Over the past year, I have repeatedly encountered legal consultations concerning disputes over motor vehicle road traffic accidents: a villager carrying vegetables to the town market was seriously injured after being hit by a car; a city pedestrian walking to work was struck head-on by a delivery e-bike and suffered fractures in both legs; someone driving a long distance home from work dozed off and hit a three-wheeled electric vehicle, causing the other party’s death; another person riding an electric bike with a child to school was hit by an SUV and disabled; and the most recent case involved a student riding a shared bicycle who was seriously injured after being hit by an electric car. Among these, the last two upset me the most, because neither accident actually happened at high speed, yet in the former case the e-bike unfortunately encountered a new-energy SUV weighing more than two tons, while in the latter case visibility at the entrance of a narrow alley was poor, and the electric car barely even lost any paint. It was precisely these two accidents that led me to focus on the issue of pedestrian protection in cars.

Frankly speaking, the topic of car safety has been argued online into almost two separate worlds. Car sellers talk about high-strength steel, how many stars a crash test received, and how much the body can withstand; car buyers calculate whether the people inside will be okay after a crash and whether insurance premiums will rise; under online videos of car accidents, car fans of various camps compare which brand’s car is harder and which car can win in a collision against another; as for those standing outside the car—pedestrians and cyclists—they are rarely part of this discussion. They appear neither in advertisements nor on most people’s car-shopping lists, and even short-video platforms’ automotive video recommendation feeds often bypass them. Only when an accident makes the news are they briefly remembered, and then quickly buried by the next hot topic.

In recent years the situation has only gotten worse. There are more and more two-and-a-half-ton, even three-ton-class large new-energy vehicles, with hoods made ever higher and blunter; on the same street, more and more delivery riders are being driven by algorithms; urban space has forced many families onto electric bicycles, with often more than one child squeezed onto the rear seat. Vehicle mass is rising, speed differentials on the road are rising, but the pedestrian’s position has barely changed at all.

After a period of careful comparison and reflection, in my view, pedestrian protection in cars has never been a purely engineering-technical question. It simultaneously involves legal rules, commercial logic, and urban space; three forces have twisted the same problem into a dead knot.

Limited Compensation Liability and Unlimited Compensation Liability

A life has a price, but half a life instead becomes a bottomless pit; a fatal accident is often easier to “close” than an accident requiring lifelong care. This is the most absurd part of this compensation system.

Civil compensation for traffic accidents is an arithmetic problem that can be calculated, and also one that cannot be calculated clearly; the difference lies only in who is injured.

If someone is killed, under the current judicial interpretation on compensation for personal injury, death compensation is calculated as the per capita disposable income of urban residents in the place of the court accepting the case in the previous year multiplied by twenty years, decreasing by one year for each year over sixty, and counted as only five years for those over seventy-five. This sum, plus funeral expenses and dependent living expenses, can be estimated in advance, and in financial terms constitutes limited liability. After the comprehensive auto insurance reform in September 2020, the death and disability compensation limit of compulsory traffic accident liability insurance was raised to 180,000 yuan. Added to a third-party liability insurance policy with a million-yuan coverage limit, the accounts of a fatal accident are calculable, priceable, and transferable for car owners and insurance companies. This is the institutional root of the public feeling that “death has an upper limit.”

If the person is seriously injured or even becomes a vegetative patient, the accounts are completely different. Nursing fees, rehabilitation fees, nutrition fees, and follow-up treatment fees occur year after year. According to judicial interpretations, rehabilitation fees and follow-up treatment fees can be sued for separately after they actually occur, and the nursing period is calculated at a maximum of twenty years. But the nursing needs of a vegetative patient will not automatically end in the twentieth year. What happens after that window is used up? The judicial interpretation gives no direct answer, and in practice one can only return to the individual case and sue again. For the victim’s family, this is an expense with no end in sight; every few years they must return to court and again prove that the person is still alive and that money is still being spent. For the party at fault and the insurance company, this is likewise an exposure that is hard to cap.

Compensation Liability Pushes Hard Cars into Becoming the Default Answer

The original intent of allocating compensation liability is to make the strong yield to the weak, but it has become a catalyst for hard-car culture; the risk is not eliminated, only pushed onto the weaker party.

The handling rule in Article 76 of the Road Traffic Safety Law has long been simplified by car owners as “if you hit a pedestrian, it’s the motor vehicle’s fault.” Strictly speaking, this judgment is not accurate, but the direction is indeed roughly so. When an accident occurs between a motor vehicle and a pedestrian, if the pedestrian is not at fault, the motor vehicle side bears compensation liability; if evidence proves the pedestrian is at fault, the motor vehicle side’s liability is reduced according to the degree of fault; even if the motor vehicle side is completely without fault, it must still bear no more than 10 percent of compensation liability, and compulsory traffic accident liability insurance also has a corresponding payment obligation under no-fault circumstances. To put it plainly, unless the motor vehicle side can prove that the pedestrian intentionally collided with the vehicle, it is very difficult to escape completely.

In legal theory this rule is called no-fault liability, and behind it is the logic of “the superior party bears the risk”: whoever controls greater kinetic energy and stronger risk-control capacity bears responsibility first. Its general direction of protecting the weak is not wrong, but its consequences have been underestimated. This constant placement of accident compensation liability on the motor vehicle side often makes car owners feel not “I must drive more carefully,” but “I may not be able to afford the compensation.”

The most critical problem is that if compensation is only paid to the other party, there is often auto insurance to fall back on; after all, third-party liability coverage for ordinary vehicles is already quite high, often 3–5 million yuan, and premiums are not expensive. Once the perspective of compensation turns inside the car, the situation is different. If members of the motor vehicle’s own side are injured in an accident, under the current auto insurance system the car owner may also face high claims, while insurance usually covers only within 10,000–100,000 yuan. Paid rides, in particular, themselves follow the no-fault liability rule. Even under the civil law rule of “good Samaritan rides,” if a passenger really gets into trouble, it is quite difficult for the driver to escape liability.

Under these circumstances, people naturally turn to another form of self-protection: buying a heavier, harder, taller car, hoping that if a collision really happens, their own side will all be safe. In “ Gas Is Too Expensive, I’m Thinking of Switching to a Pure EV, but I Don’t Know What to Choose ”, I recorded my own agonizing over switching to an EV. Looking back, all the comparisons were indicators such as battery, range, and in-car safety equipment; not one was about how friendly the car is to people outside it. This is not just my preference, but the information structure the entire market gives consumers.

Automakers have keenly picked up on this “risk-avoidance” psychology. High-strength steel throughout the body, body rigidity, and crash-test star ratings have become almost fixed talking points at new-car launches, while “what happens to the other person when this car hits a pedestrian” is rarely brought up. Occupant safety in cars has the dual endorsement of mandatory standards and authoritative ratings, but pedestrian protection design in China has long had only recommended national standards and voluntary ratings. The difference in pace between the construction of these two systems has given ample room for the marketing rhetoric of “as long as the people inside are safe, that’s enough.” After all,

Overseas Market Entry Thresholds

The target of accident compensation determines the incentive for design improvement. When automakers are never directly sued in court because a pedestrian was injured, design will naturally not put pedestrians first.

If we broaden the perspective globally, we find that pedestrian protection overseas has long ceased to be a voluntary matter for companies.

Europe and Japan have written collision protection for pedestrians’ heads and legs into the mandatory entry thresholds before a new car can go on the market. The source of EU regulations is the United Nations-level Global Technical Regulation GTR No.9, which is then applied to every vehicle through type approval, and these requirements have become increasingly stringent over the years. Even more noteworthy is that the EU’s new round of vehicle safety regulations lists automatic emergency braking as mandatory equipment for all newly registered passenger cars and light commercial vehicles, and the ability to recognize pedestrians is precisely the core direction in which such systems have been continuously tightened in recent years. In Europe, a car that turns a blind eye to pedestrians cannot even obtain market approval.

The United States has no unified federal mandatory threshold, but it has another set of constraints. A 2023 study released by the Insurance Institute for Highway Safety (IIHS) showed that pickup trucks and SUVs with higher and more vertical front ends have a significantly higher risk of killing pedestrians than low-fronted sedans, and the rise in U.S. pedestrian deaths in recent years has been almost synchronized with the rise in the share of large vehicles. Combined with the tradition of punitive damages and product liability litigation, the price automakers pay for design defects often runs into hundreds of millions of dollars. The fines and settlement amounts in cases such as the General Motors ignition switch and Toyota unintended acceleration remain industry teaching materials to this day.

By contrast, in China the national standard for pedestrian collision, issued in 2009, is still recommended. In recent years the industry has continually called for upgrading it to a mandatory standard, but what truly influences automakers’ decisions every day is still voluntary ratings such as C-NCAP. The 2024 version of C-NCAP has already attempted to merge headform and legform impacts with pedestrian automatic emergency braking scenarios into an integrated evaluation. The direction is right, but its nature is still a rating, not an entry threshold. The more critical difference lies on the legal side. The Civil Code provides for punitive damages in cases where a party knowingly produces and sells a defective product and causes another person’s death or serious harm to health, but in traffic accidents victims almost always sue only the driver and the insurance company; cases that put vehicle model design defects before a court are extremely rare.

High-Risk Non-Motor Vehicle Participants

The essence of misaligned responsibility allocation is the mismatch between control and responsibility. Whoever designs the rules, whoever determines speed, and whoever profits from accidents should bear corresponding responsibility, but this simple principle has been bypassed in the platform economy.

Everyone knows that driving on the road is backed by a whole system of qualifications and supervision: driver’s licenses, exams, insurance, point deductions, plus ubiquitous electronic surveillance, all striving to train every motor vehicle driver into a “professional” in traffic regulations. But although pedestrians, bicycles, and electric bicycles also participate in public traffic, they have almost no threshold at all. The national standard for electric bicycles long ago drew lines for speed and weight, but they require no driver’s license, no insurance, and no one has given their riders systematic traffic regulation training. The highest-risk road sections are precisely where these threshold-free participants share the same road with steel weighing one or two tons or more.

The most troublesome link in this field is food-delivery riders. Platforms use algorithms to compress delivery times to the limit; for the delivery system, running red lights and going the wrong way are often the shortest paths, and the cost of fines for violations has long been factored into the per-order price and borne by the riders themselves. Legally, most crowdsourced riders are characterized as having a cooperative relationship with the platform rather than a labor relationship, so the platform neither bears the social insurance costs of traditional employment nor basically enters the accident liability chain. After a food-delivery accident, the parties the victim can sue are mainly the rider as an individual and the per-order accident insurance, occasionally the car owner who hit the rider, while the platform, as the rule-maker and beneficiary, often remains invisible. The occupational injury protection piloted in some provinces and cities starting in 2022 was rolled out nationwide from July 2026; when the expansion was announced, cumulative participation had already exceeded 12 million people. This is an important make-up lesson at the social security level, but it protects the rider’s own occupational injuries and cannot replace tort compensation to third parties in accidents.

Families Forced onto Electric Bikes by Urban Space

The road rights that cities owe in planning ultimately become liability in accidents, repaid with their bodies by those with the least bargaining power.

There is another overlooked group of traffic participants: ordinary families pushed onto high-risk routes by urban structure. The phenomenon of “separation of jobs and housing” that has appeared in large numbers during urbanization has stretched ordinary people’s commuting distances longer and longer, while high-quality educational resources are highly concentrated; housing near schools is unaffordable, and walking and bus connections are often broken, so electric bicycles have become the only realistic choice for many families. Although the traffic management rules of various cities generally allow an electric bike to carry only one minor on a fixed seat, more and more two-child and even three-child families need to send multiple children to school at the same time. What are they supposed to do? It is impossible to drop one off and then go home to take another, so the combination of one parent, two schoolbags, and one electric bike has become a fixed street scene in many cities in the morning and evening, and also a routine violation under traffic rules.

When I previously wrote “ Why Is Almost No One in China Discussing Road ‘Speed Reduction’? ”, I mentioned that the awkward position of electric bicycles is only one side of Chinese cities’ inability to “slow down.” The absence of a slow-traffic system, mixed motor and non-motor traffic, and intersections with no room for error are not planning details but risk allocation. When the fault tolerance of the physical environment approaches zero, and the weak have no choice but to venture onto the road every day, all the law can do is assign fault after the fact, recording systemic failure stroke by stroke as individual fault.

Put Costs Back Where They Belong

After analyzing these dilemmas, one finds that they actually share the same underlying problem: cost mismatch. Design costs, labor costs, and road-right costs have not been paid to those who should pay them, so the risk is left to those who can least afford it. The direction of the solution is nothing other than putting costs back in their original place, acting simultaneously from four aspects.

The first is technical thresholds. It is still necessary to upgrade car pedestrian protection design from voluntary manufacturer ratings to a mandatory baseline, and to update it regularly like the EU. The EU can do this because GTR No.9 was turned into mandatory type approval; if an automaker does not meet the standard, its cars cannot go on the market. What China lacks is precisely turning the baseline from recommended into mandatory, making pedestrian-recognizing automatic emergency braking standard on new cars, and setting tiered requirements for high-front-end and large-body models. Every upgrade of a technical standard will become a real budget item for hood energy absorption, front compartment layout, and front-end styling, rather than wording in advertisements.

The second is the compensation structure. Personal injury compensation needs to move from one-time settlement toward sustainable underpinning. First, link the compulsory traffic accident liability insurance limit to per capita income and adjust it regularly, rather than letting a 2020 figure carry on for more than a decade; second, develop long-term care insurance and annuity-based payouts, so that the nursing costs of vegetative patients’ families no longer depend on fighting a lawsuit every few years; third, extend the use of the Social Assistance Fund for Road Traffic Accidents from emergency advance payments to rehabilitation and long-term care. As long as the compensation gap between death and serious injury remains, hard-car culture will not disappear, because killing someone and disabling someone have always been different financial risks for car owners.

The third is the return of responsibility to platforms and automakers. After occupational injury protection is rolled out nationwide, the next step should be to use the degree of algorithmic control as the yardstick for determining platform liability. The tighter the algorithmic scheduling and the deeper the control over riders’ behavior, the more the platform should enter the accident liability chain. At the same time, the door to product liability litigation should be opened for pedestrians. The provisions in the Civil Code on punitive damages for defective products cannot just lie on paper. The target of compensation determines the incentive for improvement; only when rule-makers feel pain will the rules change.

The fourth is optimized design of urban space. The slow-traffic system must be built as a public good, not decorated as landscape. Separation of motor and non-motor traffic, speed reduction at intersections, and traffic design around schools should calculate the pedestrian account first and the car account second. Cities must also directly answer a specific question: for a family with two children, besides illegally squeezing onto one electric bicycle, are there any compliant travel options? If urban planning has no answer to this kind of question, enforcement will fine a systemic dilemma as individual fault.

Writing here, I cannot help but think of the next change happening in the automotive industry. After intelligent assisted driving becomes widespread, the focus of pedestrian protection may shift from “how to cushion after a collision” to “how to judge before a collision.” When the system brakes and whom it prioritizes protecting is equivalent to writing moral choices into code in advance. But this also requires further legal regulation, for example, the responsible subject may shift partly from the individual driver to automakers and algorithm designers—though such a change is indeed very difficult.

But I believe that no matter what cars become, pedestrian protection is ultimately still a question of responsibility allocation: whether technology, responsibility, and space are willing to lean a little toward those with the least ability to fight back. This question is in fact not only something the automotive industry must face; it is a required question that concerns everyone.