Big Carrots and Big Sticks

How to encourage car manufacturers to make safer vehicles

The inside of my family’s car is comfortable, quiet, and safe. I paid tens of thousands of dollars for a new vehicle last year, and I feel good about my purchase. I drive my kids to school and to practice and have a lot of confidence that they are safe inside our car.

But I feel a greater sense of caution when I cross a busy street to get to the store, when I ride my bike alongside much bigger vehicles, or even when I’m driving and see a huge truck coming my way. I feel this sense of caution because I know that little to none of the $10,000 or $40,000 or $100,000 cost of the vehicle in front of me was spent to make life safer for people outside of the vehicle. With few exceptions, “vehicle safety” means minimizing cabin deaths — i.e., safety inside the vehicle — while paying little attention to contour deaths (deaths of those outside the vehicle).

That dynamic of safety-for-me-but-not-for-thee shows up in the data. Pedestrian fatalities have nearly doubled in the past decade. Bike fatalities have increased by around 50%. And Americans are buying vehicles (especially trucks) that are safer for drivers inside the vehicle but put other drivers at risk.

Some of my liberal friends would argue that this is because car companies are full of greedy people who skimp on safety so that they can make a few extra bucks. Only strong government regulation can keep most of us safe. This argument is often correct (but probably not in this case).

Some of my conservative friends would tell me that government regulation is misguided and makes everything slower and more expensive. We as individuals must take responsibility for our own safety. This argument is often correct (but probably not in this case).

Unfortunately, neither the standard liberal approach nor the standard conservative approach is likely to solve this problem. Regulators have to balance and understand too many variables in a complex system (note the increase in traffic deaths the last ten years), and there’s little incentive for anyone to solve the problem if we take a hands-off approach.

This problem — safety-for-me-but-not-for-thee — is a classic externality: when I choose to drive a big truck, it has the side effect of making other road users — cyclists, pedestrians, other drivers — less safe. Other externalities include a company whose factory pollutes the air, or a nightclub whose loud music late at night bothers people who live nearby.

Externalities can be addressed with regulation or by changing incentives. In the case of safety-for-me-but-not-for-thee, the problem can be solved via an approach called Big Carrots and Big Sticks.

Big Data → Big Carrots and Big Sticks

Governments have very good data on car crash fatalities and can use it to make new cars safer for the people outside of them.

According to Car & Driver, the Ford F Series was the best-selling car in the US in 2023 with 750,789 units sold. They have a huge footprint, and I posit that Ford should be either rewarded or penalized for the real-world safety of their vehicles via the mechanism of Big Carrots and Big Sticks.

Big Carrots and Big Sticks are centered around rewards and penalties: if Fords are safer than baseline expectations, they will be rewarded with large stipends from the government; if Fords are less safe than expectations, they will be penalized.

Big Carrots and Big Sticks rewards and benefits are designed to incentivize safe designs with cash payments for every life saved, and penalties for every life lost. Because there is already strong incentive to make the inside of the car safe, I propose a relatively small incentive level of $1 million for cabin deaths (people inside the vehicle), and a higher incentive level of $2 million for contour deaths (people outside the vehicle, including in other vehicles).

I’ll illustrate the specifics of the Big Carrots and Big Sticks program by example.

Safety Baseline

Every mile driven on the road involves some risk of both cabin deaths and contour deaths.

For every 100,000 cars on the road, there are about 6 cabin deaths (occupant drivers and passengers) each year. Those same cars are involved with about 8 contour deaths (people outside the car, including occupants of other vehicles, pedestrians, and bike riders).¹

Those are the approximate numbers; we’ll use them to calculate a Safety Baseline for Ford F Series vehicles.

There were 750,789 Ford F Series vehicles sold in 2023. That means that in 2024, our Safety Baseline would be:

• 45 cabin deaths (occupants of 2023 Ford F Series vehicles)

• 60 contour deaths (other people would be killed in accidents involving 2023 Ford F Series drivers)

Depending on the actual safety performance of those cars, Ford could stand to be rewarded or penalized: rewarded if their on-road safety numbers are better than the Safety Baseline, penalized if their on-road safety numbers are worse.

Big Carrots and Big Sticks for Okay, Good, and Bad

If the 750,000 Ford F Series vehicles had exactly the safety record expected — 45 cabin deaths, and 60 contour deaths — they would be neither rewarded nor penalized.

If, however, the Ford trucks deviate from the Safety Baseline in 2024, they could incur a large reward or a large penalty.

Let’s first consider a “good” safety record for Ford. There are only 25 cabin deaths and 30 contour deaths. Ford would receive a $20 million reward for their positive safety record with occupants, and a $60 million reward for their positive safety record with non-occupants — an $80 million reward, just for their 2023 Ford F Series. If other models and other years had a similarly good safety record, Ford could add $1 billion or more to their margins. Ford’s net income for 2023 was around $5 billion, and a 20% increase would make shareholders very happy while also saving many lives.

We can also consider a “bad” outcome. For the Ford F Series, a bad outcome might entail slightly better than average safety for people inside the vehicle — only 40 cabin deaths — but double the number of contour deaths (120) as the Baseline. This outcome would result in a penalty of $115 million for Ford (a $5 million reward for better than Baseline safety for occupants, plus a $120 million penalty for much worse than Baseline safety for non-occupants). This would be a huge financial hit for Ford, especially if we saw similar trends with other models and years. Under my proposed system, Ford could easily take a hit of a billion dollars in lost profits as a result of a poor safety record.

To summarize, here are numbers for one model-year:

• “Okay” scenario: no change from today

• “Good” scenario: $80 million reward for good safety record

• “Bad” scenario: $115 million penalty for Ford for a bad safety record

The total penalties or rewards would be much higher across all models and years.

Image made by ChatGPT

Who Would Implement Big Carrots and Big Sticks?

Calculating the values for Big Carrots and Big Sticks is straightforward. States have data on car registrations, so it would be easy to define a Safety Baseline². The NHTSA puts out yearly data with details about each individual crash, making it straightforward to calculate both cabin deaths and contour deaths.

A more interesting question is who would implement Big Carrots and Big Sticks. Three possibilities are private foundations, state governments, and the federal government.

Private foundations could roll out Big Carrots and Big Sticks, awarding specific companies that have exceptionally good safety records. However, they would almost certainly have to rely on vehicle makers volunteering to be part of the program. If they did that, it’s likely that only makers of safe vehicles would sign up. Those makers would have incentive to make their cars even safer, saving lives, but makers of dangerous vehicles would opt out, so the impact would be more limited. Foundations are probably not the best option.

States could roll this out inside their states, using vehicle sales statistics and crash statistics within their states. This risks being a scattered approach, but with a critical mass of states, could meaningfully shift vehicle maker behavior.

The ideal option would be the federal government. If the U.S. government rolled out Big Carrots and Big Sticks, it would have a massive impact on vehicle makers’ behavior. This may seem like a stretch in a world where Congress struggles to keep government funded, but as a non-partisan, pro-consumer measure that also provides incentives and predictability for businesses, I believe it’s worth a shot.

Small Carrots and Small Sticks → Big Carrots and Big Sticks

This Big Carrots and Big Sticks program could roll out immediately, but a quick rollout would cause significant shock to the system and would likely be quite unpalatable to car manufacturers who stand to be heavily penalized for past actions.

A better approach would be start with Small Carrots and Small Sticks. For instance, rewards and penalties could start with only the current year’s models, with rewards/penalties of just 10% of the values proposed above ($100,000 for cabin deaths and $200,000 for contour deaths). Each year, rewards and penalties could increase — from $100,000 and $200,000 to $200,000 and $400,000, then to $300,000 and $600,000 — in ten years getting to the Big Carrots and Big Sticks end state of $1 million per cabin death and $2 million per contour death.

Critiques of Big Carrots and Big Sticks

I can foresee a number of critiques of Big Carrots and Big Sticks.

1. They’re a giveaway to car companies. There is a real chance that car companies could collectively make a lot of money under a system like this. If car companies reduced deaths to zero, they could stand to make approximately $100 billion more in additional profit using the numbers suggested in this post. While that might sound bad to populist ears, it would be a small price to pay for 40,000+ lives saved, and would be a small fraction of the $340 billion NHTSA estimated car crashes cost America in 2019 (a number which would be far higher today). This sort of program is probably only politically feasible if vehicle makers see it as neutral to positive for their businesses — i.e., with as many carrots as sticks — and Big Carrots and Big Sticks is a program designed with that balance in mind.

2. Manufacturers will get blamed for crashes that aren’t their fault. Imagine, for instance, that the driver of a Toyota is driving drunk and crashes into someone driving a Kia who was doing nothing wrong. If either driver is killed, Kia is penalized for the bad behavior of the Toyota driver. Why should this happen? The answer lies in the aggregation: there are surely thousands of cases like this — some one-car crashes, some crashes where the Kia driver is drunk, some crashes where the Toyota driver is drunk. If Toyotas are made in a safer way, fewer people will die in those crashes; if Kias are made in a safer way, fewer people will die in those crashes. In aggregate, the companies will be rewarded or punished in a way that reflects their true safety outcomes.

3. Companies may not change the way they make cars. It’s possible — but unlikely — that car companies would refuse to change their designs to make vehicles safer. While this is a conceivable outcome, it is at worst comparable to what would happen today. Surely Big Carrots and Big Sticks would not push vehicle makers to make vehicles less safe; the worst case scenario here is that cars don’t get any safer, but government generates more revenue that it can use to decrease deficits or provide services.

4. This is a service insurance companies should provide. There is some truth to this argument: insurance companies should charge more for larger vehicles that are more dangerous to people outside the vehicle. And they charge a little more: about $1710 for a 2023 Honda Civic, compared to $2024 for a 2023 Ford F-150. However, this small differential is far out of proportion to the much larger safety gap, wherein the driver of the old car is ten times as likely to be killed, because (under current laws) the insurance company’s payouts wind up being similar for the two different types of vehicles.

5. Big Carrots and Big Sticks are expensive. This is potentially true: a $100 billion pricetag would be a lot of money to spend. I’d counter this argument three ways. First, the Department of Transportation statistically values a life at over $12 million so this proposal to reward each life saved at around $3 million would be a very cost effective way to save lives. Second, it’s unlikely that safety numbers will change extremely quickly: $100 billion is very unlikely to occur overnight. Third, the reward/penalty thresholds would change over time, setting higher expectations for safety in future years. For instance, numbers could be set to equalize rewards and penalties at 36,000 traffic fatalities in 2025, 35,000 in 2026, 34,000 in 2027, and decreasing in each year beyond.

   

   A (kind of) Ford F-150 and a (kind of) Honda Civic by ChatGPT

The Backdoor Benefit

Perhaps my favorite part of this plan is that it would immediately push automobile makers toward advocating for safer streets.

If vehicle makers know that every traffic fatality cost them at least a million dollars, they will suddenly have huge incentive to make streets safer³. Today, car access and convenience are most important to the Fords and Teslas of the world; in a world with Big Carrots and Big Sticks, better safety could mean billions of dollars in additional profits.

This would align incentives beautifully, perhaps nudging Elon Musk and Mary Barra to go to Washington, Sacramento, and Austin to advocate for policies that would improve the safety of American streets. And every American would benefit.

1. These numbers are super rough, chosen to provide round numbers and for illustrative purposes. They also effectively double count deaths in multi-car collisions: when a Civic and an F-150 collide and the Civic driver is killed, this calculation counts both the contour death associated with the F-150 and the cabin death associated with the Civic.

2. I’ve outlined a simple approach under which cars are all treated the same. There might be arguments to adjust for miles driven or driver demographics. e.g., young men get into more crashes, and car companies shouldn’t necessarily be penalized for that.

3. Along somewhat similar lines, car insurance companies have taken steps to discourage texting while driving.