Traffic Fatalities Are a Choice

America’s roads are more dangerous than those of almost every country in the developed world. We know how to change that.

by Abi Olvera

You are as likely to die driving on an American road as you are driving in Kazakhstan or Kyrgyzstan. On a per capita basis, the American traffic-related death rate ranks 87th in the world. At 12.8 deaths per 100,000 people, it is double that of Greece, triple that of Austria, and six times more than Japan. In 2022, more than 42,000 people died on American roads, and more than two million — 1 in every 170 — required emergency medical care from automobile-related accidents. The total economic cost, in medical expenses and loss of life, amounted to an estimated $470 billion.

This is not an inevitability. It is a policy decision that the federal and state governments continue to affirm. Indeed, if the United States had matched the rate of improvement in road safety since the 1970s seen in, for example, the Netherlands, Sweden, or Spain, it would have prevented 2 out of every 3 road deaths, saving 25,000 lives last year alone.

It’s tempting to argue — and some do — that such drastic reductions are easier to accomplish in small, dense, European countries. But even relatively car-centric countries like Britain, Canada, and Australia cut road deaths by nearly half between 1979 and 2002. In the United States, they decreased by just 16% over the same time.

What did Sweden, Spain, and other countries do that the US did not? Each of the countries that reduced their road deaths used the same method: they implemented a set of urban planning decisions — and a philosophy which informs them — now called the Safe System approach.

Safe System origins

At its core, the Safe System approach acknowledges that human error while driving is inevitable. Thus, road designers and urban planners should engineer environments to guide safer behaviors. Smart design ensures that when human error happens, it does not lead to severe crashes by, for example, physically separating pedestrians from high speed car traffic, and by designing roads that don’t facilitate high speeds.

The philosophy traces its origins to road safety activism that first began in the Netherlands in the 1970s. The Netherlands at that time had a road fatality rate 15% higher than that of the United States. It reached an apex in 1971, when more than 3,300 people, including more than 400 children,were killed in driving accidents — a fatality rate equivalent to 25 per 100,000 people. Out of public outrage was born the Stop de Kindermoord or Stop the Child Murder movement, spearheaded by bereaved parents and civil society organizations who demanded safer streets across the country.

Activists within Stop de Kindermoord were pragmatic. They worked with urban planners and local communities to develop innovative, safer street designs. One result was the separated bicycle tracks for which the Netherlands is now famous. So, too, was a new street design known as woonerf, or living street, which used physical elements like bends, curves, and trees to deliberately slow traffic and provided for shared space between pedestrians, cars, and bicycles.

Through continued activism — and partially aided by the oil crisis of 1974, which made driving less economically attractive — many designs were taken up by politicians who codified new rules on street design, funding, and governance. Over decades, the Netherlands continued to improve its road design, pioneering the behavioral-engineering approach. For example, the Dutch developed the concept of “self-explaining” roads, which intuitively guide drivers to safe speeds via engineering features.

Yet progress was not linear. Many road improvements were implemented in more piecemeal fashion, not country-wide but neighborhood by neighborhood, city by city. After early success, reductions in road fatalities stalled through the 1980s. Recognizing the need to catalyze further change, in the late 1980s, the Dutch States General set an ambitious goal: a 50% reduction in road deaths by 2010.

To meet those goals, the Netherlands enacted what is cited as the very first “system approach” to road safety — the Dutch then called it “Sustainable Safety” — which applied three principles to road design that look like an early version of today’s conception of Safe System: a hierarchical network of road types, each with its own design and speed limits; “homogeneity,” which uses speed management and directional separation to manage interactions between vehicles of different masses; and predictability for drivers via consistent design principles.

These changes required a massive implementation program. Between 1998 and 2008, for instance, the percentage of roads in Dutch cities that were zoned at 30 km/h increased from 15% to 70%, while the percentage of rural roads zoned at 80km/h dropped from 97% to 43%. Over the same time period, fatality rates dropped steadily — 5.3% per year — compared to only 1.8% in the preceding 10 years.

Safe System goes global

Around the same time the Dutch began implementing such widespread changes, Sweden adopted an even more ambitious vision. In 1997, the Swedish Parliament passed a road traffic safety bill that set a goal of zero deaths or serious injuries on Sweden’s roads. They called it Vision Zero.

But it also came with specific innovations that sought cost-effective solutions to infrastructural upgrades. Most famously, Sweden developed 2+1 roads, a three lane road where two lanes run in one direction and one in the other, alternating every few kilometers, separated by a cable barrier. This barriers reduces head-on collisions, but still permits faster drivers to overtake slower ones — without having to clear space for a full highway. Rural highways where 2+1 roads were installed saw a 79% reduction in fatalities between 1998 and 2007 compared to those that did not. Crucially, they were relatively inexpensive to install.

The Netherlands and Sweden provided proof that a systemic safety approach significantly reduces fatalities — and other countries took notice. In 2008, the OECD published a landmark report which built on findings from both countries to further codify what was now being called the Safe System approach. It is defined somewhat differently across the world, but its essential components include not accepting the loss of life, tailoring interventions to local contexts in a way that is tolerant of human error, sharing responsibility across the system rather than blaming individuals, and applying solutions across similar locations throughout a system.

Mothers join hands with their children to block the intersection of Hellerman and Walker Streets in a plea for a stop sign on Hellerman Street after a woman was killed in an auto crash at the scene. Published in the Philadelphia Evening Bulletin. Courtesy Temple University Libraries.

When adopted widely,¹ results are striking. Between 1990 and 2017, fatalities dropped in Australia by 47%, New Zealand by 48%, Sweden by 67%, the Netherlands by 55%, and in Spain by 80%.

It is common to assume that Safe System is not possible in the United States, typically due to some combination of high car ownership, car dependency, and large proportion of rural expanse. Yet there are many countries that contradict those excuses. Canada and Finland have similar or higher rates of car ownership as the US, yet have much lower fatality rates. In Canada specifically, people have a 60% lower likelihood of dying in a crash than in the United States, despite being equally spread out.

What does this actually look like?

Over the years, urban planners have refined Safe System road modifications in a way that is tailored to American urban environments. What does this look like in practice?

The U.S. model typically encourages wide lanes and corners to increase driver visibility, but this has the unintended consequence of encouraging cars to go through intersections faster, and and thereby decreasing the peripheral vision they might have retained at a slower speed. Instead, the Safe System intersection is designed to limit car speed and facilitate eye contact between users. It does so by expanding pedestrian areas via curb extensions or bumpouts, narrowing crosswalks, and removing parking within 20-25 feet of an intersection. The crosswalks and narrowing of the lanes encourages cars to slow and to stop well ahead of the crosswalk, while bumpouts shorten the distance pedestrians must be in the road.

A handful of U.S. cities have individually implemented some aspects of these safer design approaches. Chicago has installed protected bike lanes and pedestrian safety islands, which function similarly the the intersection describe above by making crosswalks shorter and increasing visibility. Baltimore and Washington, D.C.’s improved intersections use pylons and other forms of road furniture to force cars to make perpendicular (rather than angular) turns. This not only slows traffic, but permits turning cars fuller visibility of the crosswalk.

Diagram based on NACTO, the National Association of City Transportation Officials, a coalition of the Departments of Transportation in North American cities.

Whereas the Netherlands clearly differentiates roads and streets — as do Germany, Spain, and France — the US is known for having “stroads,” roads where cars reach high speeds yet must also avoid drivers entering from adjacent businesses and homes. The majority of fatal crashes in American cities happen on these “stroads,” and impact pedestrians and cyclists in particular.

Stroads, however, can be modified to serve all users, including roadside businesses, shoppers, drivers, and pedestrians. For example, La Jolla Boulevard, just a few blocks from the beach in San Diego, was previously a five lane road that served 23,000 vehicles per day. Renovations reduced it to two lanes, with roundabouts, parking, bike lanes, and varied sidewalks. Traffic dropped only slightly — to 22,000 vehicles — but the boulevard shifted from being unsafe for pedestrians to a walkable, inviting street. Average vehicle speeds dropped from 40-45 mph to 19 mph, noise levels decreased by 77%, retail sales increased by 30%, and traffic crashes plummeted by 90%.

There are other success stories. As of 2024, Hoboken, New Jersey has not had a single traffic fatality in seven years, mostly thanks to inexpensive intersection design upgrades. And larger cities such as New York, Chicago, and Washington D.C. also experienced fatality reductions through the aughts and early 2010s, via various improvements. In DC, in particular, traffic fatalities fell from 9 per 100,000 residents to 3 between 2000 and 2012.

But nationwide, some of these gains have reversed. During the pandemic, road fatalities increased, particularly during late nights and early mornings. And while congestion has come back to pre-pandemic levels, road fatalities have stayed high.

Experts seem to agree that the pandemic changed driving behavior: the increase in vehicle occupant deaths appears to be heavily weighted towards people who weren’t wearing seatbelts. Speeding, distracted driving, driving without a license, and driving under the influence of alcohol or drugs rates have also increased.

But it’s not just COVID: U.S. pedestrian fatalities have increased 80% since 2009. Many factors contribute to a deadlier environment: the increase in average car size and weight, smartphone use, and large population growth in metro areas with the highest per-capita pedestrian death rates in states like Florida, Texas, and Arizona. Another factor appears to be the suburbanization of poverty. More lower-income Americans who rely on public transit now live in areas built around the most dangerous roads.²

Vision vs. action

The United States still lacks a cohesive, national effort to reduce road fatalities. More than 60 American cities have signed onto Vision Zero commitments, but implementation of comprehensive reforms is rare. This wouldn’t necessarily mean a stall in progress — the Dutch method was also incremental — but factors within the American road system make tackling even the highest leverage changes challenging.

Cities often lack control over the roads that run through them, especially the most dangerous thoroughfares. Municipal governments only manage smaller streets, as major arterial roads and highways are generally owned or managed by state DOTs. Most state DOTs only require compliance with the Federal Highway Administration’s Manual for Uniform Traffic Control Devices, which sets baseline rules for street design and traffic control.

Prior to 2023, the MUTCD set speed limits based not on safety but on the speed that 85% of drivers instinctively follow. It wasn't until 2023 that MUTCD stopped recommending setting speed limits solely at the 85th percentile rule. It now allows for the consideration of other factors like pedestrian activity and crash frequency, but these are not mandated, and it will take years for post-2023 changes to take effect.

Ideally, pressure from Washington would force changes at the state level. The U.S. DOT claims to adopt Safe System “as the guiding paradigm to address roadway safety.” But there is no coordinated effort to align Safe System principles with federal budgets and investments — or the MUTCD. Transportation officials have adopted some approaches similar to the Safe System approach, like the "Forgiving Highway" concept designed to mitigate crash consequences, but the U.S. DOT, despite the lip service, stops far short of embracing the holistic safety enhancements championed by Safe System. The result is a piecemeal strategy that primarily focuses on mitigating crash impacts rather than preventing incidents.

Even when cities do wish to implement changes, state DOTs can delay, block, or even overrule local efforts to make roads safer. For example, in 2022 after San Antonio voters approved funding for a "complete street" redesign of two miles of a six-lane street, the Texas DOT unexpectedly blocked the project, citing its authority over the road — despite years of collaboration with the city. This reversal was part of Texas DOT’s broader policy to maintain all existing vehicle lanes, aligning with the Texas Republican Party’s stance against road diets.³

Nor is this purely a red state problem. Nine states, for instance — including purple Maine, New Hampshire, and Wisconsin — prohibit speed cameras. Few states have implemented them at all. Even in places where officials want change, the norms of highway-focused agency culture and focus on free-flowing car traffic make progress difficult. One advocate reports how some state DOT traffic engineers express discomfort with engineering roads that guide driver decisions. State inertia demonstrates the need for stronger federal support to align transportation policies with safety goals, and possibly grassroots activism to demand that goals are met. Vision Zero can’t be achieved in an environment so myopically focused on optimizing business as usual.

America still treats road safety as a separate special program, a “nice to have,” not a basic requirement of all projects. While spending $50 billion yearly on roads, the government dedicated just $2.4 billion specifically to reducing traffic deaths in 2022-2023. As former DOT official Beth Osborne puts it, "I would argue so long as you can easily pick out the safety projects, that's a problem. They should all be safety projects."

Crucially — and unlike in other countries — this is accepted as business as usual. Public (and congressional) attention on this front remains relatively muted outside of advocacy and grassroots efforts that don’t appear to have the same purchase as those in the Netherlands throughout the 1970s. The instances where road fatalities do reach media attention are more often when car companies are found at fault for flaws, such as the intense Congressional scrutiny over 19 deaths linked to sudden unintended acceleration in Toyotas. This tendency is partly driven by the U.S. legal framework, which allows for substantial settlements in cases of vehicle design faults — and otherwise almost only for driver negligence.

There is no denying that the comprehensive implementation of Safe System in America will be logistically difficult and expensive. But the alternative — one which many European countries decided not to accept decades ago — is some 40,000 deaths every year, double that of gun homicides.

America's road safety crisis demands a three-pronged transformation that addresses the core issues revealed in successful international models. First, we need a national vision and strategy to prioritize preventing road fatalities. Support at the highest levels would ensure that federal policies attach funding mechanisms to safety requirements, and possibly outcomes. Such changes can be integrated during already-planned maintenance and upgrade cycles, and prioritize the highest-risk corridors that account for a disproportionate share of fatalities.

Second, comprehensive design reforms must replace outdated planning and design standards and protocols — particularly the MUTCD — that handcuff local and state governments and engineers trying to implement proven safety measures.

Third, we need to build a culture of urgency across America. A shared conviction that road fatalities are solvable can unite safety advocates, business owners, and community leaders into powerful coalitions. These alliances can challenge the political dynamics that have allowed complacency — even in the face of a decade of rising deaths — to proliferate. Proven solutions exist. What’s missing is our collective will to demand them.

***

The views expressed in this piece are the author’s and do not represent those of the US government

1. No country claims fully comprehensive implementation, but Australia, New Zealand, Spain, Sweden, and The Netherlands have been using the approach for at least two decades.

2. Affordable multifamily housing if often built near stroads.

3. The Texas GOP’s legislative priorities for 2024-2025 include“Resisting unconstitutional federal acts and mandates that restrict transportation, including mandatory kill switches in vehicles, road diets, and restrictions on the owner’s right to repair vehicles and equipment.”