In a recent Wired post, Aarian Marshall makes the point that there are several ways of accomplishing the goal of making roadways safer for the people in them. Many jurisdictions have adopted "vision zero" plans to reduce traffic injuries. That is, they aspire to reducing traffic fatalities to none through various safety measures.
Currently, one of the highest-profile measures connected to this effort is adoption of self-driving cars. Sebastian Thrun, one of the pioneers of this technology, has said that one of its primary aims is to save lives. Robot cars will not get bored, distracted, or drunk while driving, thus enabling them to overcome human foibles that increase risk. Of course, robot cars have to acquire the judgement and common sense that humans possess, a project that is still ongoing.
As Marshall points out, there are well-established and low-tech ways of improving roadway safety. Measures such as bump-outs stimulate driver attention and increase vigilance. Narrower lanes make drivers more cautious and decrease the span of roads for pedestrians when they cross. And so on.
Yet, slowing cars down can be difficult politically. Commuters are a potent voting bloc and tend to oppose any measures that add to their time in traffic. (For Ontarians, think of former Toronto Mayor Rob Ford's talk of the "war on the car.") However, they might be convinced to buy self-driving cars, which promise to remove the tedium of communting, if not its duration.
At this point, I am reminded of Evgeny Morozov's point about "technology solutionism". In a nutshell, this concept refers to a policy of technology providers to promote technological fixes to social problems with the effect of prioritizing private consumption over collective action.
Traffic safety is a social problem, that is, a problem that is created by everyone who uses roadways, through their complex interactions. Since roadways are administered collectively through governments, the obvious way of addressing problems with them would be through government action, such as changes in regulation concerning road design. Such changs would be systemic, relatively cheap, and effective.
Another approach is to sell individuals products that will improve things for them, which seems to be a selling point for self-driving cars. The argument is that when enough of these vehicles are on the road, their improved safey features will lower risk not only for their occupants but for people in the roadway in general.
It is possible that this scheme will work as advertised. However, it may well prove to be an expensive and inequitable route to improved road safety. It may be possible to reduce road risk more quickly and cheaply, and for people of lower income, through measures such as roadway design. Yet, people who might be better served by the latter approach do not form as potent a voting bloc, so the political will to follow it is harder to summon.
So, Marshall's argument raises an issue to ponder: Is the self-driving car an example of technology solutionism? If so, should we consider approaching road safety differently?