Keith Bradsher’s book High and Mighty: The Dangerous Rise of the SUV is compelling (if depressing) reading. Bradsher describes how an obscure loophole in fuel economy standards led to the rise of the SUV, and how automakers and unions squelched efforts to hold them to the same standards as cars. Regulators estimate SUVs cause 3000 extra deaths a year–they contribute little or nothing to the safety of their occupants, but greatly increase the danger of driving for non-SUV drivers. Greg Easterbrook describes some other problems:
Traffic studies show that the typical SUV occupies as much road and parking space as 1.4 regular cars. These vehicles have converted driving . . . into a nerve-wracking Darwinian battle. . . . [They] are designed to bring out the worst in their owners while simultaneously making them feel that they are invincible. And they simply take up space, shrinking the road and parking acreage and increasing all forms of congestion.
This is an interesting attribution of negative externalities to SUVs, fleshed out a bit more by Bradsher. He says that they will increasingly become dominant because of “network externalities,” described as follows (on ixi and 170 of the book):
It is becoming harder and harder to see down the road while sitting in a car, because of the impossibility of seeing through the tall SUVs . . . . Backing a car out of a parking place between two taller vehicles has become an exercise in hope that no one is about to come barreling by. . . . The extra weight of SUVs rearranges where deaths occur in crashes, transferring deaths from the SUVs to the cars. . . The result has been a highway arms race.
I think Bradsher meant to say “positional externalities;” those occur when people invest in resources in order to best others in some zero-sum competition (such as a car crash, where the crash momentum gets distributed between the cars, and the heavier car usually suffers less). In any event, the sad story of the SUV suggests that authorities like NHTSA need to pay attention to all the externalities of technology, not just market demand for it.