Is Zero Reasonable? A Multidisciplinary Framing of Traffic Deaths

"Reasonable to expect" is doing sneaky work. It smuggles in three questions and lets you argue about all of them as if they were one.

• Prediction. Will the annual toll ever print as 0? You're guessing at the future of a chaotic system with 230 million drivers.
• Feasibility. Could it, in principle, with the right roads, vehicles, speeds, and rules? Now you're asking what's physically and institutionally achievable.
• Commitment. Should we treat zero as the organizing goal even if we never literally touch it? That's a question about targets and about who carries the blame.

Watch what happens next. The realist insists we'll never predict zero. The advocate insists we must commit to zero. They talk past each other and call it a debate. But both can be right, because they're answering different questions with different kinds of claims. Sort the questions first. The rest of this essay is just me tracing which discipline speaks to which one.

Injury epidemiology reframed road death decades ago, and everything else rests on that. To a public-health researcher a crash isn't bad luck. It's a preventable injury event with identifiable causes and dose-response curves, no different in kind from cholera or lung cancer. William Haddon Jr., the first head of what became NHTSA, built the Haddon Matrix: a grid crossing three phases (pre-crash, crash, post-crash) against three factors (human, vehicle, environment). Overnight a single death had nine cells where you could intervene instead of one, and only one of those cells was "the driver screwed up."

That quietly killed the dominant theory of the 20th century, which held that crashes come from "the nut behind the wheel," so the fix is better nuts. The epidemiological frame says you cut the disease burden by changing the agent (crash energy) and the environment, because you will never finish reforming 230 million humans. Seatbelts, airbags, guardrails, breakaway sign posts. All of them came from people who stopped asking whom to blame and started asking how to keep kinetic energy from reaching a body.

In 1997 Sweden ran the public-health logic to its end and gave it a name: Vision Zero. The engineering rests on two stubborn facts. Humans will err, and a human body absorbs only so much force before it quits, an unprotected person struck at 30 km/h usually survives, at 50 km/h usually dies. That reframes the whole goal. You're not trying to manufacture error-free drivers. You're building a system where an ordinary mistake doesn't get converted into a corpse. So speeds are matched to the crashes a given road can actually produce, conflict points get engineered away, and the energy stays below what a body survives.

Think defense in depth, James Reason's "Swiss cheese" model borrowed from aviation and nuclear plants. Driver, vehicle, road, speed limit, trauma response, no single layer has to be flawless. The holes simply can't align. And the genuinely radical part is who gets held accountable: when someone dies, the Safe System asks what the designers could have done differently, not only what the victim did wrong. From there, near-zero stops reading as utopian. It reads as an engineering target several countries are visibly closing in on.

The numbers complicate the optimism in two directions at once. First, the denominator decides the story: the fatality rate per 100 million vehicle-miles has collapsed over a century even as the raw count plateaued, because Americans drive vastly more. "Huge progress" and "stuck for a decade" are both true, they're just different fractions. Second, death is a rare event smeared across hundreds of billions of trips a year. You can drive the per-trip probability arbitrarily low and still expect a non-trivial count, the way a tiny per-flight risk still yields crashes at scale. Pushing a Poisson process to a hard, permanent zero is a categorically stronger claim than pushing it very low.

Which is where the favorite analogy, "commercial aviation is basically at zero, so why not roads?", both inspires and misleads. US airline travel really has gone years with near-zero passenger deaths per billion miles. But aviation is a closed system: a few tens of thousands of certified professional crews, redundant aircraft, rigid procedure, total traffic control, and a no-fault investigation culture that feeds every incident back into design. The road network is an open system, ~230 million amateur operators, no recurrent certification, almost no redundancy, machines sharing space with toddlers and deer. Aviation proves near-zero is possible; it also prices the ticket, and the price is a degree of control roads will never have.

Welfare economics asks the uncomfortable question out loud: is zero the efficient target at all? In the textbook model, the marginal cost of preventing one more death eventually exceeds the marginal benefit, so the "optimal" number of crashes is greater than zero. Economists make this tractable with the value of a statistical life, the US Department of Transportation uses roughly $12-13 million, which lets a guardrail's cost be weighed against the deaths it prevents. By that arithmetic, spending without limit to save the last life is irrational.

Vision Zero's deepest move is to reject the premise of that whole calculation. It refuses to treat a human life as one priced input trading against travel time and construction budgets. Economics can still do enormous work inside the commitment, sequencing which countermeasures buy the most safety per dollar, but it cannot, by itself, settle whether a fatality is an acceptable line item. That's a values question wearing a spreadsheet. (Economics also flags a feedback trap: safety gains can be partly spent as extra speed or risk, the Peltzman effect, or risk compensation, which is the handoff to psychology.)

Strip away the engineering and a moral question is left standing: who decides what risk is tolerable, and for whom? A utilitarian can live with a non-zero optimum as long as total welfare comes out ahead. A deontologist can't. And Vision Zero is essentially deontological: to use some people's deaths as the price of others' convenience is to treat those people as means, not ends. That's Tingvall and Haworth's founding axiom. It can never be ethically acceptable that people are killed or seriously injured while moving through the road system.

Distributive justice sharpens the knife. Road deaths don't fall evenly. Pedestrians and cyclists, rural drivers, low-income communities, Black and Native American Americans all show up over-represented in the toll. "Acceptable risk" is always acceptable to someone, and usually not the person carrying it. It's the same reason the popular trolley-problem framing of self-driving cars is mostly a distraction. The real ethics of automation lives in the mundane, structural choices, how fast to deploy, who gets exposed to immature systems, how liability shifts, not in some lifeboat dilemma a car meets once a decade.

About 94% of crashes involve human error. NHTSA quotes it constantly, and it's true - it just doesn't say what people think it says. It counts the last link in the chain honestly, then hands a quiet alibi to everything upstream that turned a mistake into a fatality. Behavioral science covers the immediate failures fine: speed, impairment, distraction, fatigue, plus the way people partly spend their safety gains on faster, looser driving. Fine. But look at the structural numbers and they drown that out. We kill our own on the road at roughly two to four times the rate of Sweden, the UK, or Japan. American drivers aren't uniquely reckless. What's different is exposure and design - deep car dependency, more miles driven, fast multi-lane "stroads" knifing through neighborhoods where people walk, vehicles that keep getting heavier and taller, induced demand, and land-use that makes every errand a high-speed trip.

Here's the hope buried in all that bad news. If the gap with peer nations came down to national character, we'd be hopeless. It doesn't. It comes down to roads, speeds, vehicles, and rules - and those you can engineer down.

If human error is the last link that matters most, automation is the obvious lever to pull. Today's ADAS features already cut specific crash types by measurable amounts, automatic emergency braking, lane-keeping. Mandating them is one of the highest-yield moves on the table. And fully self-driving cars? They promise to strip out most of the human-error contribution altogether.

But here's the catch: technology mostly redefines "zero" rather than delivering it any time soon. Autonomous systems swap familiar failures for strange new ones. The long tail of bizarre edge cases. Brittle behavior when robots and humans share the road. Cybersecurity. Accountability that slides off drivers and onto manufacturers and their code. Fleet turnover takes decades too, so a car bought today might still be rolling in 2045. The honest projection: automation can plausibly push deaths toward zero over a few decades while quietly changing what we even "expect." Because the real question becomes how safe we require the machines to be before we trust them, and that loops right back to ethics and economics.

Put the three readings next to each other and the question mostly answers itself.

• As a forecast: no. A literal, permanent count-zero isn't a reasonable prediction on any near horizon. Think about what you're betting against: rare events spread across hundreds of billions of trips, an open system full of amateur operators, decades of fleet turnover. A hard zero is the wrong shape of claim.
• As an engineering target: mostly yes, if you say near-zero. Peer countries and specific Safe-System cities have already pushed deaths way down using tools we own right now - speed management, separating users, vehicle standards, automatic braking. The gap that's left is political and infrastructural. It isn't a law of physics.
• As a moral commitment: yes, and probably the only stance you can defend. Reject it and you're left writing down the number of preventable deaths you'd tolerate. No official says that number out loud. It can't be said.

Here's the reframing that actually helps. Quit asking "can we hit zero?" Start asking "is any death we could have designed out acceptable?" Those are two different questions, and the second is the one Vision Zero poses. That's when "reasonable to expect" comes apart into three answers at the same time. Reasonable to predict, no. Reasonable to pursue, yes. Reasonable to organize around - that's the useful one. The disciplines don't really disagree. They're answering different questions, and any serious conversation starts by naming which one you mean.