Safer Streets Require Less Traffic Engineering, Not More

by James A. Bacon

A week or two ago, I lamented the disparity between the high cost of traffic accidents in Virginia and the paltry resources devoted to reducing their number. But to say that insufficient attention is being paid to the issue is not to say that no attention is being paid. According to Virginia’s 2012-2015 Strategic Highway Safety Plan, traffic deaths in the state fell 23% between 2006 and 2010. We are making progress.

Since my blog post, the Hampton Roads Transportation Planning Organization (HRPTO) published its own regional safety study. The HRPTO did a number of worthy things. It mapped the location of accidents across the region and honed in on the worst freeways and intersections that offered the best prospects for improvement. The planners asked a critical question: Which improvements offer the best benefit/cost ratio? As Brian Chenault with the HRPTO summarizes the range of possible solutions: “Recommendations run anywhere from restriping pavement, bicycle and pedestrian improvements, optimizing signal timing, adding additional lanes, trimming vegetation, reducing speed limits, adding extra signage/signals, etc.”

As untutored as I am in the arcana of traffic engineering, it strikes me that the HRTPO has identified a number of projects that will provide considerable bang for the buck, at least as measured by traditional traffic-engineering criteria. Some of the tactics it advances are ideas that I have championed, such as making better use of roundabouts and traffic light synchronization. Its recommendations undoubtedly would amount to a net positive for the people of Hampton Roads — and I hope readers will view it in that light even as I discuss its shortcomings.

The problem with the study is that the authors approach the safety issue from a traffic-engineering perspective. Not surprisingly, every traffic-engineering challenge has a traffic-engineering solution. Indeed, for some safety problems, traffic-engineering solutions are entirely appropriate. But for many, they are not. In many instances, the best solution is precisely the opposite of what traffic engineers recommend.

For purposes of illustration, let us look at the intersection of Mercury Boulevard and Power Plant Parkway, as shown above. (Click for larger image.) This particular intersection averaged 43 crashes per year between 2009 and 2012, the second most of any intersection studied. The HRTPO’s analysis attributes the high accident rate to a variety of micro-factors, which can be summed up by noting that the traffic patterns here are highly complex, leading to a high number of rear-end crashes and side-swipes. The report makes three recommendations:

  • Add a painted triangle yield line with YIELD pavement markings
  • Relocate stop bars
  • Add a flashing “signal ahead” sign
  • Add a pedestrian signal and crosswalk with ladder striping

What’s the common denominator here? It’s the expectation that more signage and clearer road markings will induce drivers to drive more safely.

Now, let’s take a radical leap. At the 2014 Congress for the New Urbanism conference last week, Ben Hamilton-Baillie, an English consultant who specializes in reconciling traffic movement with quality public spaces, showed a clip of San Francisco street traffic filmed days before the 1906 earthquake. There were no marked lanes. There were no turn arrows. There were no traffic lights. There were no street signs. Pedestrians, trolleys, horses, horse-drawn carriages, bicycles, pull-carts and automobiles shared the street. At first view, it seemed like chaos. People were jay walking and standing around in the middle of the street. Cars weaved between lanes. Vehicles made U-turns. It was a traffic engineer’s nightmare. But no one — at least not in this 12-minute film clip — experienced a mishap. Watch the clip. It’s mesmerizing.

Undoubtedly, some accidents did occur in those days and people did get hurt. Such incidents provided the pretext for the traffic engineers to step in. In the 1920s and 1930s, they segregated vehicular traffic from pedestrian traffic — roads became the domain of cars, and people were relegated to the sidewalks. Not only would this arrangement create safer streets, it was thought, keeping people off streets would allow cars to travel faster, thus increasing the carrying capacity of streets and reducing congestion. That is the path cities have traveled for some 80 years now, with the result that contemporary Americans cannot imagine any other way of doing things. Today, the traffic-engineering solution to every problem is more signs, more lane markings, more signals, more information overload.

improve_pavement_markingThe logical result is something like the intersection shown at left, which is pulled from the HRTPO’s list of crash counter-measures under the tactic of “improve pavement markings.” I would argue that the intersection is so complex that there is no way to improve pavement markings. Motorists will be confused regardless. Drivers encountering the intersection for the first time will be befuddled, as will elderly drivers with slower cognitive functioning, as will distracted drivers tweeting on their cell phones. As for marking the crosswalks with ladder striping, that will only confuse drivers all the more while doing little to encourage pedestrians to brave the inhospitable intersection and cross seven to eight lanes of traffic.

What would happen if someone removed all the traffic lights, lane markings and street signs? Would the street devolve into chaos? Some cities in Germany and the Netherlands have experimented with that approach quite successfully. That’s fine for smaller communities with low traffic volumes in isolated locations, one could argue, but is it practical elsewhere?

Hamilton-Baillie helped re-engineer a major intersection in the center of the English town of Poynton where there was both heavy pedestrian traffic and vehicular traffic, including heavy trucks. The roads had been engineered to maximize traffic flow at the expense of walkability. The town ended up with the worst of both worlds: the traffic back-ups were terrible while the shops and businesses in the enter of town went into decline. In desperation, the city shifted to a system of shared space, lower traffic speeds and less signage — the antithesis of the traffic-engineering solution. The result? Watch the video below.

Spoiler: The Poynton redesign was hailed as a success. Pedestrian traffic revived. And, despite the slower speeds, the continuous flow of traffic increased the capacity of the streets. Although they drove more slowly, trucks and cars moved more continuously.

The problem with conventional traffic engineering, according to Hamilton-Baillie, is that traffic engineers don’t take human nature into account. “In the absence of controls, it isn’t as if we don’t know how to behave,” he said. Bringing down speeds to a level where drivers and pedestrians can see each others’ faces changes the dynamic of driving behavior. When they perceive pedestrians as people, not obstacles, motorists drive much more cautiously. “Social protocols are more powerful than traffic regulations.”

In the traffic-engineering paradign, roads are sluiceways in which cars alternate between sitting behind stop lights and driving at 45 miles per gallon in a stop-and-go pattern. You couldn’t design a system more likely to create read-end accidents. Frustratingly, the ubiquity of stoplights means that cars never reach high average speeds. But when you slow down the speeds and introduce bicycles and pedestrians into the equation, the dynamic changes. Drivers become far more cautious, accidents are far less injurious when they do occur, street life is revitalized… and drivers cover just as much ground.

Bacon’s bottom line: We can’t leave traffic safety to the traffic engineers. While an engineering approach like the state highway safety plan and HRTPO study can make roads marginally safer, there are inherent limits to what the traffic-engineering methodology can accomplish. We should look to our own past, or to contemporary Europe, for alternative ways of organizing our streets.

Share this article


(comments below)


(comments below)


23 responses to “Safer Streets Require Less Traffic Engineering, Not More”

  1. larryg Avatar

    I actually agree with most of what is asserted here with a couple of provisos.

    you don’t put an at-grade ped/bike crossing on an interstate highway for a damn good reason… no matter who says that you need to have a “crossing”.

    we don’t try to slow traffic down so they can see “faces”

    stay with me here… what do we do when you need a “crossing”. You build a bridge or tunnel for the bike/ped – at no small cost and I’m sure everyone has seem these structures – as some have “wrap-over-head chain link to keep things from being thrown off including people.

    We actually have a good example of this down my way where Fredericksburg has built a trail – and that trail crosses two very busy roads – Route 1 and Route 3 … and the “plan” is to put up an “on-demand” countdown light that will basically immobilize auto traffic from all 4-directions.

    The current traffic signal there – at rush hour – backs traffic up for 3-4 light changes before people can clear the light.

    Imagine what is going to happen if bike/ped folks start using that crossing on a regular basis.

    there is no way to fix this without building a bridge – one sufficiently high to allow tractor trailers under it – and then you’d need a seriously inclined path to get folks from ground level to the bridge crossing.

    the reason they’re not building the bridge is money. That bridge would cost as much as much of the rest of the trail.

    Okay – so whose fault is this? how should it be fixed? this is not really about traffic engineers not understand ped/bike safety.. they’re actually quite concerned about it.

    also -bike/ped has for a long time been codified safety approach…

    look up “clear avoidance zone” to see how traffic engineers have, for a long, time approached bike/ped.

  2. you don’t put an at-grade ped/bike crossing on an interstate highway for a damn good reason… no matter who says that you need to have a “crossing”.

    VDOT is proposing this in Tysons – at-grade bike and pedestrian crossings at the Beltway-Route 123 exit and entrance ramps. It is the most foolish thing I’ve seen in years. It’s an absolute invitation for traffic fatalities.

    1. larryg Avatar

      well.. because folks like Bacon and the Smart Growth weenies and Greater Greater Washington have lobbied to do that..

      this is sorta like militant hang-glider enthusiasts demanding to use Dulles for Airport their flying!


  3. larryg Avatar

    anyone who has ever sat at an un-signalized side road to a busy highway – knows that it’s near suicide to try to force your way into the flow.

    Imagine there being an intersection like that – and or remember the last time there was a power outage and you found your self at the intersection of two busy roads .

    the funny thing is – if you put a policeman there – he can usually get traffic moving without an accident but let the policeman go away and see if the folks at the light can devise a “take turn” protocol collaboratively without a policeman present.

    In other words – there are no lights – there are no gates to hold traffic back – and the only difference is a policemen telling one stream of traffic to stop and then telling another stream t0 go.

    How come that traffic – by itself – could not agree on a similar protocol among themselves and not require someone to direct them (something almost anyone could do by assume the role of “director” – like you’d see at the scene of a crash and someone, often just a volunteer are “directing” traffic.

    we’ve heard of “defensive driving” but has anyone heard of “collaborative driving”? It’s not unheard of at all. It happens everytime someone lets you in traffic or lets you turn left across traffic or move in front of you to turn onto a side street.

    A certain number of people practice this -on a regular basis – and others -don’t have a clue.. they’r e not even thinking about it and he ones that do – they take certain satisfaction in watching you twist in the wind.

    most of all – people drive to effing fast – for conditions. they drive the same speed no matter the traffic or weather conditions and not surprisingly, come to grief …

  4. I forgot to add VDOT wants to widen the outside lanes on 123 to 14 feet so they can be shared with bikes. Now I don’t have a degree in engineer, but …….. Insanity. Some roads cannot accommodate bikes or pedestrians safely. They need to find other locations.

    1. You need a significant buffer between bicycle lanes and fast-moving traffic or no cyclist will feel safe using the lane. Without proper safeguards, building a bicycle lane is a waste of money. I can’t help but think that VDOT understands that.

      1. larryg Avatar

        VDOT basically does what they have to do but they don’t think when bikes get to major road crossings that there are any good solutions.

      2. I spoke with two people who attended the VDOT meeting and personally reviewed the drawings. This is exactly what VDOT is proposing. It’s insane. Someone should say “back to the drawing board.” You simply cannot have at grade bike/ped crossings at 123 and the Beltway. People will die.

        1. larryg Avatar

          re: ” You simply cannot have at grade bike/ped crossings at 123 ”

          it’s almost impossible at most busy intersections… and you cannot even fix it
          very easy with grade-separated because of the on and off ramps.

  5. I just emailed the VDOT drawing to Bacon. Post it.

  6. larryg Avatar

    re: ” very easy with grade-separated because of the on and off ramps.”

    NOT easy because of RAMPS.

  7. larryg Avatar

    I think VDOT and FHWA do a pretty good job with cars.

    But I much like the Hampton Roads idea of capturing accident data by site.

    and I agree – some simple, cheap things can be done – not the least of which is deploying red-light cameras at accident prone locations that are seeing T-bone type (red light running) accidents.

    and I would propose that there be 1: a generous time gap.. and 2. – two warnings before you get the 3rd one that is a real ticket.

    totally post the fact that you have a red light camera and that a one second grace time exists to tamp down rear-end collisions.

    then – if the rear-end collisions continue – consider re-configuration of the intersection to a round-a-bout.

    basically people drive these days too fast for conditions in my view.

    done it myself on occasion..but the bottom line is – when we push the envelope – we risk running into someone doing what we are doing and when that happens – we have an accident.

    I’m amazed. Everyday in a county of about 125K people – we have multiple accidents.. even though we are down from a year ago – we still have a lot of accidents and while some of them may be due to road safety issues – most of them are due to just too fast for conditions and bad judgement.

    I’m all for identifying accident prone locations and taking actions to reduce accidents.. there are some places that do confuse people.

    One of them is left turns against oncoming traffic.

    Until recently VDOT was using a green arrow to indicate a protected turn and a green ball to indicate a yield to oncoming traffic turn – which even if a sign saying full-green yield to oncoming traffic – we’ve seen that really bad accidents.

    But now I notice that on some lights, VDOT is using a yellow blinking arrow – which ought to be more intuitive to people especially the elderly and folks who are of other cultures and languages.

    but once we get away from auto traffic – and to ped and bike traffic – the better entities are the cities and towns – and Arlington County.

    VDOT tries sometimes but at the end of the day – good, safe bike and ped are expensive..despite Bacon’s claim.

    anyone who thinks – a pedestrian can run across 6 lanes of traffic- even with a median from his hotel to a restaurant – is living on another planet.

    anywhere there is a road – there are huge risks for ped and bike…

  8. TheWalkman Avatar

    San Francisco 1906 worked because everyone was moving at human speed. The street worked for all users despite the apparent chaos.

    Unfortunately, for the past 60 years VDOT (the Highway Department) has been designing roads for fast moving cars not people.

    Things haven’t changed much.

    As we see the demand for other transportation modes, like bike/ ped, transit, even streetcars, it will be expensive to safely retrofit these roads to make them safe (and we all know that no public official will have the spine to agree to slow traffic down to safe levels.)

    Some might argue it will never work but compare traffic in downtown DC now to 1980. Build a safe, accommodating infrastructure and the users will come.

    In VA, though, look at the weak, “vision” VTRANS2035 makes to create a vision for a workable 21st Century, we will continue to head down a path to transportation chaos.

    VDOT needs to realize its real role on land use, human psychology and planning and rather than higher the next graduating class of civil engineers from VMI, spend that money on a couple of well educated planners who have the vision to stop the insanity.

  9. larryg Avatar

    Some of this comes about because of different views of what roads should be and not be – in purpose and design.

    Obviously most folks work agree it would be loony to try to add bike/ped facilities to interstate highways.

    And obviously most folks would strongly disagree that a residential subdivision should allow speeds of 50 mph.

    but then thinks get murky as the opposite goals of arterial utility vs local access move closer to each other.

    there are more than a few downtown areas that have either totally closed off the street to cars or put in roundabouts and other traffic calming structures that force cars to go very slow and gives bike and peds more, better, and safer access.

    VDOT simply believes that major arterials – primary purpose is to move not only cars – but large delivery trucks , fire equipment, etc and that bike/ped often end up degrading the other _intended_ benefits so we get advocacies like wanting to convert a designated US highway – even ones that are coast to coast – as essentially a human-speed road in some places – essentially cutting up the road for it’s original intended purpose – which is not simply a hollow designation – real designs and real dollar funding go into trying to maintain the road – towards it’s original intended purpose – of moving – not only people – but goods – like groceries, and other necessities ….

    I don’t know what the answer is and I can be influenced by arguments on both sides of the issue but I just don’t think it’s reasonable to advocate taking a busy road that has a lot of traffic on it – especially at peak hour – an re-designing it – in a way- that traffic is going to be much worse – and more dangerous. that seems to be not a win-win tradeoff..

  10. larryg Avatar

    let me also point out here – that cities and towns in Virginia control their own streets ( but not the US signed roads nor VDOT primary roads) – but all other streets in a town or city (and 2 counties) in Va are controlled largely by the local government.

    So it’s a little bit disingenuous to blame VDOT for not making major US/Va signed highway more ped/bike friendly when the locality has tremendous latitude for all other roads in their jurisdiction.

    Any road in a city/town/Arlington/Henrico that is a 600 series road is not a US or VA designated road and there is substantial latitude for the locality to re-configure/prioritize such roads for bike/ped.

    The same is also somewhat true with counties roads.. though VDOT does have a role – if a 600 series road has, over the years, become – by county choice – a major arterial or collector.

    VDOT went from an “opt-in” for bike/ped on new or improve county roads to a “opt-out” which means the bike/ped is in the design by default unless the locality opts-out…

    People do not understand who the responsible parties on and the localities are more than happy to let people be misguided as to who they should blaming or advocating change.

    VDOT is likely never going to prioritize bike/ped for US- signed or Va primary roads unless the locality insists absolutely on it – at which point, VDOT will often suggest that a local parallel road be targeted for bike/ped – on the localities dime of course.

    VDOT has a responsibility to maintain US-signed and Va Primary roads for their original purpose – of supporting the movement – of GOODs (as well as people).

    The locality has the ability and authority to pick any local (non State/Fed) roads they wish to prioritize for bike/ped.

    Like with the “stroad” argument – people do not understand the significance and purpose of roads like US 29.. or US 50 or Va 7.

    I see the advocacy of insisting that bike/ped be put on these major roads -rather than parallel roads as – hard to understand – if the goal is to provide a bike-ped alternative – to get from A to B.

  11. cpzilliacus Avatar

    James A. Bacon wrote (with emphasis added):

    One CNU session highlighted an example of “lean sprawl repair” for the Oak Hollow Mall in High Point, N.C. That project visualized transforming an abandoned mall into a business incubator with space for live-work studios, artisan workshops and a culinary institute.

    The above bothers me a lot – note – not Jim’s reporting, but the occupations that seem to be preferred by planning types. Arts-and-crafts-type work (along with dry cleaning establishments, another business that planners seem to really like) are not likely to provide gainful employment on the large scale that our nation needs, yet they always seem to appear as part of redevelopment proposals for reasons not clear to me.

    1. Pat, I think that the people working on the High Point mall make-over were attuned to the realities of the local skill sets. High Point, as I’m sure you know, is the center of what was once a thriving furniture manufacturing industry. As I recall, the “artisan workshops” involved a lot of advanced woodworking.

  12. cpzilliacus Avatar

    James A. Bacon also wrote:

    We should look to our own past, or to contemporary Europe, for alternative ways of organizing our streets.

    I have traveled – quite a lot – in Northern Europe (that’s where my parents were born), and can assure you that there is plenty of traffic engineering work (and there are traffic engineers as well), even in the generally successful downtown areas of the Nordic nations.

    larryg wrote:

    VDOT basically does what they have to do but they don’t think when bikes get to major road crossings that there are any good solutions.

    Northern Virginia has some of the nicest bike/pedestrian trails I have seen (comparing quite favorably with those in the Nordic nations), and one of the busiest, the Custis Trail, was designed, engineered and built by VDH&T (predecessor agency to VDOT) in the late 1970’s and early 1980’s. More recently, VDOT, in close cooperation with the Maryland State Highway Administration, built an all-new trail along the reconstructed Woodrow Wilson Bridge between Alexandria and Oxon Hill, Maryland – a trail that sees plenty of use by both pedestrians and bicyclists, even though the distances between “stops” on the trail is rather long.

    1. larryg Avatar

      I like the separate trail idea much better than trying to force-fit bike in areas where it’s really problematical. I favor the trails and then strategic connections to places without trail …

      I think you can build a “smarter” bike network by using existing parklands and riparian and water/sewer/power line corridors and yes parallel streets to main thoroughfares.

      this pedestrian bridge is in Fredericksburg over Route 1 – it connects the main Campus with rental dorms and a shopping center with restaurants, grocery, bank, etc…

      it meets VDOT standards for height even though it does not look it.

      it was not cheap. Not sure I’ve seen anything like it in Cville.

  13. cpzilliacus Avatar

    larryg wrote:

    it meets VDOT standards for height even though it does not look it.

    Reminds me of this nice bridge, also designed and built by VDOT, to cross very busy U.S. 50 (Arlington Boulevard) in the Seven Corners area of Fairfax County.

    1. larryg Avatar

      thanks. I think it’s that kind of infrastructure that you have to be thinking of for crossing major highways as opposed to the idea that the highway itself needs to be reconfigured to handle bike/ped.

      People don’t pay attention to the ‘US” part of highway designations but those roads are different from other roads – and have specific purposes and standards.

      what’s happened over time – is that towns and cities – have chosen to target their “places” to these highways… and that sets up opposing philosophies of what the road should be – based, in part, on ignorance about the purpose of US-signed highways to start with.

      To be fair and honest – when the US designated highways were set up – there was significant lobbying going on by cities who wanted the US routes to come through their cities – to bring the road users into their towns for services.

      The highway folks knew that taking the US highways through downtown city streets (as opposed to bypasses or beltways) was problematical but back in those days they were confident that they could impose access-management standards (via highway funding with “strings”) to maintain utility but over time – with localities making land use changes – many of the urban US routes ended like like US 50 in NoVa – a road no longer truly part of a nationwide network…but outside of NoVa – Route 50 is still a major route – both west and east – east it becomes the Chesapeake Bay-Bridge to the eastern shore.

      VDOT – and other DOTs have – over and over tried to convince localities to build parallel roads for commercial and bike/ped and maintain the US route if there are no alternative routes. In NoVa – you do have the beltway. In Cville – there are no real alternatives…and both sides claim US 29 for totally opposite purposes…

  14. larryg Avatar

    I continue to think there is a disconnect, a lack of understanding as to the original purpose of US-signed and State-signed (in Virginia – Primary) roads.

    Both were designed explicitly for automobiles and specifically to be capable of moving not only passenger vehicles – but commercial vehicles that would be carrying all the goods that any location would need.

    You cannot accomplish that goal if you take parts of those roads and turn them into what is essentially local-access roads – replete with bike and pedestrian facilities.

    If you go outside the urban areas on roads like US 50 or US29 or VA7 – you’ll not normally see ped and bike facilities on those roads – because those roads were never intended to transport pedestrians and bikes but rather cars – and 18-wheelers.

    I’m not saying that people don’t ride bikes on them – they do – and VDOT accommodates them PRIMARILY by putting up “share the road” signs with bikes on them – even though there are no added lanes for the bikes – out in the hinterlands… UNLIKE some roads out west in places like Colorado which have miles and miles of bike lanes on major highways.

    why this is – why VDOT does not put bike facilities in on these primary roads – at the same time they are putting up “share the road” signs seems contradictory.

    but if we start cutting up the primary roads – many places that depend on trucks to supply their needs – are going to be affected – in fact already are.

    I’m not advocating that we not have bike or ped facilities. I would like to see more of both – and more safety incorporated into both – in places where doing those things – will not essentially degrade the other intended, original purpose of those roads.

    In urban areas – in Va – it’s truly up to the locality which has far more influence and even a level of autonomy than the counties do in Virginia.

    The place where VDOT will assert itself is where US-signed or Va primary roads are involved – and they do this not to purposely harm bike/ped but to preserve and maintain a level of utility that was originally in the reason for the road to start with – and they do that also with regard to commercial development – they use access management to prevent commercial development from taking over the road completely and essentially severing it as a network road.

    in some places, where there are interstates and bypasses – these roads can essentially converted from their original intended purpose to local access with more/better bike/ped… but in places where the only way “through” is that US or Va-signed road … it’s not a solution – it’s trading one thing for another – and there is a cost – you trade “connectability” for local access and nowhere in Va is that juxtaposition more apparent than US-29 in Charlottesville. VDOT was willing to run US-29 business lose – let it be used for local access – if they could get a bypass.

    well they lost – but looks what’s happening now.. US-29 is going to be rehabilitated to gain back some of it’s original purpose – and local access is going to be changed to grade-separated intersections – with ramps – which are not the best for bikes/ped either.

    where am I wrong on my viewpoint here?

Leave a Reply