Large, heavy vehicles are safe, as everyone knows. If you’re going to be in an accident, would you rather be in a Miata or an Escalade? More large vehicles on the road make us safer, and we should worry about anything which reduces vehicle sizes. Notably, fuel economy standards decrease vehicle size, so we will become less safe as fuel economy standards become more strict. See for example Crandall and Graham (1989).
Or so conventional wisdom goes, but it turns out the truth is more subtle, according to recent research.
The issue is not how big cars are on average, but rather the dispersion of vehicle types and weights. A large body of evidence shows that people in small cars are much less likely to be killed in a collision with another small car than with a larger vehicle, particularly a truck or an SUV. Heavier vehicles are safer for their occupants, given a crash occurs because heavier vehicles are favored by the laws of physics in a crash. Trucks and SUVs gain an additional advantage because they are also tall. When a truck or SUV hits a smaller car the force of the crash impacts on the upper bodies of the occupants of the car rather than on the steel frame of the car below. Further, pedestrians and cyclists are more likely to be killed if they are struck by a truck or SUV than by a car. Finally, trucks and SUVs tend to have relatively poor braking and maneuverability and, given driving styles, may cause an increase in accidents.
Heavy vehicles, particularly trucks and SUVs, are not safer than smaller cars, they’re only safer for their occupants conditional on a crash occurring. For everyone else trucks and SUVs are a hazard. An increase in the proportion of trucks and SUVs on the road could increase or decrease overall safety, depending on the mix of vehicles already on the road. Note that an increase in the proportion of trucks and SUVs which has no effect on overall safety involves a perverse outcome: the occupants of the new trucks and SUVs experience more safety as a result of their vehicle selection, but pedestrians, cyclists, and occupants of smaller vehicles experience less safety. A Prisoner’s Dilemma writ large arises: we would all be better off if we all cooperate and drive small cars, but anyone can defect and buy an Escalade. An “arm’s race” occurs in which we all wind up buying vehicles which are inefficiently too large.
These considerations are troubling given the market share of trucks and SUVs has risen from 17% in 1981 to 50% in 2006, at least in part because fuel economy standards are lower for trucks and SUVs, effectively a subsidy. Several recent econometric studies, including Anderson (2007), Li (2009), and Jacobsen (2010), suggest that the increase in the proportion of trucks and SUVs has cost many lives (all statistics in this post were drawn from one of these papers).
Anderson (2007) estimates that the elasticity of the mortality rate to truck and SUV share is about 0.34 (or 143 deaths per percentage point increase in trucks and SUVs), that about 80% of the increase in deaths as more trucks and SUVs hit the road are to pedestrians, cyclists, and car occupants, and that a Pigouvian tax on trucks and SUVs would come in at just under $4,000. He concludes:
Overall, light trucks pose a significant hazard to other users of the highway system but on average provide no additional protection to their own occupants.
Li’s (2009) estimates suggest that about 12% of truck and SUV sales can be attributed to the arm’s race for private safety and that a tax on trucks and SUVs should be set at about $2,500. Jacobsen (2010) provides selection-corrected estimates of the effects of changes in fleet composition on safety and finds that a one mile per gallon increase in CAFE standards costs 164 lives due to the discrepancy in standards for trucks and cars, but a unified standard has little effect on overall safety.
These estimates are remarkably consistent given these authors use a variety of data and empirical methods. The conclusion is statistically robust and well-grounded in theory: large vehicles, particularly trucks and SUVs, pose substantial external costs and should face large corrective taxes, roughly $2,500 to $4,500. Further, these estimates likely underestimate how dangerous trucks and SUVs are as they assume away, due to data limitations, an effect of driving a truck or SUV on a given driver’s behavior: if people in such vehicles feel safer because they are in large, heavy vehicles, they may respond by driving more recklessly.
A question I have after perusing this literature: how about a uniform, possibly nonlinear, per-kilogram tax on all vehicles? Such a tax would not solve the vehicle configuration issue (an additional tax on trucks and SUVs would be required) but it would solve the arm’s race in vehicle weight.