Over the last two months we’ve been treated to stories speculating that fully autonomous vehicles — read, “self-driving cars” — will be mass marketed by 2020. Others assert that it will happen more gradually: Certain new safety options essential in such a vehicle, such as automatic braking, will continue to be offered as they are perfected, transforming our automobiles over a generation into completely autonomous vehicles.
What we know is that the cameras, lasers, computers and the other parts that allow the self-driving vehicles on our highways cost somewhere around $75,000 today. That price will have to come way down before customers consider the option on the window sticker that reads, “autonomous control.”
Much more interesting than the price, however, is all the speculation the possibility of these cars is causing: Among other questions, what will happen to our economic society once these vehicles flood our highways?
Years of the Horse
Last week the Detroit News covered an investment paper in which Barclays’ Brian Johnson took the position that overall new car sales will fall by 40 percent over the next 25 years as autonomous vehicles become the norm. Why? Johnson is convinced that families will no longer own two or more vehicles: One self-driving car could handle a whole family’s demands. Barclays’ analyst is also convinced that, as families downsize their automotive fleets, 140 million vehicles will disappear from our highways.
And Johnson has what he calls historical proof of his prediction that self-driving cars will diminish the size of America’s vehicle fleet. That is this: The size of the horse population dropped sharply once automobiles came along. Sounds like it’s true. Forget that it’s a flawed analogy; it’s not completely accurate, either.
First, as the Civil War came to an end, it was estimated that there were 8 million horses in America. And their number continued to grow until the estimated peak year of 1915, when it is believed that 21.5 million horses existed here. By 1960, we were down to a mere 3 million horses — in spite of the overwhelming popularity of CBS TV’s Mr. Ed. However, today it is believed that there are 9.9 million horses in this country. So, if one compares our horse population today with that of 1867, or of 1960, despite the addition of 240 million vehicles on the roads there are also more horses. And in the past 55 years, as a percentage the horse population has grown almost in lockstep with the size of our automobile fleet.
Johnson also believes that self-driving cars will slam GM and Ford the worst, claiming in his “Disruptive Mobility Study” that GM will have to slash production by up to 68 percent, while Ford will lose 58 percent of its production and sales. His belief led to even more speculation by Fortune’s Kristen Korosec, or at least by whoever wrote the headline to her article on Johnson’s study: “Uber Wins, GM Loses When Driverless Cars Rule the Road.” The long and short of this subchapter of super-speculation is that Uber will no longer need drivers to conquer and vanquish the taxi and livery industry. Further, GM becomes a fond memory in a future where Corvettes and Camaros can drive themselves.
What? Why would anyone buy a self-driving Corvette or Camaro — where’s the fun in that? Precisely.
Then came the second theory on how self-driving cars will change America, in a Slate magazine piece bylined Joseph Coughlin and Luke Yoquinto. This theory has it that once self-driving cars come to market, the entire concept of living in the suburbs changes. Now, instead of working in Fort Worth and living in Aledo, one might consider a five-acre ranchette near Vernon.
Yes, according to these guys, living a couple hundred miles out in the country becomes a piece of cake if you own a self-driving car. They use the example of living in Bennington, Vermont, and working in Manhattan, and making that 180-mile commute in just 90 minutes. Wow!
Somebody Save Me
All this automotive illogic was just killing me. I don’t know who was worse, those at Slate who didn’t understand how self-driving cars work or those at Barclays who don’t understand how the average American family works.
First the Slate story. For someone to own a self-driving car and make it from a home in Vermont to Manhattan in 90 minutes, that person must travel at 120 mph. But one of the safety features hard wired into self-driving cars is that they will follow all local speed limits; and so far, to the best of my knowledge, New York’s Interstate 87 still doesn’t have a posted speed limit anywhere near 120 mph. Come to think of it, Hoosick doesn’t allow 120 mph motoring on NY State Highway 7 through their rustic village, either.
Second, even if it were possible under the absolute best of conditions to make that 180 miles in 90 minutes, heavy rain or snow would most certainly slow even the best self-driving car down to the speed those conditions dictated. Not to mention that self-driving cars don’t even work in snow as it’s impossible for their cameras to see the road or highway markers. But even here it doesn’t matter; even under the optimal conditions, few individuals are going to buy the gasoline needed to commute 360 miles to work and home each day, or want the depreciation on a car one is driving 100,000 miles a year.
But even if someone would pay for that much gasoline, or put that many miles a year on a car, few would spend three hours on the road going to work at 120 miles an hour in a car — even if they didn’t have to manually drive it.
That scenario also conflicts with Barclays and Johnson’s theory that families will need only one car when self-driving cars come to market. How would that work? The husband has the family car drive him 180 miles to work; does he then set it to drive back home — so the wife can use it to go shopping — only to drive back almost 200 miles to pick him up?
Even in Barclays’ scenario, what if the husband and wife both have jobs, but one works 20 miles east of their home and the other 15 miles west, and both have to be at work at 8:00 a.m.? Of course, there’s nothing to stop a family in which two adults work from sharing a car today, self-driving or not.
A Huge, Invisible Niche
Let’s look at the reality of self-driving cars using a Honda Accord LX sedan for our example. Today if that car were made to be self-driving and the equipment to do so were added at cost, an autonomous Accord would sell for $97,000. Now, let’s say that in a decade the cost of all of that gear to make a self-driving car work came down to $5,000; that would mean the option for an autonomous Accord would likely add $10,000 to the window sticker. So, here’s the question; would you buy a $33,000 self-driving Accord sedan or the regular one for $23,000?
Keep in mind that with the self-driving Accord you can’t go 20 miles over the speed limit on Interstate 30 anymore. At least, not in self-driving mode; but then again, if you habitually go fast, then paying extra for the self-driving option seems a bit silly.
As for the speculation on autonomous vehicles, the reality is that there will still be a huge market for them. Not because anyone wants a self-driving car, but because the elderly will have to either use a self-driving car or go live in a nursing home. There’s the most viable market niche, and it’s growing faster than anyone thinks.
© Ed Wallace 2015
Ed Wallace is a recipient of the Gerald R. Loeb Award for business journalism. He hosts Wheels, 8:00 to 1:00 Saturdays on 570 KLIF AM. E-mail: email@example.com; read all of Ed’s work at www.insideautomotive.com.