Blog

The evidence didn't change. The website did.

In July, the Federal Highway Administration quietly removed five items from its list of Proven Safety Countermeasures. The list went from 28 measures to 23. The five that disappeared:

• Road diets (roadway reconfigurations)
• Bicycle lanes
• Appropriate speed limits for all road users
• Speed safety cameras
• Variable speed limits

There was no announcement. No press release. Reporters found it by comparing the live page against an archived version.

If you work in traffic safety, you already know why that list matters. It isn't regulation — it doesn't control a dollar of funding directly. What it does is settle arguments. When a city engineer proposes narrowing a four-lane arterial and a council member asks “is this actually safe, or is this a bike thing?”, the Proven Safety Countermeasures list has been the answer. It's federal. It's evidence-based. It ends the conversation.

That backstop is now gone for five of the most contested tools in the field.

What the research actually says

The frustrating part is that nothing about the underlying evidence changed in July.

Road diets have shown crash reductions between 19 and 47 percent in the studies that put them on the list. Adding a bike lane cuts crashes by up to 30 percent on a two-lane road, and up to 49 percent on a four-lane road. Converting a painted lane to a protected one with delineators reduces bike-vehicle crashes by 53 percent. Fixed speed cameras have cut crashes on urban arterials by as much as 54 percent. In New York City, speeding at camera locations dropped roughly 94 percent since the program started in 2014.

Those numbers came from FHWA's own published research. They're still true this week. They were true last week, when the agency was still hosting them.

The other side of it

USDOT frames this differently, and it's worth stating their position accurately rather than caricaturing it.

The department describes the change as part of an ongoing administrative review to ensure guidance aligns with current DOT policy and the administration's priorities. A spokesperson pointed out that roughly 80 percent of Americans drive to work and to medical appointments, and said the department is reviewing federal guidance to make sure families have the infrastructure they need. Transportation Secretary Sean Duffy has been direct that he views a category of bike and climate projects as misallocated spending, and in July redirected $1.73 billion in grants toward roads and bridges.

There's a real argument inside that — about who a corridor serves, about whether federal guidance had drifted toward one set of modes, about local versus federal authority over street design. Those are legitimate policy questions and reasonable people land in different places on them.

The problem is that removing a measure from a list titled Proven Safety Countermeasures isn't a statement about priorities. It's a statement about evidence. And the evidence didn't move.

What this means if you're the one at the podium

Your projects don't get less safe. Your arguments get harder. The design still works. What you've lost is the shortcut — the ability to point at a federal page and move on. You'll now be re-litigating settled questions in public meetings, and you should expect to.

Build your own evidence file. NACTO has republished the original FHWA documents for all five removed measures — the research briefs, the crash modification factors, the design guides. Download them. Local sources matter too: Boston's bike lane crash analysis, Seattle's speed limit results, DC's before-and-after evaluations. A local number beats a federal citation in a council meeting anyway.

Watch what cities do, not what changed on a website. Bellevue, Washington published its safer speed limit work in early August. Austin cut speeds by 20 mph near a neighborhood park. Honolulu built a roundabout that improved safety and stormwater at once. Portland is connecting neighborhood bike routes across busy arterials. None of that paused in July.

Expect the fight to be about lanes. Four of the five removed measures are, at bottom, about the same thing: how much space goes to vehicle throughput and how fast that space invites people to drive. That's the whole Vision Zero argument compressed into a right-of-way cross section.

The through-line

This is exactly why the Vision Zero Safety Series exists. Federal guidance can be edited overnight. What can't be edited is what a practitioner in Durham or Richmond or Arlington has actually seen work on their own streets, in their own political conditions, with their own council.

That knowledge lives with the people doing the work. It's more durable than any list.

Try it yourself: the Street Lab on our homepage lets you build a corridor curb to curb and see what your choices cost. Give four lanes to cars on an arterial and watch what happens to the crossing distance. It's the same trade-off every one of these five measures is really about.

The Dutch didn't get lucky. They changed course.

The most common thing said about Dutch cycling infrastructure is also the most wrong: that it works because of culture. That the Netherlands is flat, small, and full of people who simply like bicycles — so nothing there transfers here.

Every part of that is false, and the history is worth knowing precisely because it's so often used to shut down the conversation.

The Netherlands built for cars first

Through the 1950s and 60s, Dutch cities did exactly what American cities did. They widened roads, demolished buildings for parking, and pushed cyclists to the margins. Cycling's share of trips collapsed. Amsterdam had active plans to fill canals for highways.

Then traffic deaths climbed, and in 1971 more than 400 of the people killed on Dutch roads were children. A campaign formed under the name Stop de Kindermoord — stop the child murder. It was deliberately blunt. Combined with the 1973 oil crisis, it forced a genuine policy reversal.

The infrastructure people point to today is the result of fifty years of sustained, expensive, politically contested decisions. Not culture. Choices — the same kind of choices available to any city council in America.

What they actually figured out

Separate by speed, not by mode. The organizing principle isn't "bikes get a lane." It's that when speeds and masses differ significantly, users get separated; when they're similar, they can mix. A 20 mph street can be shared. A 40 mph arterial cannot, no matter how much paint you apply.

Sustainable Safety. The Dutch framework predates Vision Zero and shares its logic: roads should be self-explaining (their design tells you how to behave) and forgiving (a mistake shouldn't be fatal). If drivers are speeding, the answer is that the road is asking them to.

The protected intersection. Most North American protected bike lanes still dump riders into mixed traffic at the intersection — which is where most crashes happen. The Dutch solution uses corner refuge islands, setback crossings, and forward stop bars to keep separation through the conflict point. This is the single most transferable piece of Dutch design, and it's now in NACTO's guidance.

Networks, not segments. A protected lane that ends after four blocks doesn't produce four blocks of cycling. It produces almost none, because people plan whole trips. The Dutch built connected networks; American cities often build disconnected demonstration projects and then conclude demand isn't there.

Woonerf. The "living street," where pavement, planting, and furniture are arranged so cars move at walking pace and the street is legible as shared space. No enforcement required — the design does the work.

The honest counterarguments

Dutch cities are denser, older, and more compact than most American ones. Trip distances are genuinely shorter, which makes cycling viable for a wider range of people. The Netherlands also has national funding structures and a governance model that make sustained investment easier than in a country where street ownership is split across city, county, and state agencies.

Those differences are real, and pretending otherwise costs you credibility. But they're arguments about pace and degree — not about whether the design principles work. Geometry doesn't stop at a border. A protected intersection reduces conflict in Atlanta for the same reason it does in Utrecht.

Where to see it for yourself

The Learn more section has the Dutch Cycling Embassy, CROW's design manual, and BicycleDutch — a channel that has filmed the same intersections before and after reconstruction for over a decade. Watching a junction get rebuilt is more persuasive than any statistic.

And if you want the argument in book form, the Bruntletts' Building the Cycling City documents the reversal directly, including what happened when five North American cities tried to copy it.