The Federal Highway Administration removed five evidence-based safety countermeasures from its “Proven Safety Countermeasures” list without publishing a research review, citing contrary data, or making any public announcement. The list went from 28 to 23 items. Three of the five removed strategies remain unnamed in the public record. The two identified are bike lanes and speed cameras.

The list itself does not directly control how billions of dollars in federal highway funds are spent. But it carries significant weight with state and local transportation agencies. Josh Naramore of the National Association of City Transportation Officials described the functional loss as “taking tools out of the toolkit.” When a safety measure carries the federal “proven” label, local transportation departments use it to justify project funding and secure political cover. Its removal does not directly defund those projects, but it withdraws the federal endorsement that made them administratively easy to approve.

The governing logic at work is straightforward: the political priority of driver throughput overrides the evidentiary standard the list’s “proven” designation was designed to represent.

The FHWA’s own 2021 documentation states that speed cameras can reduce crashes on urban arterial roads by up to half, and that adding a bike lane could cut crashes by as much as 30 percent on a two-lane road and 49 percent on a four-lane road. Michael Griffith, who worked in the FHWA safety office for more than a decade before retiring in 2022, confirmed that the agency had concluded the supporting research was rigorous. “‘Proven’ is basically backed by sound research, research that we have confidence in,” Griffith said. The evidence was not re-evaluated, disputed, or superseded as part of the removal. The FHWA spokesperson’s rationale cites alignment with “current DOT policies and the administration’s priorities,” not a reassessment of crash-reduction data. The removal was a withdrawal of legitimacy, not an information update.

The structural vulnerability of the list traces to its status as non-binding agency guidance, which has allowed political directives to override established research baselines. The “Proven Safety Countermeasures” list and the “proven” designation itself are not anchored in statute and are not protected by a procedural requirement to provide evidentiary justification, public notice, or peer review before removal. The list operates as guidance, not regulation, and no evidentiary gate applies. The process was designed to add evidence, not to resist removal; it assumed the list would only move forward, never be used as a lever. That assumption rested on a shared institutional belief that any future leadership would treat the evidence base as binding, making procedural safeguards against removal appear unnecessary. The culture of evidence-based decision-making within the FHWA’s safety office was never stress-tested against an administration hostile to the bipartisan consensus that sustained it.

The apparent absence of on-record career-staff endorsement of the removal reinforces that the workforce would not have certified the change. The structural arrangement of decision rights allows political-appointee leadership to modify a technical artifact without technical concurrence.

The list change occurred alongside a thematically aligned policy environment hostile to the removed strategies. On July 7, the Department of Transportation announced $1.7 billion in discretionary grants with no funding for bike lanes or pedestrian projects. The same day, Transportation Secretary Sean Duffy wrote on X that the Biden administration had “used YOUR MONEY for DEI bike lanes and climate change.” The FHWA spokesperson told NPR that the Department of Transportation is “taking action to reverse the last administration’s policies that decreased lane capacity and increased congestion” and that drivers who pay taxes and vehicle fees expect their money to be spent on roads, “not social initiatives that burden their commutes.” Safety advocates noticed the changes to the list late that week, after the grant announcement. The temporal and thematic clustering of these actions constitutes a coherent administrative posture.

The PSC list functions as an evidentiary shortcut for state and local agencies. Its removal raises the justification burden for adopting the same strategies, even though the underlying research remains intact. What is severed is the federal endorsement that made that data actionable for resource-constrained agencies. The change re-labels which research counts as federally endorsed rather than substantively downgrading the research itself. For a local planner facing a city council, the absence of the federal “proven” label can be decisive. The burden of proof shifts from the objector to the proponent. A planner presenting a speed-camera proposal to a city council in 2027 will face: “the federal government just removed that from its proven-safety list, so why are we doing this?”

The backdrop sharpens the consequence. More than 36,000 people were killed on U.S. roads last year. Pedestrian deaths, while down from a four-decade high in 2022, remain above pre-pandemic levels. The measures identified as critical tools for state and local planners — the very measures just removed — are the ones whose crash-reduction evidence is most directly applicable to addressing these fatalities. Stephanie Pollack, who served as acting FHWA administrator under President Biden and oversaw the list’s expansion to 28 items in 2021, characterized the decision as arbitrary: “It’s hard for me to understand how you could say you’re putting safety first, and then make arbitrary decisions about what does and doesn’t improve safety.”

The list’s structural vulnerability leaves it exposed to multiple vectors of political override. The current administration has pursued ideological reclassification, labeling bike lanes “DEI bike lanes” and characterizing the removed countermeasures as “social initiatives that burden their commutes.” A future administration could pursue the same outcome through formal notice-and-comment rulemaking to rewrite the “proven” standard, or through a pure throughput-first policy pivot. All three vectors converge on the same symptom: the “proven” label detached from the evidence that earned it without any measurement-based threshold being crossed.

Until the list is statutorily insulated from administrative changes that lack evidentiary justification, public notice, and an inter-agency peer-review trail, the same pattern of removal could recur in future administrations. A potential structural feedback loop compounds the risk: if the removal of the federal designation leads to fewer local projects being justified and built, the resulting absence of local implementation data could subsequently be cited as evidence of low demand or impact, reinforcing the case for further removal. The credibility of a safety toolkit that depends entirely on persuasive federal endorsement rather than regulatory mandate has been revealed as the binding constraint.

The three unnamed removed strategies remain unidentified in the public record, limiting full assessment of the removal’s scope. Whether state or advocacy groups will challenge the removal as arbitrary and capricious under the Administrative Procedure Act is an open question. The extent to which other states or municipalities have independently adopted bike-lane and speed-camera programs that would persist regardless of federal list changes is also not publicly known.

The key questions for the next phase of the policy debate: Will the administration publish the evidence review that led to the removals, or will it continue to justify the change solely on policy grounds? Will any local agency that had relied on the list to plan a speed-camera or bike-lane project now face a successful challenge? Will the structural design of the list — a guidance document with no evidentiary gate for removal — survive the political transition that follows the current administration? Will the administration codify the list reduction through formal notice-and-comment rulemaking, or will it remain a silent webpage edit — and what does that procedural choice signal about the expected durability and defensibility of the policy?

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Red-Team Assessment
Models a capable adversary probing a plan for the seams they would exploit.
Relationship Mapping
Extracts the network of ties among people, institutions, and entities.
Root-Cause Analysis
Traces a symptom back along its causal chain to the conditions that actually generated it.