State says it has no plans to restripe street where one person has died per year

Since you’re asking specifically for feedback from a traffic engineer here I’ll take a shot at this, but I think it’s useful to first take a step back and look at what traffic engineering actually is.

One way to describe engineering might be to say that it’s the practice of applying the scientific method to try to solve a specific problem or accomplish a specific task. Note that this distinguishes engineering from pure science in a critical way: Scientists ask more general questions about how the universe works and will largely follow evidence wherever it leads, while engineers have a specific goal in mind at the outset, i.e., an agenda. Often that goal is obvious and noncontroversial, and success is easy to define. If the bridge does not fall, the structural engineer has succeeded. If the rocket deposits its payload into the proper orbit, the aerospace engineer has succeeded.

So what about the traffic engineer? What is our goal, and how do we know if we’ve succeeded?

Traffic engineering is a relatively new field that began to arise as something distinct from other branches of civil engineering during the go-go era of road building in the middle of the 20th Century. At that time, the amount of driving correlated nearly perfectly with economic output, and so the massive expansion of our auto infrastructure that occurred in the three decades after WWII was the backbone of the rapid economic growth that occurred during the era. In 1950, the first Highway Capacity Manual was published, which standardized the methodology for quantifying throughput of vehicles. Fifteen years later the first edition of AASHTO’s “green book” was published, standardizing many elements of highway design to ensure efficient traffic flow. Thus the profession was born, complete with its original sin: It would concern itself with throughput and travel times for motor vehicles, and little else. The goal of traffic engineering—maximizing throughput and minimizing delay—would remain the same and would not be seriously questioned for decades.

Today, of course, things are changing, and changing rapidly. The profession has largely rebranded itself as “transportation engineering,” and many if not most of us are asking a different and more complicated set of questions. Instead of, “How do we maximize throughput and minimize delay?,” we’re asking, “How do we maximize safety? How do we maximize the utility that a street can provide to its community?”

Of course, the old attitude is far from dead. Allowing for exceptions to every rule, the capacity-first engineers tend to be older (and thus have more seniority and more power) and tend to work for state DOT’s (which, generally, were formerly called “highway departments”) rather than cities. Exacerbating the problem, measuring capacity is a relatively easy thing to do from an analytical standpoint, especially given the 65 years of refinement and standardization of the practice. Safety, by contrast, is far, far harder to quantify and attempting as much has a far shorter history; the first Highway Safety Manual was published in 2009, nearly 60 years after its capacity-concerned counterpart.

In this context, everything coming out of ODOT makes sense and has been consistent from the get-go. ODOT continues to practice this old-school style of traffic engineering, where they’re concerned primarily or even exclusively with automotive throughput. The question they asked over the course of this study is, “How might a Barbur road diet affect capacity and delay?” This is what informed the methodology they used to collect speed data—calculating an average speed over a distance is a direct measurement of travel time. Calculating instantaneous speeds at a particular point really reveals no useful information regarding travel time. While we know that speed and crash risk/severity are directly related, and thus a before/after measurement of speeds at a given point on the roadway might be an excellent proxy for teasing out safety effects, ODOT simply isn’t interested in answering this question. This is an agency, after all, that recently claimed that road design doesn’t influence safety. The engineers there still view their job as one of maximizing throughput to accommodate economic growth. Since nearly all crashes have an element of human error, they can simply shrug and label them “accidents”—occurrences that become statistically inevitable as volumes increase.

That the study was designed to measure exclusively the delay and capacity effects of the road diet makes it harder, but by no means impossible, to use the findings to answer broader questions about the safety and utility of the street. But this clash of old-school and new-school thinking is something that plays out over and over again in any number of ways in my profession. This is what informs much of the analysis I’ve written on this (how do we do an apples-to-apples comparison of delay costs and safety benefits?) as well as my constant push-back against use of the term ‘accident’ and other symptoms of engineers being lazy in their analysis of safety implications.

With this in mind, Mikael Colville-Andersen, Jeff Speck, and their ilk have concluded that traffic engineers are The Problem, but this is simplistic. I believe that transportation engineers have a crucial and unique role to play in fixing our streets, provided we can start asking (and then seeking to answer) the right questions. ODOT has clearly asked and sought to answer the wrong ones here. While I can promise you I and many, many others are working diligently to eradicate this mindset, the fact that Barbur (and so many other streets like it) continue to claim victims while we toil is utterly heartbreaking.