Driverless cars are quickly moving from the pages of science fiction novels to the real world. Different organizations around the globe are working hard to develop a driverless car that meets the qualifications to become available to the public. The Defense Advanced Research Projects Agency known as DARPA has been sponsoring an annual competition for the best autonomous vehicle since 2004 (Wikipedia). The DARPA Grand Challenge is an opportunity for the all of the developers of autonomous vehicles to compete and see how their prototype compares to others. In recent years the tech giant Google has dominated the competition. While owning a driverless car delivers considerable utility to the owner, the macroeconomic implications are much more complex.
If driverless cars became available for a reasonable price they would immediately cause a huge amount of structural unemployment. Taxi drivers, truck drivers, and basically anyone else who drives a car for a living could potentially lose their job. While this economics implication is more obvious it goes much deeper.
Considering that driverless cars would have to be basically 100% crash proof to be commercially viable it is safe to infer that their would be no more car crashes. All of the jobs, and car parts responsible for making cars safer in wrecks, or repairing wrecked cars would disappear(Forbes). Car sales would initially hit an all time high, but after a while car sales would drop below their current level since people would no longer total their cars(Forbes). This could be offset by the desire for consumers to have the latest technology in their cars. Personal cars could be customized to be a mobile workspace, or be fitted for other kinds of customizations.
The article linked below is the second part of a seven part series on driverless cars. The author goes into more depth of the negative macroeconomic impacts of a car. Later this week I will blog from a more optimistic perspective discussing in more detail the positive macroeconomic implications of the cars.
ForbesGoogle’s Trillion-Dollar Driverless Car, Part Two of a Seven-Part Series, January 24, 2014.
3 Comments
It is interesting that we will have driverless cars soon. I thought this was only possible in movies or dreams. However, it will take a long time for them to be affordable by many people. By that time, I hope, there can be other kinds of jobs available that the unemployment rate does not go up due to the development of driverless cars.
I still think that driverless cars is a nascent industry which will be further away from implementation than we think. The leaps which have been made are truly astounding yet marketizing these will most likely be very toiling. Before we have AI driving on the roads amongst us there may be other places more viable, for example trains. The world we live in is highly complex and the way most of these robots “learn” is by researching google in their brains to develop understanding and memory around them. As humans we know that if we see deer around the road we know to be more cautious however until the deer runs out in front of a car the “robot” driver is unable to be cognizant of the impending situation. The laws of physics still apply and a car cannot stop on a dime. If we are unable to implement electric cars on a larger scale how is it even possible to have “robot chauffeurs?”
Most of the hardware required is already on vehicles – lane departure warning, blind spot detection, adaptive cruise control, backup detection, and GPS. E(lectronic)-steer is standard on many smaller vehicles; ESC (electronic stability control) requires the ability to electronically activate brakes. Vehicle transponders are one missing piece; you really want to be able to detect vehicles obscured by other vehicles or buildings. GPS would ideally be high-grade systems, but that’s a matter of governmental approval. I don’t know whether road-edge sensors [cf. the RFID tags used in store anti-theft systmes] would be needed.
Software is the real barrier, hence Google.
But two cautions.
One is how big is employment of taxi drivers and truck drivers? You can go to the BLS Current Employment Survey to get the data. Taxis strike me as a particularly hard technical challenge – recognizing a “fare” waiving at you in snowfall – and while the long-haul interstate routes would seem easier to automate – the “Class 8” (semi) end – some trucks carry hazardous materials. (You may not realize that new trucks have adaptive cruise control and anti-sway mitigation.)
Second, the roll-out will take many years. The US has 250 million registered vehicles, with an average age for cars of 11 years (trucks are older). So even if every new vehicle has these technologies, it will take 10 years before the overwhelming majority of vehicles have autonomous capabilities. Furthermore, it will also take many years to roll over the fleet. Models for 2018 are already in development, and so what type of systems they contain will soon be decided. Since some cars aren’t redesigned but once every 6-8 years (especially pickup trucks), even if the new technologies become available this year, it would be after 2020 before all vehicle could have it. Then there are suppliers: with production of perhaps 100 million vehicles by 2020, you need a lot of ability to make the components. That capacity today is closer to 0 for some of the elements needed for autonomous vehicles. So cost issues aside, realistically you’re looking at 2030 before autonomous vehicles would be standard.
That mutes the macroeconomic impact on employment and (what you did not mention) productivity.
Comments are closed.