Public Health Experts Say Narrow Lanes Should Be the Default on City Streets

(Source: Johns Hopkins Bloomberg School of Public Health.)

An expansive new study from the renowned Johns Hopkins Bloomberg School of Public Health adds to the body of evidence that wide lanes on urban streets promote more crashes, while narrow lanes—as narrow as nine feet—are safer. The jury is in. It is long past time to make this our default approach whenever a street is restriped or repaved.

It’s especially important that it be the default, suggest the study’s authors, because street and highway agencies so often act on autopilot, preferring rote standards to context-sensitive approaches. Engineers are allowed (and, in fact, ethically required) to use their discretion in designing roadways. But most often, our policies treat an 11- or 12-foot lane as the default, and allow the engineer to make a case for a narrower lane width.

This is backwards. A 10-foot lane, at the widest, should be the “default setting” for any sort of urban street: a place lined with homes and businesses, where traffic should flow slower than 35 mph. This should be understood as the risk-averse approach. If the engineer wants to make the lanes wider, they should have to justify the choice.

This is still a controversial change. It shouldn’t be: all the evidence points to the safety benefits of narrow lanes and traffic calming. But conventional U.S. engineering “wisdom” has said the exact opposite, for decades.

That fact ought to be a huge scandal. To put it bluntly, an idea that has been at the foundation of 70+ years of American street engineering is routinely and consistently contradicted by empirical evidence. And its consequences are a disaster.

That disastrous idea is “forgiving design,” the belief that by designing streets to give drivers more physical margin of error, we can improve traffic safety. It’s an approach that has merit in the context of open roads and freeways, where the environment facing drivers is simple and the design priority is the smooth, fast flow of traffic.

But American engineers have taken a highway design principle and applied it to hundreds of thousands of city streets, resulting in crash injury and death rates that dwarf those of many of our peer countries around the world.

Forgiving design is more than just wide lanes: it’s clear zones at the side of the roadway, wide curb radii at turns, any feature that increases a driver’s room for error and sense of comfort. The problem is simple: drivers take that room for error and behave in riskier ways, mainly by driving faster.

Wide lanes are the central element of forgiving design that the new Johns Hopkins study puts in its crosshairs.

Up to 1.5 Times More Crashes When Lanes Are Widened to 12 Feet From Nine

The study team, led by Dr. Shima Hamidi, PhD, analyzed a random selection of 1,117 street segments out of a pool of more than 7,600 in seven cities: Dallas, New York City, Philadelphia, Salt Lake City, Miami, Denver, and Washington DC. They cataloged design attributes of those street segments, including number of lanes, the presence of medians, and on-street parking. They then used regression analysis to test the relationship of lane width to crash rates, independent of those other factors.

The researchers found a significant increase in crashes—approximately 1.5 times higher—when the lane width increases from nine feet to 12 feet. There was no significant difference in crash rates in very low speed zones of 20–25 mph. But in the 30–35 mph range, “traffic lanes with 10-, 11-, and 12-foot lane widths have significantly higher crashes than lanes that are nine feet wide,” according to a summary of the study published by Johns Hopkins.

The study strongly recommends a default width of no more than 10 feet for streets intended for speeds of 35 mph or under. Priority should be placed on making this change on urban streets in the 20–35 mile per hour range without significant bus or truck traffic. Where many large trucks are present, 11-foot lanes might be appropriate: this can be a context-specific consideration.

Why DOTs Don’t Make More Progress on This Issue

The study authors also conducted a national survey of American Association of State Highway and Transportation Officials (AASHTO) committee members, working for a large number of state DOTs. What they found is illuminating as to the barriers to redesigning urban streets under state jurisdiction to make safety a top priority.

All respondent DOTs—100%—indicated that they have a design exception process where lane width reductions can be proposed, reviewed, and approved. But barely over half of the survey participants (53.8%) reported that a lane width reduction project was underway or completed in their jurisdiction, and a full 30% reported no such projects at all. Surveys revealed a large amount of uncertainty about the effects of such projects where they have been implemented, often due to a lack of follow-up data.

Existing provisions for narrower lanes are rarely used. Vermont is the one state whose standards already allow for nine-foot lanes in some circumstances—state roads through small towns that, as is typical in New England, are often quite built up and urban in form. However, the engineers from Vermont did not report a single example of the use of nine-foot lanes in practice in a new build or reconstruction project. Fear of liability was cited as their main concern.

Surprisingly, it was Florida, roughly the stroad capital (and pedestrian death capital) of the universe, where the state DOT has become a leader in revising street design practices. In 2022, FDOT adopted a context-sensitive design manual that sets a default minimum lane width of 10 feet in a wide range of urban and suburban contexts where the design speed (the intended travel speed) of the road is less than 40 miles per hour.

FDOT uses a context classification system to define a range of urban vs. suburban vs. rural environments. This gives engineers confidence that they can use a standard appropriate for the context of a street, without fear of liability or professional consequences if they do not adequately justify their choices.

At the same time, FDOT’s new manual encourages that design of these streets use a holistic approach, where the context is used to determine a safe and desired speed, and then multiple traffic calming techniques, not just narrower lanes, are used to slow cars to the desired speed. The long list of these techniques includes traffic diverters, pedestrian islands, and bulb-outs, as well as psychological things like the use of street trees and terminated vistas to shrink a driver’s field of vision.

At Strong Towns, we think this is a very positive step. We’d go even further. Street design should always start with non-engineering considerations: what are the priorities of the community that has to use a street every day? What local knowledge do they have about where and how people will interact with the street? Ultimately, what are our values: safety first, or speed first? We’ve seen time and time again that the people who actually live and work next to a street will consistently choose safety.

At a certain point in the design process, the engineer will enter the picture. And when they do, we must make it as easy as possible for them to implement a design that puts safety over speed. This should be the default, never the exception.

---

---

RELATED STORIES

---

Change everything.

Here’s the program that will revitalize your city.

---

Sign up for emails to get more of our top content:

Email address*

Postal code

We use your postal code to send you the occasional email letting you know about Strong Towns events in your region or city.

SUBSCRIBE

Thank you! We’ll add you to the list.