Light Rail Increasingly Dangerous

A pedestrian was killed by a light-rail train in Denver last Thursday, February 12. The very next day, another pedestrian was killed by a light-rail train in San Jose.

According to the Bureau of Transportation Statistics, 40 people were killed in light-rail accidents in 2012. This is the most since at least 1992 (the earliest year for which I have numbers available). While the numbers vary from year to year, in all the years since 1995, light-rail accidents killed 333 people.

A few days ago, the Antiplanner mentioned that auto accidents kill about 34,000 people a year. That sounds horrible, and it is, but unlike light-rail numbers, auto fatalities have been declining. More important, light rail carried just 26.7 billion passenger miles in all the years between 1995 and 2012. By comparison, highway vehicles traveled nearly 3 trillion vehicle miles in 2012 alone. At an average occupancy of 1.67 people per car (see page 33), that’s 5 trillion passenger miles.

In other words, light rail kills 12.5 people for every billion passenger miles carried, whereas buses kill just 4.5 people per billion passenger miles. Urban roads and streets, by comparison, kill about 8.2 people per billion vehicle miles, which works out to 4.9 per billion passenger miles. While buses are slightly safer than cars, light rail is 2-1/2 times more dangerous than cars.

Transit officials are quick to blame the victims when rail accidents kill. “Did he dive under the train? We don’t know,” said a Denver official. (It turned out the man was slightly disabled.) But the real problem is putting 100,000-pound vehicles (or, worse, 300,000-pound trains) in the same streets as 150-pound pedestrians.

Denver compounded the problem by stupidly building its downtown light-rail tracks on one-way streets with the trains moving in the opposite direction from the rest of the traffic. This means pedestrians looking for cars coming from one direction may not see the train coming from the other direction. Last Thursday’s fatality took place near this location, but as officials weren’t even certain where the accident took place, we don’t know if wrong-way travel played a role.

Unfortunately, transit agencies and rail advocates care less about public safety than in getting their projects built. I once debated the head of the American Public Transportation Association who put up a chart showing that light rail was far safer than driving–for the light-rail occupants. What happens to mere pedestrians apparently isn’t important.

Share

28 thoughts on “Light Rail Increasingly Dangerous

  1. metrosucks

    I once debated the head of the American Public Transportation Association who put up a chart showing that light rail was far safer than driving–for the light-rail occupants. What happens to mere pedestrians apparently isn’t important.

    This is exactly what I expect the usual transit apologists to do. Pretend that light rail is safer than buses & cars because it happens to be safer for the rail occupants. Nice little sleight of hand huh.

  2. FrancisKing

    “In other words, light rail kills one person for every 12.5 billion passenger miles carried, whereas buses kill about 4.5 people per billion passenger miles. ”

    How does that work? We start with ‘highway vehicles’ and end up with ‘buses’. Buses are far more dangerous than cars. People wander around supermarket car parks just fine, but bus depots have up signs saying ‘Extreme Danger’.

    My understanding is that light rail is substantially safer than buses. Light rail, unlike buses, have anti-crush guards on the wheels.

    “Transit officials are quick to blame the victims when rail accidents kill. “Did he dive under the train? We don’t know,” said a Denver official. ”

    Light rail vehicles in Croydon, London, UK, are fitted with video cameras. And yes, the video footage shows that pedestrians do amble across in front of light rail, and are surprised when it doesn’t stop in a few yards. At least they don’t say, as did a car driver when he pulled out on an HGV, “Sorry mate, didn’t see you”.

  3. rjmason

    “In other words, light rail kills one person for every 12.5 billion passenger miles carried, whereas buses kill about 4.5 people per billion passenger miles”

    I think you mean 12.5 fatalities per billion passenger miles, not one fatality per 12.5 billion passenger miles (which would be much safer).

  4. FantasiaWHT

    Did you mean 12.5 fatalities per billion passenger miles for light rail? Because the way you have it written now, light rail is about 50 times safer than buses and cars.

  5. gecko55

    I wouldn’t discount suicide. In Switzerland, jumping in front of a train is one of the most common methods, (since handguns aren’t ubiquitous) — although that’s typically suburban trains (S-bahns) not trams. I’ve heard that’s also the case in Japan, and would guess in the UK as well (FrancisKing?).

    Sure, trams in busy urban settings will sometimes mow down an innocent pedestrian. But 333 fatalities in 20 years? That doesn’t sound “increasingly dangerous” or even remotely dangerous to me. (And as others have noted, the AP has botched the math in any event.)

  6. JOHN1000

    The urban progressives continuously emphasize walkability as a measure of the greatness of cites over suburbs. It is probably their strongest argument for cities.

    I love to walk, but, based on some of the answers here, I am now considered to be an annoyance by light-rail supporters. If I get hit by an automobile, it proves that automobiles should be banned. If I get hit by a light rail train, it proves that pedestrians should be banned.

    As the light-rail group tends to be the same people who claim to love walkability, the hypocrisy is overwhelming.

  7. bennett

    “Lies, damned lies, and statistics”

    Mr. O’Toole’s borderline fanatical obsession with per passenger mile calculations so callously obfuscates the dangers of automobiles it’s becoming laughable. 1,000 times as many people are killed by cars than by trains in a given year. The most dangerous thing most Americans do every day is get behind a wheel or step foot on a road. The dangers multiply when you take your car on the beloved highways. But the argument is that more people drive so driving is safer. Of course an American is more likely to die in a auto related accident than any other accident or premeditated attack.

    Frank brings up a great point too. Fatalities aside, driving related accidents are the most likely cause of injury in America. My guess is that ppm calculations would favor transit as well, which is why it’s omitted from anti-transit analysis.

    I also think occupancy rates comparing transit and autos are skewed. We don’t count transit operators as occupants, and rightly so. I live in a starter home neighborhood in south Austin that is dominated by 30 something parents. So much of our existence behind the wheel is that of a transit operator. The trip purpose is solely to transport someone else. For these trips, let’s say 5-15% of daily trips for parents of more than one kid, why should the driver be counted? I would also claim that the rate of these types of trips is directly proportional to the separation of land uses. That is, the suburban form requires more transit operator trips of parents compared to urban forms that have parks, rec centers, shops, etc. within a distance an 11 year old can bike to. But that’s another tangent for another day.

  8. bennett

    “What happens to mere pedestrians apparently isn’t important.”

    Hello pot, meet kettle. Coming from the guy who opposes every pedestrian safety effort, most notably traffic calming on small surface streets, this is rich. Similar to John1000’s pointing out the hypocrisies of planning advocates, the same can be said for antiplanning arguments. What happens to mere pedestrians apparently isn’t important if it impacts road flows. Heard it here several times.

  9. gilfoil

    “What happens to mere pedestrians apparently isn’t important.”

    Pedestrians are absolutely not important. There is no possible justification for restricting drivers’ mobility in favor of such a low-mobility mode of transportation as walking. Walking is perhaps the least efficient, slowest, and most expensive modes of transportation per passenger mile. Until pedestrians start paying for their lavishly-constructed, taxpayer subsidized sidewalks, they need to consider joining the rest of the world and buying a car.

  10. Sandy Teal

    Trains, planes, and automobiles are all safest when they are separated from each other and from pedestrians. The only reason that dedicated separated highways are dangerous is because they are so safe that people drive very fast on them. If people drove 30 mph on interstate highways there would be almost no fatalities on them.

    The problem is that the whole idea of urban cities is to cram everything in close together. That has many costs and many benefits. Personally I think adding light rail to many cities adds another complication to getting around as it changes parking and traffic lights and traffic flow at a huge cost and very little ridership. Busses have such less complication because they flow as all other traffic (except for bus stops).

  11. The Antiplanner Post author

    Bennett,

    Per passenger mile is the only valid stat. Yes, more people are killed by cars than by light-rail trains, but that’s because the latter are so rare.

    And evidently you never really read my critiques of traffic calming, which show that a lot of traffic calming actually makes streets more dangerous for pedestrians.

    All: As many people pointed out, light rail kills 12.5 people per billion PM, not one per 12.5 billion. Thanks for pointing it out.

  12. metrosucks

    I think Randal gets very passionate about his subjects and can sometimes overlook an argument or two. I can certainly relate to that. One of the best ways to tackle the whole light rail kills people debate is this…deaths, whichever way they are caused, are undesirable and expensive for society.

    However, light rail, with its very low utility and existence as a corrupt giveaway to contractors, is the most insulting and pathetic way to die. It’s essentially saying people are dying so Stacey & Witbeck can improve its bottom line and some New Urbanists can feel selfish satisfaction. Light rail doesn’t improve our life in any way. It is a overall drain on society and a giveaway to the already rich & corrupt. It is a horrible tradeoff for its costs, including the deaths it causes.

  13. ahwr

    @AP
    Measuring safety based on passenger miles alone implies that rail trips replace identical car trips, that transportation infrastructure has no impact on land use, that it has no impact on mobility, which is ridiculous. Ignoring injuries, even serious life altering ones, and only focusing on fatalities ignores a great deal of carnage. For people struck by trains are they more or less likely to die than those involved in automobile collisions?

  14. C. P. Zilliacus

    FrancisKing wrote:

    Light rail vehicles in Croydon, London, UK, are fitted with video cameras. And yes, the video footage shows that pedestrians do amble across in front of light rail, and are surprised when it doesn’t stop in a few yards. At least they don’t say, as did a car driver when he pulled out on an HGV, “Sorry mate, didn’t see you”.

    This reminds me (and not in a good way) of what was reported to have happened just prior to a multi-fatal Metro-North Railroad (railroad, not light rail) crash in Valhalla, Westchester County, New York (north of New York City) earlier this month. Supposedly a car was passing through a grade crossing (you might call it a level crossing) when the gates came down, doing some minor damage to the vehicle.

    Foolishly, the driver of the vehicle got out to look at the damage while her vehicle was stopped on the tracks with a train coming. That was literally a fatal mistake.

  15. metrosucks

    Basically, planners are saying that deaths caused by light rail are merely the “egg-cracking” necessary to create a wonderful new social contract based around the faultless, gold-plated light rail systems they are pushing. If you get my meaning.

  16. OFP2003

    By now this conversation is probably dead, nevertheless I’ll add that in the first half of the 20th century a relative of mine was killed by a street trolley. The resulting PTSD in the survivors manifested and affected the family well into the second half of the 20th century.

  17. cecc0011

    I challenge the use of passenger-miles. It doesn’t matter to a person in the back seat of the car, on a bus, or a pedestrian/cyclist how far the driver who hits them has traveled. A car who drove 20 miles into the city kills a pedestrian the same way a car who just left the neighborhood block on a 2 mile jaunt to the baseball park. Passenger-trips by mode is a much better metric. Since not every trip by differing mode has the same distance, but usually carries roughly the same utility (commute to work, grocery run, bringing the kid to daycare, whatever), this is far more relevant.

    Even when viewed in that regard, LRT has a fatality rate of 82 deaths/billion passenger *trips* (with 83% being non-occupants) vs 54 deaths/billion passenger *trips* for passenger vehicles (non-light/heavy trucks – only 14% of which are non-occupants). Compare that to 16.3 LRT deaths/billion passenger-miles & 55 vehicle-related deaths/billion passenger-miles (my numbers are a little more specific than yours, drilling down using NTD data and a few other sources). The gap is much higher when looking at this through a per mile vs per trip lens.

    Further, LRT and urban buses operate in environments almost exclusively where pedestrians/cyclists legally have the right to be present. Many miles are driven by cars on grade-separated roads or highways/arterials far from urban development. When you take this into account, LRT, at 82 deaths/bn pass-trips comes even closer to cars at 70 deaths/bn-pass-trips.

    In all these scenarios, urban buses and heavy rail beat out cars by a large margin. Of course, non-occupant death rates as a share of total for all transit is much higher than passenger cars. We should absolutely work to build better transit/streets, hold drivers accountable, etc to improve safety. But I think your framing of the issue is intentional and somewhat misleading. I’m happy to share my data if you’s like.

    The Antiplanner Reply:

    Actually, it does matter if you go 5 miles or 10 miles. If you can go 10 miles in a decent amount of time, you have access to four times as many destinations as if you can only go 5 miles in that amount of time. LR trips average about 5 miles; auto trips about 12 miles, so auto trips can access roughly six times as many destinations. But, as you say, rail loses by either measure. When calculating auto fatality rates, it is important to count only urban driving (which is safer than rural driving), because rail transit is found almost exclusively in urban areas.

  18. cecc0011

    “Rail” doesn’t fail by either measure, LRT does, and barely when comparing per-passenger trip rather than per pass mile.

    Counting urban driving on all road classifications masks the true rate of pedestrian deaths (measured by trips or miles), because there are grade separated highways and interstates that make up a huge share of VMT where pedestrians have a near-impossible job accessing it to get hit in the first place, whereas buses and LRT run almost exclusively at-grade in urban environments. Weeding that out is important. Additionally, excluding “rural” VMT misses traffic on local and minor collectors found within townships/cities that fall under the population threshold, yet nonetheless have real pedestrians being killed by real cars every day. Those environments are every bit as car-oriented as the same road design/population density that happens to fall in an “urban” area.

    And the whole point of your post was under the guise of pedestrian safety. It literally does not matter if, as a pedestrian, the car that hits me traveled 2 miles or 10 . It’s the fact that they’re making a trip by a particular mode that counts. Evaluating those modes’ deadliness relative to how many people they’re moving from A to B (regardless of how far A & B are apart) is the important stat. Otherwise we could just convince everyone to drive 100 miles for every trip, not lower the pedestrian fatality total, and look like we’re all of a sudden 10x safer than we currently are (2009 average car trip length is 9.7 miles).

  19. metrosucks

    I see that every rail fanatic is busy trying to bend the numbers to suit their fetish and disregard anything that shines a bad light on the precious choo choo trains.

  20. Frank

    metrosucks, yes,yes, but I’m still perplexed as to why the AP gives msetty’s “tweets” play here given msetty’s trolling nature, use of “slurs,” peevishness, and general douchbaggery (sic).

    I’m still perplexed why the AP will answer a neophyte’s post before replying to my question of why fatalities are the metric for safety rather than injuries. I’d genuinely like to be educated. “Srsly” to quote my favorite planner boy.

    Yet msetty and this guy, likely msetty’s butt buddy (not Biden), get responses.

    Things that make you go hmmm.

  21. Andrew

    Randall:

    Its well known that a small segment of people use rail lines as a way to commit suicide, while people generally don’t jump in front of cars, trucks and buses (although they do drive them off bridges and cliffs and into lakes and rivers and even occasionally in front of train).

    I would love to see a calculation made that excluding suicides on both sides, as this obviously distorts the numbers.

    Specifically, I’d like to know the danger to pedestrians – i.e. whether light rail or road vehicles kill more pedestrians per VEHICLE mile, and I’d like to know the danger to riders/drivers – i.e. whether light rail or road vehicles kill more riders and driver/operators per VEHICLE mile and per PASSENGER mile.

    I’d also like to know how deaths from light rail-road vehicle accidents are treated in your numbers. It would seem to me that deaths from a road vehicle that violates a rail vehicles operating right of way should go in the road vehicle column and be subtracted out of the rail column, and that is likely the great majority of such accidents. Rail should get credited with accidents when its vehicles leave their tracks and strike innocents or kill riders from a rail system malfunction.

    An example of this is the recent Metro North accident in New York. A stupid driver violates a grade crossing that derails a train, causing her death and the deaths of 5 others. It seems clear to me that these should all be dumped into the road death column, but I suspect your statistics treat it as 6 rail deaths (all who died in the accident) and 1 road death (the driver counted a second time).

  22. ahwr

    Andrew:

    http://www.ntdprogram.gov/ntdprogram/data.htm
    http://www.ntdprogram.gov/ntdprogram/pubs/MonthlyData/S&STimeSeries-October2014-02022015.xls

    2002-2014: of the 336 fatalities involving light rail vehicles 74 were classified as suicides, 1.055 billion railcar (not entire train) revenue miles, 25.694 billion passenger miles, 5.4 billion passenger trips (4.8 miles/unlinked trip), 10,190 injuries. Of the fatalities 10 were passengers, 34 were “revenue facility occupants”, 1 was an operator, 1 was other employee.

    2002-2014: of the 1169 motor bus fatalities 34 were classified as suicides, 24.454 billion vehicle revenue miles, 258.364 billion passenger miles, 67.749 billion passenger trips (3.8 miles/unlinked trip), 175,597 injuries. Of the fatalities 103 were passengers, 87 were “revenue facility occupants”, 15 were operators, 9 other employees.

    @AP Going twice the distance gives you access to four times the number of geographical points. But that isn’t the same as destinations, you’re assuming constant land use.

  23. gilfoil

    From the Antiplanner’s post:

    In other words, light rail kills 12.5 people for every billion passenger miles carried, whereas buses kill just 4.5 people per billion passenger miles. Urban roads and streets, by comparison, kill about 8.2 people per billion vehicle miles, which works out to 4.9 per billion passenger miles.

    So if I understand correctly this paragraph correctly:

    light rail kills 12.5 people/billion passenger miles
    buses kill 4.5 people/billion passenger miles
    urban streets and roads: 8.2 people/billion passenger miles

    How are the streets and roads killing people? Does he mean cars on urban streets and roads? Why are highways not included?

  24. EyesWideOpen

    Why are all collisions involving a train pegged as being the train’s fault? These numbers are meaningless, as they are totally cherry picked. The number of train riders who are dying in accidents while riding, is in fact miniscule.

  25. The Antiplanner Post author

    gilfoil,

    Sorry about my imprecision. Motor vehicles on urban roads and streets are involved in accidents that kill about 8.2 people per billion vehicle miles or 4.9 people per billion passenger miles.

    EyesWideOpen,

    Collisions involving trains are not the trains’ fault. In my opinion, they are the fault of those who put the trains in the streets. You are right: few people die while they are aboard a light-rail train. That doesn’t mean you want one running through your neighborhood.

Leave a Reply