What If Traffic Engineers Were Held to Safety Standards Like Carmakers?

Twenty-one bicyclists and pedestrians were struck by cars on just an 8-block stretch of Tampa's Hillsborough Avenue over a four-year period, including two 15-year-old high-school girls who were killed in two separate incidents. But NHTSA won't be issuing any fines and there won't be a class action suit against the road designer. Image: Google Maps
Tampa’s Hillsborough Avenue has an appalling safety record, including the recent deaths of two 15-year-old high school girls who were killed in  separate incidents. But NHTSA won’t be issuing any fines and there won’t be a class action suit against the street designer. Image: Google Maps

It’s been a rough few days for auto makers.

News broke last week that Volkswagen will be fined because the carmaker manipulated the data from its diesel vehicles to make emissions look lower, deceiving U.S. environmental regulators.

And on Thursday, General Motors reached a $900 million settlement with the Justice Department for covering up a defect in its ignition switches that claimed the lives of at least 120 victims over 17 years. Almost 1,500 victims have been pursuing a civil suit against the company.

Most of the chatter about the GM deal has been in the vein of this USA Today editorial calling the punishment too lenient. It looks like all GM’s employees will escape criminal charges, even though the company knew about the problem before the faulty cars even went into production in 2002. The newspaper noted that the company’s stock was up 11 cents on the news of the settlement’s details.

Auto defects are a serious issue. According to Edmunds.com “63 million passenger vehicles were recalled in 2014,” the biggest year for recalls ever. Most of the defects are not as serious as what GM is under fire for. The ignition switch problem affected 7 million vehicles over 17 years, Edmunds reports.

What’s interesting, from a safety perspective that encompasses streets as well as vehicles, is that the auto companies are held to account for dangerous conditions to a much greater degree than the engineers who design our streets.

After all, about 33,000 people are killed on American streets annually — a much higher rate than most of our peer countries — and street design is a major factor. But the designers of streets are rarely, if ever, sued when someone is killed because of dangerous, high-speed conditions.

The GM settlement comes out to $7.3 million per confirmed death. If those types of penalties were applied to cities where street designs have led to demonstrably higher rates of injuries and fatalities, how would that affect public safety and the built environment?

Consider Tampa’s Hillsborough Avenue, a wide, high-speed arterial street that runs past a local high school. From 2008 to 2012, motorists struck and injured 21 people walking or biking on just an eight-block stretch. Two of the victims were Middleton High School girls, Shenika Davis and Norma Velasquez-Cabrera, both 15, killed in separate incidents. One month after the second death, an 18-year-old classmate was gravely injured in the same area. As a result of the public outcry, the city is pursuing some design changes that will — hopefully — make the road safer.

But no one will face repercussions for the deaths of Davis and Velasquez-Cabrera. The National Highway Traffic Safety Administration isn’t investigating Tampa’s streets department for negligence or failure to ensure public safety. The victims’ families won’t be awarded a multi-million-dollar settlement.

The United States could prevent thousands of deaths each year by designing streets to better safety standards. But engineers enjoy a level of immunity we do not afford corporate entities that sell consumer products.

Why is that? We asked civil engineer Chuck Marohn of Strong Towns.

Marohn said there is really no mechanism for holding designers of public roads responsible when their product kills people. Engineers, and the municipalities that employ them, enjoy a large degree of legal immunity. “They would need to bring a lawsuit against the city, and our legal system gives huge deference to legislative bodies (city councils) and the decisions they make,” Marohn said in an email.

As long as they are abiding by local design standards (or if no local standard has been established, national standards like the Manual on Uniform Traffic Control Devices), engineers generally don’t need to worry about being held responsible. In fact, many engineers worry that if they deviate from American standards and design streets in accordance with standards that have proven safer in other countries, they will be at risk of getting sued.

Marohn says that despite the long odds, it would be worth it to legally challenge cities that have created dangerous conditions despite sticking to the standards they’re comfortable with.

“I’d like to see people be more aggressive with lawsuits where engineers, even if they follow standards, clearly ignore the safety of different classes of individuals,” he said. “I think they could win.”

44 thoughts on What If Traffic Engineers Were Held to Safety Standards Like Carmakers?

  1. Thought-provoking post, Angie. But Chuck Marohn seems to contradict himself, noting that municipalities “enjoy a large degree of legal immunity” on matters like street design, particularly via the shield of national standards, but then holding out hope that a lawsuit for ignoring dangers to pedestrians and cyclists “could win.” Which is it?

  2. “In fact, many engineers worry that if they deviate from
    American standards and design streets in accordance with standards that
    have proven safer in other countries, they will be at risk of getting

    This passage answers the question. These engineers are following the standards they’re given by varying government bodies and have little latitude to deviate from them in most cases. Suing the municipalities and other government bodies, while daunting, would be the more sensible route. If we want to see real change, it needs to come from the top down. No use suing or imprisoning the guy who was just following what he was taught and the standards he was given, flawed as they may be.

  3. I wouldn’t hold engineers to the same standards as carmakers. As far as I know, no one is suing the engineers who worked for GM or VW; the responsibility lies with the company and with the executives. I believe the same should apply to municipalities and states.

  4. How can you sue the engineers here? They likely discovered the problem, perhaps even designed a fix, but were prevented from implementing it by the bean counters. As an engineer myself, I’ve bumped up against this sort of nonsense more times than I care to admit. At least with the stuff I make, people don’t die if something which has a known defect is sold.

  5. I think Chuck’s position is that municipalities and traffic engineers should, in the future, be held accountable for the dangerous designs–not something that really happens now. Only then would we start to see movement away from the throughput bias built into our systems for SOVs.

  6. Two words: breakaway poles

    Placing one in a sidewalk is the ultimate indefensible act. You acknowledge a car going off the roadway is such a common and violent occurrence that expensive mitigation is needed, yet you intentionally design systems that place people in this same space.

    I believe that to be gross negligence, a willful disregard for the welfare of an entire class of people.

  7. If you are a PE then you should be lieble

    same as a accountant, Architect, or MD

    or plumper,

    pleading ‘I was only following orders’ went Out a while ago

  8. While I like the idea, I think there are two important considerations here:

    1) traffic engineers would love to fix everything. It’s their job. Money is limited though and dollars often have to be stretched, sadly they can’t do as much as they want due to budgetary constraints
    2) 90% of crashes are because of driver error…hard to pin that on an engineer in court

  9. “90% of crashes are because of driver error”

    And a lot of those errors are a result of drivers initially feeling comfortable with doing these risky driving maneuvers that cause these crashes, due in part to how the road is designed.

  10. The first rule of design is that there is no such thing as user error, only bad design that induces users in error. It’s not 100% true, but it’s a good rule to follow. You might rephrase that by saying that the #1 sign of a bad designer is blaming people for not using their design well.

    It’s the job of the designer, and the engineer, to make something that is easy to understand and that people know instinctively how to use. If an engineer designs a road and the crash rate is far higher than elsewhere, chances are that the problem lies with the road design and not the intelligence of the people who drive on that particular stretch of road.

  11. You mean the fragilized bases that are put under traffic light and street lighting poles? It’s not really that expensive though, and, at least here in Québec, they’re not supposed to be put in places where pedestrians might be.

    I would rather point out the idea of object clearance, the idea that any “fixed object” has to be put far away from the road in case a driver goes off the road… but then we build sidewalks right on the curb of the road, where we wouldn’t tolerate a pole or a tree to be. Can’t have a car hitting a pole there, but hitting a pedestrian? Hey, stuff happens!

    I’d say that sidewalks should never be located in object clearance zones… but I know some engineers would just not build sidewalks then. Better to say that clearance zones should only apply to highways in undeveloped areas, on city streets, there should be no object clearance zone at all, any driver who swerves off the street in an urban area deserves to get his mirrors torn off at the least.

    Check the image below for what I think should be proper street design on streets with sidewalks,

  12. In the case you mentioned, yes, it is the designers fault, and analysis should be done to improve the safety. This happens all the time, they are called RSAs (Road Safety Audits).

    I agree that blaming user error every time is not good practice, humility is important in every profession, but “you can’t fix stupid”. Engineers will try, and have come up with some brilliant ways to keep motorists safer (cable median barriers, clearance intervals, sign retroreflectivity, raised pavement markings, on and on) at the end of the day, if you drive impaired or distracted and get in a vehicle these are impossible to design for completely.

    This is exactly why driverless vehicles are soooo important. They are the key to saving lives and $$$$$

  13. “This is exactly why driverless vehicles are soooo important. They are the key to saving lives and $$$$$”

    Until there is a hardware or software failure and the vehicle goes at full speed on the sidewalk. And then who is responsible?

    The best solution is to get people out of cars as much as possible.

  14. Not in the short run, but in the long run, it certainly is, with the right policies. If you want to keep dreaming about the magical self-driving car that will be cheap yet never have a single hardware and software failure to solve problems, go ahead. But history says that relying on technological magic bullets to solve current problems is a fool’s gamble.

  15. Many of these road systems we design are safe on paper, but people do dumb things. We (I’m a transportation engineer) spend months and even years trying to ensure the systems we design are as safe as possible for such an unsafe activity *AND* meets the agency’s goals. A lot of the time our hands are tied by that second criteria and I deal with it every day. Nobody wants to take a risk and nobody wants to the be guinea pig for a new design. If that design fails or has a single issue, we get sued, it makes the front page of the paper, people lose their positions of power, and we’re out of a job. If someone is killed, it may make the 3rd page and is considered a statistic. Gruesome, yes, but that’s the reality of transportation engineering meeting politics and lawyers.

    As a designer, drink driving, texting and driving, changing the radio, eating a burger, changing lanes without signaling, running red lights, distractions, tailgating, emotions, road rage, speeding, etc. are some of the many things out of our control and we can’t possibly design for all of them without spending massive sums of limited money. If people driving followed all the laws and drove with any sense of safety and in the moment, everyone would theoretically be safe. However, people act irrationally and behave unexpectedly, which is the #1 reason why self-driving cars are so incredibly safe; no humans involved.

  16. What we need are laws like they have in the UK that allow suits for either design or maintenance if they cause bodily injury, property damage, or death. Infrastructure that causes death can result in criminal charges of corporate manslaughter, but IANAL so I have no idea how that works out.

  17. Not in urban areas. Armco barriers and cables are not there to stop a car, but to slow it down, they wouldn’t be enough to protect pedestrians from vehicles swerving off the streets. They would also be bothersome for pedestrians crossing the road mid-block.

    Outside of urbanized areas, on high-speed roads, then you can put these barriers, and if you need a sidewalk, have it be a foot/bike path a bit further out from the road, OUTSIDE the object clearance zone.

  18. It’s absolutely straightfoward to bring a lawsuit against the city and against the engineer in his official capacity (ex-parte-Young style). Since there’s no evidence whatsoever behind the “American standards”, a sufficient collection of documented expert evidence would be solid enough to press the suit. The trick is winning it.

    What’s needed is a test case: something where the “American standards” are absolutely egregiously dangerous, and where the engineer following them obviously should have known better, and was therefore recklessly endangering lives. Something where the jury’s gonna go “We can tell you shouldn’t have designed it that way”.

    Something like a design with no sidewalks, rebuilt within the last couple of years with no sidewalks, where pedestrians are getting killed.

    It would be worth hunting for a really solid test case.

  19. I think road designs with high traffic speeds and no sidewalks are probably the best target for a test case, becasue that is *inherently* unsafe, and it’s entirely the engineer’s fault: there’s no way to walk safely on a 55 mph road, even if the drivers are “responsible”.

  20. Well, if you designed it and the city refused to let you implement it, that’s one thing…

    …but I can think of plenty of examples where engineers designed blatantly and obviously dangerous roads. Mostly high-speed roads with no sidewalks, but also things like slip lanes which obstruct the crosswalk.

  21. If we’re trying to win a test case, I would target roads which are rebuilt with no sidewalks at all, and with very high speed limits for cars. Given that *every* public road except an expressway is a legal walking route, it is gross negligence to build a road with cars going 55 mph and pedestrians just walking loose in the road. And even a jury can understand that.

    Yet the “manuals” are full of “rural road designs” which do exactly this: no sidewalks, high speed limits, deathtraps.

    If you want to destroy the “I was just obeying the manual” defense, I would target these designs. It’s also worth noting that the small municipalities where these are being implemented typically don’t have huge legal defense budgets, so won’t want to drag it out for aeons.

  22. Those are legislative decisions: who to provide service to and where. A decision to not include sidewalks is discretionary, debatable and thus defensible where cities are given huge legal deference, which they are.

    It is a design decision to acknowledge pedestrians and facilitate their movement by building a sidewalk and then make that unsafe by designing for high speed traffic adjacent. The designer essentially admits the unsafe situation being created by installing a breakaway pole.

  23. In what sense is it discretionary?

    It’s the well-established law in every single state and even at the *federal* level that pedestrians have the right to travel on public highways — PERIOD. This dates back for *centuries*, and is so well-established that they had to make specific exemptions for expressways.

    So the designer doesn’t *have* to make an exclusive pedestrian facility… but if he doesn’t design a sidewalk, then designing a road with a 55 mph speed limit is blatant recklessness and endangerment of pedestrians. Not legally defensible in any way whatsoever.

    In practice, I’ve figured out that a lot of the roads without sidewalks were never designed at all — they were horse paths with asphalt dumped on them — which relieves anyone of responsibility. But it’s different when a brand new road is built. It would take some care to find an appropriate test case.

  24. You’ve written that 90% figure twice now, with no source for the statistic or explanation of how it was found. How can we even know that?
    For most crashes, the police take statements from drivers and witnesses and try to find someone at the scene who broke the law. They aren’t looking for design problems nor are they qualified to find them.
    The occasional obstruction in the road, large multi-fatality crash, or area with higher-than-expected crash rates might get an engineering analysis, but I haven’t seen investigations looking for anything other than driver error routinely done. Do you know of a study where all potential causes of a representative group of crashes was investigated?

  25. Wouldn’t the result of a “winning” breakaway case like that be for later engineers to just not acknowledge pedestrians and not install sidewalks? The no-sidewalk design probably would be cheaper and have a higher Level of Service.

  26. And then there is its corollary, designing a road that encourages a high traffic speed and then placing a much lower speed limit on it. People respond to cues and will drive in a corresponding manner. If the road looks like you can go 55-60 and feels like you can go 55-60, people will go 55-60, even with a 35 MPH speed limit.

  27. It won’t happen because the design brief for designers is throughput, not safety.

    Something that needs to change.

  28. “The first rule of design is that there is no such thing as user error”

    Building any system on top of such a grossly flawed assumption, it’s a wonder any of us get out of our own door without meeting a catastrophic fire-y doom.

  29. We have thousands of perfect cases : they are intersections in NYC where the engineers put pedestrians in harm’ s way by letting cars turn at the same time pedestrians have the walk sign. It is equivalent to turn right on red for cars, and banned in NYC because it’ is obviously dangerous for cars … But when it comes to pedestrians, it is Ok to put them at risk…
    In this case the city and the engineers clearly design dangerous intersections.

  30. I guess I don’t buy this as being an engineering issue but a legislative one. I know there is discussion and some examples of all reds at lighted intersections allowing peds free range. I suppose the biggest problem there–having walked it–is the speed at which people drive and often turn so it’s much more likely to be serious or fatal. So in that respect slowing the cars on the street prior to turning could be an engineering issue.

    Is the right on red banned because sightlines are so poor that it causes too many collisions? This type of wreck is not a frequent thing in the west despite it being allowed except at intersections that forbid it (which I think is triggered by sightlines, high speeds of cross traffic or significant ped movement),

  31. Can you explain what you mean by a “legislative one” ? I ma not following you. DOT analysis of pedestrian injuries and fatalities in New york city has demonstrated that 27% of crashes involved pedestrians in the ped crossing with the walk sign…Still DOT is fine letting the cars at the same time as they cross. Why does the DOT install traffic signals to divide the time between cars but not between cars and pedestrians? I am baffled !

  32. How do you engineer to protect the vulnerable this situation? If it is a right or left hook then the only way to eventually prevent it is to outlaw the turn when people are crossing which of course then needs enforcement to make the behaviour stick. My position presumes you can’t provide a four-way walk all auto red environment.

  33. Be careful what you wish for.

    If the system you propose were implemented, the pedestrian phase everywhere would be as brief as standards permit. A fully law-abiding pedestrian walking multiple blocks would have to wait at every intersection, aside from the small number that he or she reaches during the several-seconds-long walk phase.

    Much as most pedestrians today in New York don’t slavishly obey the pedestrian signals but instead cross when it seems safe to do so, so too would pedestrians under your scenario. The problem is that the don’t-walk signal (which is the phase almost all pedestrians would encounter on arrival at an intersection) would be rendered ambiguous and therefore meaningless – it might mean that cross traffic is approaching, or it might simply mean that an occasional driver could turn, or it might mean that the system is transitioning from one to the other.

    And that ambiguity could easily lead to a reduction in safety.

    Furthermore, drivers would have no obligation to yield to the vast majority of pedestrians, who enter the crosswalk outside of the brief walk phase.
    Drivers know that it’s illegal to turn left, even on a green light, across oncoming traffic. (Only at the busiest left turn locations are specialized signal phases provided.) The exact same law requires drivers to yield to pedestrians in crosswalks while turning. The problem is that the law goes virtually unenforced, so lots of drivers either don’t know our don’t care about it.

    The problem isn’t with the signals. The problem is with drivers who break the law and with a law enforcement system too weak to encourage those drivers to obey the law.

  34. The solution is a fully protected left turn signal, where the turning lane has a red arrow during the time the pedestrians have the walk signs. So no conflict . Meanwhile the the thru traffic has a green.
    This configuration is used on many left turns on the highways, and on various intersections Chelsea and Hell’s Kitchen. It has been very effective and seniors really like it. it does not impact negatively the thru flow.

    A recent study of 68 intersections in NYC. Has shown a 70 % reductions in crashes at these intersections.

    (There are many more ped crashes on left turns so that is where you start ).

  35. A fully law abiding pedestrian today has to wait at every intersection since the signals are timed for a green wave for cars – which by the way induces dangerous driver behavior in trying to beat the lights to catch up to the wave…
    The Don’t walk signal is not ambiguous under any scenario. Green and red lights are not designed to be obeyed only when there is a lot of traffic, otherwise drivers would cross lightly trafficked intersections all the time….

    And yes the drivers do not have to yield when there is a don’t walk just like anywhere in the city and pedestrians who cross at the don’t walk take their lives in their hands – but they know it and are extra careful about it… Instead of believing they are protected by a walk signals and finding out they are at the mercy of drivers who could not care less…

    Vision zero is all about engineering the roads to prevent mistakes… Not to push the problems down the line to enforcement, which we cannot afford.

  36. Where the vast majority of our crashes involving people on foot or bike and people driving are right turns. Probably due to the one-way system in the city. But here’s where we have issues. The vulnerable users do not abide by the walk signal and get hit (especially in limited or no light).

    I saw, for the first time in years, a four way walk intersection the other day. It is a very dense, small environment with significant ped movement, large numbers of people driving wanting to turn and thankfully a very narrow roadway. I have been here many times but the signals were new. Because I pay attention to signal sequences and am involved with traffic safety–plus I was not in much of hurry–I sat back and watched it through a cycle. The peds screwed it up for the turning autos because they just ignore the ped signals and walk regardless. Thus huge auto queues for several blocks. Interesting behaviour study.

  37. The Dot studies show the majority of pedestrian crashes with injuries being on the left turn, not right.

    The government’s mission is to protect citizens who abide by the law first and foremost. If a driver goes at 150 miles per hour and kills him/ herself on a curvy road where the speed is limited to 50 you would not worry about the road design as much as if drivers would kill themselves while respecting the 50 limitation.
    Same for pedestrians: we need to focus first on protecting those who abide by the law but where the design of the intersection put them at risk.

    I am not suggesting an all red intersection, just a red left arrow for left turning cars.

    Huge auto queues happen everywhere in the CBD and the good news is that huge auto queues do not kill pedestrians… While drivers turning at green light kill 27 % of pedestrians….. So what is the proper decision here ? Less queues or less dead ?

    In midtown there are so many pedestrians crossing at certain intersections that drivers never get a chance to turn, in this case a red/ green arrow would give them a fair amount of the cycle and reduce the queues…

  38. Our state and particularly where I live there have been almost no left turn crashes. Almost all of them are right hooks. Lefts are probably more often fatal here though because of the auto speeds involved. Portland, which has a significant one-way system in it’s downtown core hasn’t, to my knowledge distinguished between lefts or rights. Too most of their fatals are people riding bikes.

    I have no problems rejiggering the bias away from autos–it was merely an observation about what was happening with a new design. Here however we have right and left turns (onto a one-way from either a one or two way) legal on red–been that way since the beginning and like sales tax and pump your own petrol will be nearly impossible to change.

    That car belonged to an Iowa collector but it fits Portlandia so well. We had a 69 V4 (better engine, no smoke) same color when we first moved to PDX in the early 80’s. Would love to have it back but it wouldn’t replace our i3.

  39. Tampa, Florida: One of the most dangerous intersections in Tampa is at the intersection of the James A. Haley Veterans’ Hospital and the University of South Florida’s Shimberg Health Sciences Library and the iQ Luxury Apartments.

    The above intersection is located where Richard Silver Way joins Bruce. B. Downs Blvd.

    What is required is road and highway improvements and traffic signals.

    You can help expedite road and highway and traffic signals installation by signing anonymously or by name the Facebook GoPetition at . Thank you!

Leave a Reply

Your email address will not be published. Required fields are marked *


Pretty Please: U.S. DOT Asks Carmakers to Limit Onboard Distractions

Transportation Secretary Ray LaHood’s signature issue has been distracted driving. He’s spent the last four years amplifying the heartbreaking voices of those who have suffered the consequences of this highly dangerous habit. The stories of the needless loss of so many people, especially children and teens, are tragic. Clearly, it’s time to take decisive action to stop […]