Killed by a Traffic Engineer: Shattering the Delusion that Science Underlies Our Transportation System

by Wes Marshall

The Vietnam War lasted from 1954 to 1975, and 58,209 Americans died in that conflict. I remember watching TV shows like The Wonder Years and thinking about how horrible that time period must’ve been. And it was, but over that same time span, drivers killed 974,567 people on US roads by accident.b These aren’t accidents; these are results.

In truth, more Americans died on US roads than in all US conflicts combined, including the American Revolution.

The Centers for Disease Control and Prevention (CDC) regularly points to traffic deathse as the number one cause of unintentional injury death for all Americans aged 4 through 34.2 This isn’t new. I found a 2002 paper saying that “in the US, injury, both intentional and unintentional, is the leading cause of death from age 1 to age 45. Because it so disproportionately strikes the young, it is also the leading cause of lost years prior to age 75, more than either cancer or heart disease. Motor vehicle injury is the largest single component of these losses.”3 I also found a 1983 paper lamenting that “automobile crashes are the leading cause of death for all persons between the ages of six months and age 47.”4 And a 1965 paper warned us that “traffic accidents are the leading cause of death among all persons from age 5 to 31!”5

million years of life lost annually on our streets.f WHO also estimates that for every road death, there are at least 10 to 15 people hospitalized due to road crashes.7 This means 13 million to 20 million hospitalizations and millions more that probably should’ve gone to the hospital.

In other words, some of early decision makers were more concerned with selling cars than with, for instance, accurately assessing safety.

subsidy. Here are some of my favorite least favorite examples of engineer speak. “Safety is a shared responsibility.” When the public tells the engineer that they don’t feel safe crossing the street here or bicycling there, traffic engineers will remind everyone that safety is a shared responsibility or that personal responsibility needs to play a bigger role. In other words, they don’t plan to fix the problems with design changes; they just want everyone to follow the rules we’ve laid out for them. I’ve found versions of this almost as far back as the traffic engineering literature goes. A 1955 paper tells us that “safety is everybody’s problem” and that “if each accepts his individual share of the responsibility, the task [of improving road safety] will be easy.”11 Some of today’s more progressive traffic safety movements use this phrase as well. They want “safety is a shared responsibility” to mean that engineers and policy makers need to share in this responsibility of creating a safe transportation system. Unfortunately, traffic engineers and policy makers continue to co-opt it in a way that shifts blame onto road users. “We need to take a balanced approach.” In many cases, the public wants more space for pedestrians and bicyclists. Or they want the engineer to help slow cars to make streets safer for other modes. The engineer reminds the audience that we need to balance everyone’s needs, which in theory means that we need to find an appropriate balance between things...

“The standards say . . . .” This is a big one. Traffic engineers like to tell the public that whatever the public is proposing won’t work because the standards say this or the standards say that. The truth is that they are referring to guidelines, not standards. Other than the Manual on Uniform Traffic Control Devices (MUTCD) and the ADA Standards for Accessible Design, none of the guidance encased in our massive national manuals should be considered “standards.” Even the MUTCD—often used to tell the public that a crosswalk or pedestrian signal isn’t warranted—gives the engineer discretion to overrule the infamous warrants with engineering judgment.g If

The traffic engineers already aware of this disconnect are guilty of implying that safety is built into our so-called standards when, in reality, “knowledge of the crash frequency and severity consequences of design decisions played no discernible role in the formulation and in consecutive revisions of important design standards and procedures.”2

were not part of our process. Despite manuals lacking the requisite science to back them up, traffic engineers put their faith in these manuals and adhere to them. Traffic engineers also don’t take too kindly to those criticizing these holy texts. It wouldn’t be hard to argue that this sounds more like a religious cult than a science-based discipline. The unfortunate reality? The system isn’t broken; we built it this way.

who don’t do anything about it.” Now that we know better, traffic engineers have become the not-so-innocent bystander. If traffic engineers continue to conduct business as usual, more people will die on our streets, indirectly killed by the generations of traffic engineers who came before them.

Driver education tends to be tied to driver’s licensing programs, which date back to 1903 in Massachusetts. By the 1930s, about half of US states had them. By the 1950s, Wyoming was the lone holdout. Heck, we didn’t even share driver records between states until 1961. As a 1949 Traffic Quarterly paper points out, “Driver-licensing is regarded primarily as a revenue-producing measure.”1 It isn’t much different today.

it? In the same 1958 report, AAA says that “it is time that we become concerned with pedestrian violations and unwise walking practices” and then highlights all the progress on this issue in cities like Detroit, which arrested 19,765 pedestrians for crossing against the signal but only 8,662 drivers for violating the pedestrian’s right-of-way. The report noted that San Francisco arrested 165 pedestrians for crossing between intersections as compared to 7,304 drivers arrested for violating the pedestrian’s right-of-way. But don’t let the numbers fool you; San Francisco also arrested 32,968 pedestrians for public intoxication.5

20 thousand people??? ever? per year? wild, either way. slave patrollers gonna patrol, eh?

are hindered and delayed by traffic congestion”).7 In the 1930s, Paul Hoffman urges “better enforcement of legal restrictions upon pedestrians” and for “cities to arrest and convict pedestrians.”8 Midcentury Portland, Oregon, also decided that the best way to “protect its pedestrians” was to arrest 15,196 of them for crossing against the signal and another 9,923 for crossing between intersections. In a 1952 Traffic Quarterly paper, the mayor of Portland says that starting in early 1945, “police officers began to make straight arrests for pedestrian violations.” What do they mean by a violation? They mean everything because “there is no such thing as a ‘minor’ traffic violation.” But guess what? The mayor then tells us that “the city of Portland has nearly 100 percent pedestrian observance of traffic signals.”9 What happened to traffic fatalities in Portland? Between 1944 and 1945, they increased by more than 50 percent, from 46 deaths to 70.

5, but for now, I want to introduce one of the most insidious offenders of this mentality: design speed. What is design speed? It represents the theoretical safe limit that we are designing the street to, akin to a design load for a bridge or building. For most of the years since the design speed concept originated in the mid-1930s, we defined it as the “maximum safe speed” for a given section of roadway. The idea behind design speed is simple enough. Let’s say I’m trying to design a major street where I’m expecting the speed limit to be 30 or 35 mph. As a designer, I’m aware that some people will drive faster than that. My guidelines3 tell me that I need to make sure those people are safe too: “The selected design speed should fit the travel desires and habits of nearly all drivers expected to use a particular facility.”e That is good to know, but what does it mean for design speed? What the guidelines say is that other than on local streets, “every effort should be made to use as high a design speed as practical.”4 So after setting the design speed as high as possible, I then design everything to make sure people can safely drive at that speed.

exact design speed either, but they adapt to meet it.f When the design speed is higher, drivers drive faster. When drivers drive faster, we lose more lives on our streets. Killed, albeit indirectly, by a traffic engineer.

To get a Professional Engineering license, you take the Professional Engineering exam. It includes a breadth section, which covers the gamut of civil engineering, and a depth section. Even though my undergraduate concentration was structural engineering, most of my professional experience leaned toward transportation, so I decided to take the Professional Engineering licensure test with the transportation depth. When I received my Professional Engineering license in 2003, I had taken a grand total of one college course specific to transportation. That’s it. But I was professionally licensed to design any street or freeway in the state of Connecticut.

A 1971 Traffic Quarterly paper makes it clear that “since death rate is dependent upon motorization, it can be reduced simply by increasing the number of vehicles on the highways. The formula leaves no room for any other conclusion.” Accordingly, a city like Driveton “could expect to lower its death

AASHO’s 1957 manual gets a bit more specific about the causes of urban safety problems. It says, “Pedestrians and congestion, augmented by intersectional and roadside interferences, are chief causes of urban accidents.”3

district have come about in the form of blighted areas,

existing one. The city as a whole has suffered.”7 In a 1964 Traffic Quarterly paper, Boston traffic engineer William McGrath—the namesake of a big, bad arterial that runs through Somerville, Massachusetts—casually refers to the work that we do as slum clearance.8 Another 1964 Traffic Quarterly paper then argues that we need to evolve past the “limited concept of slum clearance” and start looking for “blight in non-residential structures.”9 How do we measure nonresidential blight? According to the American Public Health Association, traffic engineers should be on the lookout for a “lack of off-street parking.”10 A 1972 Traffic Quarterly paper even weaponizes our work: “Modern transportation . . . has provided an important weapon for attacking slums.”11

they told us we should envy. In the 1939 report Toll Roads and Free Roads, MacDonald and the Bureau of Public Roads called “the whole interior of” Baltimore a city that is “ripe” for these changes.12 Years later, a 1967 Traffic Quarterly paper brags about the plan to put 20 miles of freeways right through Baltimore’s slums. “Good news! An enormous stride has been taken toward creating heaven on earth.”13 That’s a bold statement Baltimore. Tell me more. “An examination of the present condition of the areas under consideration for a right-of-way shows that they are in a deplorable mess, . . . host to more thriving slums and crime than many areas now classified under urban renewal projects.”14

ones.” Slums weren’t the only problem in Detroit; it was also the street network: “The street layout established in the urban renewal plan will consist of fewer, but wider, streets—because the project will provide a large amount of public and private off-street parking space—

traffic.”15 What’s the solution, Detroit? “Many projects planned for residential redevelopment will wipe out obsolete gridiron patterns and replace them with curvilinear systems such as are found in the best of suburban subdivisions.”16 But as a 1963 Traffic Quarterly paper tells us, obsolete, gridded street networks can be a blessing in disguise: “It is obsolescence, however, that can be a blessing, an opportunity to convert whole areas to the needs of man in the large modern city.”17

there any specific projects you want to mention? “Detroit’s Gratiot project will convert a slum area of about thirty small blocks into a ninety-acre superblock penetrated at intervals by cul-de-sac streets.”18 What about the people who live there? Not long before, “the customary acquisition procedure was to make an award of only one dollar at the time of the taking.”19 But in ...

of asthma, lung disease, heart disease, and cancer. If we dig into New York City’s history of big roads to the time when Robert Moses was still in college, you’ll come across Madison Grant. Grant was not a traffic engineer, but he was behind the construction of the Bronx River Parkway that opened in 1908. The parkway was sold to the public as an “environmental improvement,” but soon thereafter, Grant published a book called The Passing of the Great Race that became one of Adolf Hitler’s favorite books. What happened next? As L. J. Aurbach explains, “Grant became an internationally known leader of the white-supremacist and anti-immigration movements, and one of the most prominent racists in America.” Unfortunately, Aurbach continues, “Grant’s fellow parkway commissioners echoed his bigoted sentiments,” and they took pride in their parkways cutting through low-income, immigrant neighborhoods.8 They

wasn’t wrong.b Future senator Moynihan scoffs at the notion that we didn’t know what would happen to these cities: “It is not true, as is sometimes alleged, that the sponsors of the interstate program ignored the consequences it would have in the cities. Nor did they simply acquiesce in them. They exulted in them.”6 The 1956 Clay Commission Reportc that President Dwight D. Eisenhower forwarded to Congress praises urban freeways for helping “to create a revolution in living habits.” Because of these urban freeways, “our cities have spread into suburbs, dependent on the automobile for their existence.”7

limit” designation.8 For the sake of an example, let’s say we combine “exceeded speed limit” and “too fast for conditions” into a single speeding-related variable. Over a recent six-year period, 28,642 pedestrians lost their lives on streets with a posted speed limit.9 How many of the drivers who killed these 28,642 pedestrians were speeding? According to FARS, the answer is 2,015. Go look at almost any road. Does it seem within the realm of possibility that the speeding problem is limited to 7 percent of drivers? Of course not, but somehow we are supposed to put faith in data that tells us that drivers were speeding in only 7 percent of pedestrian deaths.e

have a case. A lawyer who used to prosecute organized crime wrote a 1968 Traffic Quarterly paper complaining that there is less justice in transportation than organized crime. He says that “speed and right-of-way violations lead the list of those offenses which have a causal relationship to accidents resulting in personal injury. Yet most of the persons cited for such offenses never see the inside of a courtroom.”8 The author is flummoxed as to how “someone can run a red light, hit a pedestrian in the crosswalk, totally disable the victim, and never even have to make an appearance in court. The offender simply writes out his check for $10 which he posts as bail and then forfeits.” He goes on to say that “of course if the unfortunate pedestrian dies, the motorist will be charged with vehicular manslaughter, but it takes a homicide to get you into court unless the underlying moving violation is an extremely serious one.”9

But in Maryland, the consequences for the driver have nothing to do with the outcome for the pedestrian. According to a 1992 Transportation Quarterly paper, “The motorist who goes through a red light and kills someone pays the same fine as if no injury had occurred.”10 It then gives an example of a 1987 case where “a driver who killed a blind couple and their guide dog” paid a $40 fine and got three points added to their license.

created negligently, over time, became negligent. Given how many of our streets predate any and all industry standards—as well as traffic engineers altogether—you might think the lack of standardization left us with streets that were dangerous when built and that have only become more deadly over time. But the opposite is usually true. Streets built before the advent of traffic engineers are some of our safest, often with far fewer fatalities and severe injuries than the new-and-improved, fully traffic-engineered versions. Find yourself an example of an old, narrow street in your favorite part of any older city. Watch how people use this street. Take some measurements. Then ask a traffic engineer if they’d design a “substandard”—yet safe—street with the same dimensions. Most traffic engineers wouldn’t go anywhere near this because “tort liability arises when substandard designs results in accidents.”13

Even today, decent sidewalk data is hard to come by in most cities. Back in 2009, Denver’s sidewalk data was a shocking find. But the sidewalk data had not been updated since it was first collected by the city in 2003. I asked somebody with the city why this information wasn’t being updated regularly like everything else. The reply? It’s not allowed. The city’s lawyer said not to. Wait, what? The lawyer’s thinking? If the city knew there was a problem, it would be liable. If the city stuck its head in the sand, it could feign ignorance. How can it possibly fix a problem it doesn’t know it has? So the city looked the other way on sidewalks while averaging nearly 250 pedestrian injuries each year. Unfortunately for the city, the old ostrich trick doesn’t work. Nor does trying to pawn sidewalk maintenance onto the adjacent property owners.b Like many cities with similar sidewalk policies, Denver was sued under the Americans with Disabilities Act. The city settled the case, promising to install at least 1,500 new sidewalk ramps each year for as long as it takes.

Why is that the case? Many states have open range laws, which means that ranchers don’t have to keep their cattle off the road. Not only can’t animal owners be held liable if somebody drives into their cows, some states—including Virginia—force the driver to pay the rancher damages. If I’m driving and hit a cow, I’m at fault. If I hit a pedestrian, the pedestrian is.

the scofflaw pedestrians and bicyclists. But here’s the thing. We know that drivers running red lights is a massive factor in crashes that result in an injury or fatality.7 Waymo—or the artist formerly known as the Google Self-Driving Car Project—even had to program its autonomous fleet to pause for a second or two after a light turns green because actually going when the light turns green turned out to be the biggest crash risk. But the connection between safety outcomes and scofflaw pedestrians/bicyclists? It’s dubious at best. Here is Kenneth Todd in a 1992 Transportation Quarterly paper about pedestrian safety: “The notion that compliance with the law saves lives is a myth.”8 Other researchers agree with him that “the commonly held notion that lawful behavior equals safe behavior is not borne out by statistics.”9

The Netherlands made it legal for pedestrians to jaywalk in 1997. Back then, more than 130 pedestrians lost their lives each year. The population of the Netherlands has since grown by more than 10 percent, but now the Netherlands sees fewer than 50 pedestrian deaths annually.1 Transportation infrastructure in the United States isn’t quite the same as what you might find in the Netherlands. But another interesting difference is how the Netherlands handles assigning blame in a crash, as John Pucher and Lewis Dijkstra explain in their 2000 Transportation Quarterly paper: “Even in cases where an accident results from illegal moves by pedestrians and cyclists, the motorist is almost always found to be at least partly at fault.” And when the pedestrian involved in the crash is a kid or an older person, “the motorist is usually found to be entirely at fault.”2 Pucher and Dijkstra go on to say that “in almost every case, the police and the courts find that motorists should anticipate unsafe and illegal walking and cycling. Having the right of way by law does not excuse motorists from hitting pedestrians and cyclists.”3 In terms of liability, how do insurance companies in the Netherlands handle crashes between cars and pedestrians or bicyclists? “The insurance company for the motorized vehicle automatically pays the damages, regardless of guilt. The only exceptions are cases where the pedestrian or cyclist can be proved to have deliberately caused the accident.”4

to become a squeaky wheel issue. If your neighborhood has the wherewithal to set up petitions and complain to the right people, you now have a shot at seeing one of our safer designs. If your neighborhood doesn’t have that sort of time or political clout, tough luck. The reverse is also true—and to a greater extent. Sometimes, all it takes is a few vocal constituents to squash a safer design. Given some of the atrocities perpetuated by traffic engineers in the past, taking community input seriously makes sense. But safer streets shouldn’t be subject to a popularity contest. Opting against a safer design after a few drivers grumbled about it forcing them to drive slower? Canceling a protected bikeway after a few neighbors moaned about losing some on-street parking or the perceived aesthetics of the street? Such decisions are common for traffic engineers, and they should invite a liability problem. Even an unfrozen caveman lawyer could argue that basing what gets built on the whims of a few neighbors—instead of a “rational process”—could fall under the umbrella of “arbitrary” or “capricious.”

signs, placed by the powers that be. The Colorado Department of Transportation (CDOT) has an official program for these memorials. Once you die on a Colorado street, your loved ones can fill out the appropriate form. If approved, CDOT will take their $100 and fabricate, install, and maintain the sign for six years. The bottom half of the sign will say “IN MEMORY OF,” followed by your name. The top half of the sign varies. Based on the DOT form, your loved ones will have four options. With a driving under the influence conviction or a toxicology report, they could opt for the “DON’T DRINK AND DRIVE” sign. If you weren’t wearing a seat belt, the “PLEASE BUCKLE UP” sign becomes an option. If you were riding a bicycle or motorcycle, your family could select a “PLEASE RIDE SAFELY” sign. For any other fatality, or for any of the above, the “PLEASE DRIVE SAFELY” sign is the default. One of these signs used to sit at the corner of Leetsdale Drive and Monaco Parkway in Denver. It commemorated Amelia Bates, a 15-year-old high school sophomore who was killed there in 2008. George Washington High School sits near one corner of this intersection. On the opposite corner is a grocery store. Amelia was a writer for the school’s newspaper and had walked over to the grocery store to buy a birthday cake for her editor. On her way back, Raj Sundaram hit and killed Amelia in the crosswalk as he made a right turn. The day after Amelia died, the Denver Post wrote an article about Amelia and the ...

sciences. The original intent of design speed, for example, was to mathematically link design speed with curve radius, superelevation,h and the side friction factor. In fact, the title of Barnett’s 1936 paper in which design speed was first introduced is “Safe Side Friction Factors and Superelevation Design.”5 Design speed was not the main attraction. Instead, Barnett’s goal was a “balanced design” where all these factors would work in harmony to counteract centrifugal forces and allow drivers to take corners without slowing down. The same basic superelevation equations we use today date to 1920.6 At the time, “there was considerable public hostility to providing any superelevation due to the opinion that it encouraged high speeds and reckless driving on curves.”7 Despite the lack of consensus, AASHO made up its mind in 1930 that superelevation on curves was their official recommendation. The next step was to figure out the side friction factor, meant to represent the coefficient of friction between the vehicle’s tires and the road. In 1935, the Bureau of Public Works called on volunteers to participate in a study. “Several hundred drivers” were asked to take their own cars and ride along curves of known superelevation and radius.8 These drivers were then asked to report the speed at which they felt the car begin to “pitch outward.” As you might expect from such a study design, the results were all over the map. Nevertheless, the researchers averaged whatever drivers told ...

It doesn’t have to be this way. As we know, the MUTCD gives traffic engineers leeway to use engineering judgment. If we want an all-pedestrian phase, for example, “an unacceptable number of conflicting pedestrian movements with right-turn-on-red maneuvers, especially involving children, older pedestrians, or persons with disabilities” would do the trick.9 The MUTCD doesn’t give us much to go on, so how we define an “an unacceptable number” is entirely up to us. Take the case of sixth-grader Eliza Almuina. Eliza was walking to her school carnival in Albuquerque, New Mexico, the day before school break. Doing so meant crossing Louisiana Boulevard, a high-speed multilane arterial. Eliza tried to cross at the unsignalized crosswalk adjacent to the school. Not that it should matter with a crosswalk, but it was 5 p.m., outside of normal school hours and beyond when the school zone would’ve been active. One lane of traffic stopped, and Eliza started crossing with her best friend, Hailie. The driver in the second lane didn’t stop.f Hailie made it across. Eliza didn’t. The city reacted to this catastrophe by saying they’d study the problem. They considered a HAWK beacon but didn’t find enough pedestrians crossing Louisiana Boulevard to meet the MUTCD warrant. The city installed it anyway.g Once it did, the number of pedestrians crossing there increased to where it would’ve been warranted anyway. So

Start with these facts and the keys to making a transportation system safe for kids begin to emerge. We need slower streets.h Fewer lanes of traffic. Simpler intersections with protected pedestrian crossings. Stop bars painted farther back from the crosswalk. Parked cars set way back from the corner. And not just during the fleeting minutes before and after school. We need it all the time. I came across a school zone speed limit sign from Fort Lauderdale, Florida, that lists four short bursts of time when the school zone limit applies. And another from Detroit, Michigan, that lists six separate time periods, including school zone times like 7:52 a.m. through 8:22 a.m. If you happen to arrive at school between 8:23 and 8:36 a.m., good luck, because the school zone speed limit doesn’t apply. In Denver, many school zones don’t extend to the arterial abutting to the school. Time and time again, we choose to allow the allure of speed and efficiency to override safety. This prioritization isn’t conjecture; it’s what we do. Here is a 1952 Traffic Quarterly paper telling us how a good engineer deals with the “real and imaginary fears in the hearts of parents”: “While they sympathize with parents’ fear, they also know their responsibility to taxpayer and motorist to provide an efficient, economically controlled street and highway system, free from unwarranted stop signs, traffic signals and other installations that may cause delay and congestion; or cause unnecessary expenditure of...

A 1956 Traffic Quarterly paper advises us to dismiss such fears, saying, “An influential and intelligent but difficult citizen whose 7-year-old daughter was to enter a new school . . . argued the Parent-Teacher Association into petitioning the council of the city for a fixed time signal to protect the children crossing the highway.”12 Unfortunately, this parent was only “an average road user” and not experienced with traffic problems like the paper’s author, Fred Herring. Yet, for reasons unbeknown to Fred, this parent “persisted in his demand for the fixed time signal.” Fortunately, “in the end, reason prevailed.” Peak-hour traffic was just too damn highi to help these kids cross the street, and from Fred’s perspective, “the meeting ended in complete harmony.”13 Maybe, but I’d bet this kid didn’t walk to school the next day.

In a 1952 Traffic Quarterly paper, the mayor of Portland, Oregon, brags about quelling fears in a similar situation.14 Portland had a “high-speed arterial state highway” located adjacent to its school for blind children. The blind kids nicknamed this arterial “Death Valley.” Portland’s solution? They made the pedestrian beg button bigger. Does the bigger button bring abou...

Sigurd assumed that the world where he grew up was long gone. But when he returned to Germany in the early 1990s, his hometown had undergone an extensive program that can best be described as a combination of traffic calming and living streets, or woonerfs. The idea is to design streets that force drivers to adjust their behaviors, specifically in terms of speed, and give other road users priority over cars. According to Sigurd, the result was that “to drive faster than 30 kph [kilometers per hour] is illegal, unthinkable, and not really possible.”17 That speed, 30 kph, is 18.6 mph. Imagine a world with streets where 20 mph is “not really possible”? Wouldn’t that be a safer world for kids to walk and bike? Wouldn’t that be a safer world for everyone?

The transit planners wanted to extend the dedicated bus lanes farther toward DC to shave off another 15 minutes. The traffic engineers didn’t like seeing big, empty headways between buses. Why should these buses—which represent less than 15 percent of the traffic—consume 25 percent of the space? The answer? The buses held 82 percent of the people.6 Nevertheless, they downgraded the bus-only lane into a combined bus/high-occupancy-vehicle lane. The buses lost their advantage and eventually, their riders.