ll never see coming in enough time to react to them.

    Is this about control? A train on tracks has a given set of behaviors, others are expected to work around that, so assuming the train operator doesn’t deviate from what’s expected and there are no mechanical failures pedestrians and drivers can safely navigate around the train in a way others can’t with cars, buses, or trucks?

    A train can be made inherently nearly 100% safe, including the use of failsafes to account for human error on the part of the train driver. Grade separate the railway tracks at all crossings, and other users are mostly safe as well unless they make a bunch of deliberate decisions to violate the right-of-way. No other mode can claim this. That’s the difference. With other modes, a typical human error can result in death. With trains, it often takes deliberately, consciously putting yourself in harm’s way.

    For how long will that remain true, assuming it still is?

    For as long as it takes cancers to metastasize in people already exposed to motor vehicle exhaust. Even if we had a ZEV fleet tomorrow, we would still be seeing a gradually decreasing number of deaths from motor vehicle emissions for another 50 years.

  • ahwr

    A train can be made inherently nearly 100% safe, including the use of
    failsafes to account for human error on the part of the train driver.

    What about human error on the part of those not in the train? What about human error on the part of train passengers?

    No other mode can claim this.

    So? Is what is technically possible so important, or what would be financially feasible? Besides, give it another decade and you could say the same about cars, buses, and trucks.

    Except there’s that element of random chance. The world’s best driver
    can still die if a drunk driver decides to do something like drive up a
    highway entrance or exit ramp in the wrong direction.

    You missed the point. I didn’t mean that driving is 100% safe for a sober and alert driver who pays attention, his brakes could fail as he goes into a tight turn too. My point was that these statistics don’t tell you the risk for that driver. They don’t tell you the risk for a drunkard, or a teen with little experience and a bunch of immature friends in the car either. The risk is highly variable and dependent on the driver, his state of mind etc…The risk for transit passengers is much more uniform, though not completely so, it doesn’t make much sense to compare the two, if as you said, you want “information [for] someone who might want to know what is the safest way to travel,” since it depends so much on the person travelling.

    The premise of this article is simply to show transit is safer than driving

    For who? For an experienced driver? Maybe, I don’t know. The data here isn’t fine grained enough for you to make such a determination. Maybe it’s close and would be dependent on time of day, the vehicle being driven/which transit agency or vehicle on the other side, what corridor etc…For a teenager with a bunch of idiot friends in the car distracting him, and in a car without modern safety features? As a class, sure, but the data used isn’t fine grained enough to tell you that, and even if it was it isn’t obvious that it would apply to all members of that class.

    For the populations of bus/rail riders vs car/SUV drivers/passengers? Sure, but are you comparing comparable populations or trips? Not really, so it isn’t all that useful until you break it down into comparable groups.

    Which is exactly why [the figures a]re worthless as anything other than talking points.

    Yup.

  • Joe R.

    Statistics by definition can only be used for prediction in the aggregate, not in the individual case. Still, it gives some basis for making a rational decision, even if it can’t predict the exact outcome of that particular decision, only the average outcome.

    The risk for transit passengers is much more uniform, though not completely so, it doesn’t make much sense to compare the two, if as you said, you want “information [for] someone who might want to know what is the safest way to travel,” since it depends so much on the person travelling.

    You have something else called the standard deviation. This is basically how much variation there is from the average. Sure, there might be higher standard deviation for driving versus taking transit due to the variability of driver skill levels. Complete statistics would also include the standard deviation. The ones above don’t.

    What about human error on the part of those not in the train? What about human error on the part of train passengers?

    To be clear, I’m talking here about deaths/injuries caused by train collisions. Those can be virtually engineered away with rail. They can’t be with other forms of transportation. Human error won’t cause a person to wander on to a grade-separated or fenced railway, for example. It would instead take a deliberate choice. Human error on the part of train passengers may result in falls or other accidents but this has little to do with the mode.

    So? Is what is technically possible so important, or what would be financially feasible? Besides, give it another decade and you could say the same about cars, buses, and trucks.

    It’s quite possible and feasible to engineer virtually 100% safe railways. We can’t do it with motor vehicles now or at any time in the future. The issue here is two degrees of freedom versus one. The train always goes along the same path. In fact, it’s constrained to by the nature of railways. Motor vehicles can wander anywhere. Loss of traction can mean a motor vehicle ends up in an unpredictable location. By definition that can mean injury or death. Self-driving vehicles won’t change this fact. Nor can self-driving vehicles avoid all external objects which might suddenly wander into their paths. To be fair, neither can a train, but a railway can be inherently designed to separate the train from everything else to make such an occurrence highly unlikely . That can’t be done for motor vehicles on city streets.

  • ahwr

    a railway can be inherently designed to separate the train from
    everything else to make such an occurrence highly unlikely . That can’t
    be done for motor vehicles on city streets.

    You can build Els/subways, you can also build freeways. Neither is cheap or feasible except on highly travelled corridors. Grade separating all mainline rail crossings would cost a significant sum, and isn’t feasible. Just putting in crossing gates isn’t financially feasible all the time. All trains can derail, especially around switches.

    Those can be virtually engineered away with rail. They can’t be with other forms of transportation.

    You can fence railways that run at grade. You can also fence roads, only allow crossing at fixed points, make the pedestrian wait for a light to change. Give all cars a red when you want pedestrians to cross. Worried someone will run the light? You can engineer that away in the near future. Worried a car will blow out a tire and fly through the fence? Put in a better barrier.

  • Bolwerk

    Well, break this out a bit: with FRA heavy rail, cars are supposed to yield at a grade crossing. Drivers should know this. Even a passenger train can take more than a mile to come to a complete stop.

    LRVs probably have to follow the same rules as buses in mixed traffic, so the rules of who can be at fault are probably similar.

  • Bolwerk

    I don’t think there is any attempt here to measure discrete risk. All we have is something of a rate of death by mode, not even a meaningful measure off risk. (Injuries appear to be ignored.)

    Strictly speaking, a grade crossing death is by both automobile and train. The judgment you make about it – fault – maybe can go either way. However, objectively, someone was killed and a train was involved. (Mayhap the train deserves a Nobel peace prize, if the victim was death-worthy enough.)

  • jbl

    Number of LIES in Pines post is 3

  • jbl

    Number of LIES in pines post is 3

  • dahszil

    The yellow part of the bar chart graph is pedestrian and other vehicles passengers killed. the blue part of the graph pertains to those killed travelling inside form of transportation. The commuter rail is almost completely yellow. meaning very few are killed using commuter rail. The car and light truck bar chart has a much larger number of people killed per miles traveled of those driving and riding in the car and light trucks(for example a pick up truck). Also if you just looked at the height of bar chart for heavy trucks(e.g. tractor trailers, semis. lorries they call them in the UK) is just an eighth of drivers and passengers killed compared to cars and pick up trucks. yes the big yellow part means commuter rail and even more so heavy truck are the biggest danger to pedestrians and motor vehicles. However many of these deaths are due to cars trying to go around red and white stop for train board, a lot of pedestrians try to cross rail tracks and get killed, and a significant percentage are suicided whether in car or on foot throwing yourself in from commuter rail or big tractor trailers on the truways.

ALSO ON STREETSBLOG

America Has a Terrible Traffic Safety Record Because We Drive Too Much

|
Even though the U.S. traffic fatality rate per mile driven has fallen by two-thirds in the last 50 years, America today still has the deadliest road system per capita in the developed world. Much of the improvement from safer driving and better emergency care has been wiped out by increases in total traffic. The American approach to traffic safety has emphasized seatbelt use, vehicle standards, […]

Deadly Year for European Rail Still Safer Than the American Average

|
Does the recent train derailment in Spain, which killed 79 people, justify America’s onerous approach to regulating rail safety? Federal Railroad Administration safety rules are designed to maximize “crashworthiness,” making U.S. passenger trains heavier and more expensive than their counterparts in Europe, where the safety approach is based on crash avoidance. So what do we have to […]

Fred Barnes: Americans Mainly Want to Stay in Their Cars

|
After yesterday’s electoral drubbing, the Obama administration will have to deal with a starkly different Congress when they make their expected push for a multi-year transportation bill early next year. We know that some influential House Republicans, like John Mica, don’t necessarily believe that bigger highways will solve America’s transportation problems. And we know that […]