Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Another Way Driverless Cars Might Boost Economic Growth
One person can only know so much. And one person can only create relatively simple products on their own. Complex products require networks of people, sometimes called companies, and even networks of networks, such as supply chains. And one way of evaluating an economy is by its ability to create complex products.
In his book, “Why Information Grows,” Cesar Hidalgo writes about economies as “collective computers” whose computational capacity is either expanded or limited by the size of social networks. (I reviewed the book recently.) And the ability to create denser networks is helped or hindered by communication and transportation technology, among other things. All of which came to mind when reading the new working paper “The Role of Transportation Speed in Facilitating High Skilled Teamwork” by Xiaofang Dong, Siqi Zheng, and Matthew Kahn:
This paper argues that China’s investment in High Speed Rail creates an integrated, regional system of cities close enough to travel by fast train but far enough to not be car friendly. We have studied the productivity impacts of cross-city transport improvements by focusing on publication and citation patterns of China’s university researchers. The empirical results in this paper show that once a city is connected into the HSR network, the researchers in that city will experience significant productivity increase in terms of quantity and quality of journal publications.
We find that travel speed facilitates matching and idea flows between two HSR-connected cities. Larger productivity gains are observed for the secondary cities close enough to the mega cities to access them by HSR. We find larger productivity effects for social scientists and for the incumbent coauthors (the intensive margin). For the subsample of migrants, we find that they are more likely to choose those secondary cities that are directly connected with mega cities by HSR, compared to other secondary cities. These empirical findings bolster our claim that cross-city speed facilitates learning and matching across cities. This finding has implications both for efficiency and for equity in the modern Chinese economy. In a human capital based economy, high speed rail induced reductions in transportation costs increase regional productivity by improving matching and lowering the cost of face-to-face interaction. . . .
Our main finding that faster cross-city commuting speeds enhance productivity extends the original Gaspar and Glaeser (1998) work in a new direction. They argue that cities and information technology are complements and not substitutes. The benefits of face to face interaction increase if strangers recognize that once they have met that they can subsequently connect again by phone, Skype and email. Cities exist because they economize on transportation costs. The boundary of a city’s agglomeration area is endogenous and hinges on transportation speed. If new technologies such as high speed rail effectively make nearby cities “closer” to superstar cities (through moving at a faster speed), then agglomeration benefits spread out further across space.
Now I am not suggesting the US build its own pricey high-speed rail network. But in a world of widespread autonomous vehicles, one can imagine that many more cities might be considered car friendly and thus reduce the cost of in-person interactions.
I recently took an Acela high-ish speed train from Washington’s Union Station to Wilmington, Delaware. If I owned an AV, I almost certainly would have chosen that instead of a train. Indeed, a 2016 Boston Consulting Group report concludes that AVs will “constitute a tangible threat to passenger rail within the next one or two decades.” Or as Matt Ridley has written, “I just cannot help feeling that a very fast train, built at glacial speed (half a mile a week) over many years of consultation, review and challenge as it punches through Nimbyland, and at up to nine times the cost per mile of French high-speed rail, feels like a white elephant waiting to happen.”
Published in Economics, Technology
Driverless cars are a solution looking for a problem.
Hear, hear!
Ah increased productivity. That stuff of life. Our true raison d’etre.
If we just build high rises over the Chinese factories, we can reduce commute time 250%. Production of our stuff will sky rocket. Autonomous vehicles will be as popular as dense urban centers and with the same people, I suspect.
Where public transportation works, it works because it allows large numbers of people to move predictably between particular places — places where large numbers of people would presumably travel anyway. But this involves a trade-off. That trade-off is control. If I choose to use public transit, I lose my ability to control where I travel (beyond the list of predetermined stops) and when I travel. I exchange unfettered movement for movement along a fixed route. I trade spontaneity for certainty.
Self-driving cars might mitigate this problem, but they can’t solve it — at least, not if they allow users the sort of control they’ve come to expect from good, old-fashioned hands-on-the-wheel driving.
Fully autonomous cars might be useful for commutes, but then, most driving isn’t commuting.
Suppose I’m running errands, and I want to change my destination en route. Suppose I realize, as my new Waymo whisks me past Kroger, that I forgot to add toothpaste to my grocery list, and that I ought to stop by. Suppose I want to park near the Lowe’s home-and-garden section, because it’s closest to the item I happen to need. Suppose I want to choose one parking spot rather than another. How, exactly, is a fully autonomous car equipped to handle these things?
And what about sightseeing? What about vacationing? What about visiting Grandma Wilson, who happens to live on a rutted dirt road in an area without reliable cellular reception? What if I want to stop by the roadside and snap a quick photo of some lovely, listing abandoned farmhouse? What if I have no destination in mind?
I’m with Charles Cooke. The futurists can pry the steering wheel from my cold, dead fingers.
You know, we’ve spent 5,000 years trying to define cause and effect. We still have a hard time. That a key component of the Chinese study happens to be the impact of a state built high speed rail line, makes me look askance.
I know the roster’s crow does not bring up the sun. I know wet streets do not cause rain. I know elephants can’t fly. And I know governments are always looking for someone to validate their infrastructure decisions.
That is the exact phrase used about Multimedia Computing in the 90’s.
They did eventually find a problem for the multimedia computer to solve with MP3s, evolving into Napster, ITunes and finally netflix. (or at least finally for now.)
I now see the light, the driverless future will be here as quickly as Napster.
There are lots of ways to approach the commute, for years I had a driver and that was a superior way to commute, and like trains and buses, driverless cars are going to be far superior to driving for long commutes. They’re not superior to short commutes or rotating car pools and HOV lane. I just spent the winter in a Latin American city where one gets a sense of just how complex city traffic can be. I didn’t drive, I no longer have the reflexes that would allow me to drive in that kind of chaos but it works and was amazing to behold, but it will ultimately grind to a halt from shear congestion. Presumably driverless cars would eliminate the chaos but not the growing congestion and only by eliminating cars with drivers. Traffic is like an economy, or all emergent systems, it’s an infinitely complex system of chaos with hundreds of thousands of individual actors accommodating each other ultimately producing order. Drivers get instant feed back on everything going on around them. That can’t be programed for any more than an economy can be managed through five year plans, or controlled by a bunch of remote bureaucrats with no real time information. We could do it of course by imposing driverless cars on all big city commuters, but to do so we’d have to eliminate choice. We have choices with car pools, and mass transit and those choices get more attractive as congestion grows. I can’t imagine Tokyo or Manhattan without mass transit, their streets could not accommodate cars with or without drivers had they no mass transit and a willingness to walk intermediate distances. Freedom and choice, those things run the world and every time we try to eliminate them because we know we’re so clever and technologically advanced we replace ordered chaos with disordered chaos. Ordered chaos is the way our universe runs. Disordered chaos is what you get when a clever technological elite think they can do better. They can’t.
They are two completely different things. I am a software engineer and had the opposite view on “multimedia computing in the ’90s.” That doesn’t mean I’m right about driverless cars, but it is my considered opinion that they will not be a major economic force in the near future.
They are mis underestimating the complexity of coordinating this. It’s on par with the vincible ignorance displayed by experts who presume an economy can be orchestrated like a symphony.
What about tunnels through mountains, or under rivers? What about being stuck in traffic en route to the hospital with a baby on the way (think @BethanyMandel). Will the autonomous car obey the command to drive on the shoulder?
I can’t imagine what the legal issues will be. Who’s liable in an accident? The owner of the car? The builder? The programmers? The makers of the radars? The passenger (who might not even have a driver’s license)? I guess the legal profession will be the group that sees the economic growth the most.
I apologize if I sound too harsh about some of James’ posts. I miss his old podcast, and wish he’d jump back into them.
I hate to sound like a liberal, but I would love to see passenger train service make a comeback in the country.
However, all the current talk is usually about high-speed rail service, which is fine for light-rail commuting. For some of us, the trip is about the experience. I would love the leisurely, two-day journey from Chicago to San Fran on the California Zephyr. One of my goals in retirement is to make said trip. Hmmmmm . . . this could be a two-day, Ricochet Meetup-On-Rails . . .
But back to the comparative costs with France. What is the reason the costs in the US are 9 times that of the French? What do they do right that we do wrong?
You must be referring to the cost paid by the passenger as being 1/9 the cost. I doubt the true variance is that great. Governments subsidize transportation. We just do it a bit less than the Europeans.
In the US, 90% of rail traffic is freight. The distance from Paris to Moscow is roughly the same distance as LA to Kentucky. Your future microwave oven doesn’t care about arriving in Paducah at 2:12 AM.