We Once Had Self-Driving Cars

(Photo: Elliott Brown)

A frequent response to the dysfunctions of American air travel is technological: namely, self-driving cars (also see this article). In a self-driving car, you can relax, even sleep, while being driven safely to your destination at 60 mph. We once had such a system. It’s called a train network.

Compared to air or car travel, a decent train network is cheaper, more environmentally friendly, and quicker. As an example, I’ll compare two door-to-door, city-center-to-city-center journeys.

The first is by plane, from Boston to New York City:

0:40 Get to Logan airport (airports are typically far from the city center)
1:00 Participate in security theater, etc.
1:10 Sit in loud sardine can
0:30 Get luggage
1:00 Get from Newark airport to destination in New York City
—- ———–
4:20 TOTAL

The journey is 215 miles, for an overall speed of about 50 mph. The current cheapest advance-purchase, nonrefundable round trip costs $150; that’s 35 cents per passenger-mile. For comparison, car travel is about 50 cents per passenger mile (the IRS-set reimbursement rate).

The comparison journey is by train from Paris to Lyon (I need a decent train network for comparison purposes). I made this journey when our family lived for a summer in Lyon, and I learned French by street-fighting methods. The ticket told you where to stand on the platform. Two minutes before scheduled departure, the train arrived. You stepped into the train car, found your reserved seat, and, almost before you had stowed your luggage, the train departed.

0:20 Get to station in downtown Paris
0:05 Arrive early to validate ticket
2:00 Sit on comfortable train with leg room, or stand up without hitting head
0:25 Get to destination within Lyon
—- ——–
2:50 TOTAL

The journey is 289 miles, for an overall speed of about 100 mph. A current advance-purchase, non-refundable round trip costs 70 euros or $100; that’s 17 cents per passenger mile, or one-half of the plane’s cost. The train journey is not only cheaper and quicker, but you get more time to think and read: 2:00 on the train versus 1:10 on the plane.

Forget self-driving cars! If we can print trillions of dollars to create moral hazard by bailing out the gamblers who nose-dived the world economy, why not print money to extend and upgrade the rail network? The U.S. and U.K. rail networks were once twice as extensive as they are today.


The TGV is wonderful, but uh, dude, that 17 cents per passenger-mile includes $1B per mile capital cost in building the money-losing railway. Maybe you should like study "economics" or something.

Enter your name...

And how much did those two airports cost? (Hint: Logan alone had a four billion-with-a-B dollar project a few years back).

How much of your tax money is spent on the FAA and air traffic control?

Rail is always cheaper. The only problem is that even high-speed rail isn't fast enough for cross-country travel.

Jamie M

When a family of four travels by train they buy four tickets. When they travel by car, the cost is the same as for one passenger. Also include the travel time of getting to the train station.
When I travel to Lyon (from Geneva) I travel by car. It's cheaper, faster, and more convenient than the train.

Your article is great for single travellers, but does not cover the very common family scenario.


Self-driving cars will be able to travel faster and at closer distances than people. this would decrease the cost (drafting) and time of car travel. That would change the equation a bit from using IRS reimbursement rate and using a simplistic 60 mph.

Plus, the infrastructure is already there (during the transition, we could make the left-lane on highways self-driving cars only, much like HOV lanes) to support self-driving cars as a mode of transportation so little additional public money is needed to get it going.

Kyle Anderson

Yes, trains work great for travelling from city center to city center when cities are not that far apart. But most travel isn't from Boston to New York, it is from some suburb somewhere to some other suburb (or city or town) somewhere. And developing a train network, that can have all those hubs, and still get people where they want to go quickly is not easy or cheap. The future of transportation needs to be about taking people where they want to go, not to some central point in a large city. That is the appeal of self-driving cars.


First of all, what makes you think the US wouldn't have the TSA for trains if that were the popular form of transportation? So add an hour of useless security theater for rail. Now multiply the fact that the majority of air travel isn't 215 miles between Boston and New York, it's between Atlanta and Seattle. The time cost of all of the security theater and baggage really pay off on the longer trips like we have in the US, but are less common in Europe. And finally, rail is generally much more government subsidized than air travel, thus distorting the whole price information that you based this all on.

Homer Simpson

Europe also has a very established, condensed geography. Cities and living arrangements were established around ancient travel infrastructure. America, especially the American West, sprouted in the age of urban sprawl, personal automobiles and air conditioning. Trains are an excellent solution to some problems, but they're not a universal panacea. Furthermore, trains are so infrastructure-intensive that they're often only feasible when heavily subsidized (See California high speed rail for a current case study). This inevitably leads to a pork-barrel mentality, which is the mortal enemy of efficiency.

In my opinion, buses suck, at least in their current iteration, but they are a much more realistic public transport solution for the American landscape. They have the advantages of trains- i.e. the ability to sleep, work or otherwise spend travel time in some way other than driving, and they also have the unique advantage to change with the times. Train tracks cannot be rerouted as cities or regions fall out of favor. A train network in currently vibrant fracking communities would probably be useless in a few years when the boom has moved on. Buses are adaptable. If there's a need, the market will fill it. If there's no need, the market will stoically move on.

Trains are good, in some circumstances. Unless they can be 100% privately funded and operated, they should not be implemented here in the U.S., as doing so with subsidies is to introduce the most expensive opportunity cost imaginable.

I'm really surprised that this article was on Freakonomics, I usually find the articles here very well informed and rational. Not today.

For a better analysis, see the Monorail Simpsons episode.


Joel Upchurch

I actually think we will have robobuses in the future. People will register their destination and when they need to be there. The system will send a text message with the pickup time during the night. The route will be adjusted based on demand. The trouble with current buses is that they run all day, even when there is no demand. The manual drivers are the worst part of the current bus system.

Eric M. Jones.

We have had self-driving cars since June of 1942: See:


That's the Popular Science June 1940 cover article that promised them in two years.

But seriously, in "Automation" there is the principle of "1-2-Far", and I think self-driving cars is it. I love trains. One only has to live in NYC for a while to see how life is better not looking for a parking place.

I have published previously that the money given to Wall Street (>$100/Earthling) would buy a twin track 10,000 km maglev train system including 300 kph trains, tunnels, bridges and stations, between most major US cities. Average construction cost would be only about $50 million per kilometer. About 1200 maglev cars would be needed. The rolling stock would cost only $20 billion.

Make no half-vast plans.

[WORDPRESS HASHCASH] The poster sent us '0 which is not a hashcash value.



I love train travel, don't get me wrong. But 99% of my trips are within one city, not between cities. I'm grateful that Montreal has a really good public transit system, but if I need to do groceries or visit someone in the suburbs, I need to use my car. I'm still waiting for the day that it drives me around and lets me read my book during the trip.

Doug Goodman

This seems like a red herring. Taking the $107 midday Acela fare from Boston to New York, that's 50 cents per passenger mile for the 215 mile trip. Slightly less expensive than the shuttle trip to Newark (which is $150/215, or 70 cents per mile). It's no faster -- even the Acela is 3.5 hours, plus travel to/from the train stations within town. And that's probably the single route in the US where train travel has the biggest advantage.

Comparing US train travel to French is at best unfair, and the conclusions are misleading. European rail prices are well under market rates due to generous subsidies:

From the Amtrak Office of the inspector general: "After examining a representative sample of European Passenger Train Operations over a multi-year period, we found that:
a) When all revenues and expenses for the entire passenger train system are taken into consideration, European Passenger Train Operations operate at a financial loss and consequently require significant Public Subsidies, and
b) The average annual subsidies for European Passenger Train Operations are much higher than those for comparable Amtrak services."




arrive at boston train station: 30 minutes (most folks dont live right downtown close to central train stations)
pre-boarding: 15 minutes (who arrives for a train trip with only 5 minutes to spare?)
trip: 4 hours
pen station to destination: 30 minutes

total: 5 hours 15 minutes.
cost $100

not as shiny, but not bad.

David Rademeyer

Actually, we still have them: Greyhound bus.

0:20 Get to South Station in downtown Boston
0:15 Arrive early and get in line
4:20 Sit on slightly cramped bus
0:20 Get to destination within New York
—- ——–
5:15 TOTAL

The overall speed is low, but you get over 4 hours of reading time, and the cost (at ~$50 round trip) is a can't-be-beat 12 cents per passenger mile.


Aren't trains hugely subsidized by government in England and Europe, and isn't our unwillingness to do this here vs. for-profit airlines the reason we don't have more affordable and better train service?


There are several things wrong with this. First of all, most Americans do not live in the city center, but rather in the suburbs, so a 40 minute drive is not typical, my own experience with security theater is closer to 20-30 mins, and if you carry your bags with you there is no wait afterwards. Basically you are taking a worse-case scenario and presenting it as typical.

Furthermore, while you compare Boston to NYC, how representative are such short hops of most airplane flights? Maybe they are, but no data is presented. Try comparing flying from Boston to Chicago, Atlanta, Dallas, Denver or anywhere on the West Coast and let me know how the flight vs. train comparison stacks up.

In addition, the idea that spending billions on train upgrades is justified because billions were spent bailing out banks is nonsensical. It is precisely this kind of attitude that has helped produce the massive debt and deficits we now face.

None of this is to suggest I am anti-train. I have taken high-speed rail around Spain (which is highly subsidized -- my vacation was paid for in part by Spanish taxpayers) and the Acela from DC to NYC (although only because my employer picked up the tab -- when traveling for personal reasons I take the bus which is FAR cheaper and only marginally more time consuming). Rather, I think we should privatize Amtrak, as well as airports, and let the chips fall where they may. Instead of having central planners and observers like Sanjoy Mahajan determine how the public should travel and placing their thumbs on the scale with government dollars, let's let the market -- which is to say the people -- decide (for the record I also support things like tolls, mileage and congestion fees for cars so that autos pay their own way and account for externalilties they generate).



I like the idea. My problem is the livability of the swath of land where the train goes ripping through at 100 miles an hour. I live in a northern California long narrow beach city kept narrow by mountains. It is one of the most beautiful places in the world with a range of income levels which goes pretty low as well as somewhat high. A huge number of folks live right at the train tracks. One of my friends can waive to the engineer from her kitchen window (and he waives back). A commuter train will make life horrible here.

There is a similar town in southern Cal. which has a commuter train, and you can see the destruction it has caused. Basically it has cut the town in half and ruined the community feel of it.

The notion of improving train travel is focused on travel, not on the folks who live near the tracks. It makes me nervous. Perhaps if they go right down the middle of the freeway and have quieter horns?



Apples vs Oranges often works out in your favor. You need to add "security theater " to US train travel -the TSA will want to expand if rail networks expand! So now your benefit is only 30 minutes; maybe someone can do a cost/benefit analysis on that.

Don't forget the cost of car rental; a self driving car is not only point to point transport sans TSA. Even long trips may benefit from Self driving cars, since they can drive much faster safely than a human driver, and without rest breaks.

Bring on the robots!


For Boston to NY, I agree and have used the train. It's nicer than the airport/airplane experience but not by much. Have you BEEN to Penn Station? It's about the only place I know that makes LGA look nice by comparison.

Also.....security theater has come to AmTrak. :-(

Caleb B

Oklahoma City to Dallas would be an example of why rail doesn't work all the time. City center to city center means nothing, I still need to drive once I get there.

Now a train that could carry my car? That would be an improvement. I'd pay to avoid driving, saving on gas, getting free time, etc.

Joel Upchurch

If you look at energy cost per passenger mile then trains have a small and decreasing advantage over planes. According to the Transportation Energy Data Book (ed. 31 table 2.14). Planes cost 2.85 btus per passenger mile and a trains run 2.52 btus. This is because the aircraft companies are much better at filling their seats than trains. I also live in Orlando and as near as I can tell there is over 100 cities I can fly to. With one change of planes, I can travel to most of the earth. Trains look lame compared to that.