Modal differences in travel time

[cimmee]

https://www.nature.com/articles/s41598-020-61077-0
"Cities worldwide are pursuing policies to reduce car use and prioritize public transit (PT) as a means to tackle congestion, air pollution, and greenhouse gas emissions. The increase of PT ridership is constrained by many aspects; among them, travel time and the built environment are considered the most critical factors in the choice of travel mode. We propose a data fusion framework including real-time traffic data, transit data, and travel demand estimated using Twitter data to compare the travel time by car and PT in four cities (São Paulo, Brazil; Stockholm, Sweden; Sydney, Australia; and Amsterdam, the Netherlands) at high spatial and temporal resolutions. "

I've been looking at efficiency studies (car v train) and have always found them lacking in that they are one dimensional. In other words, solely focusing on energy consumption ignores an important variable (time). Time has a cost as does energy. Ignoring time as a cost factor when determining efficiency invalidates whatever "study" is presented.

I would posit that rail is much less efficient (in the USA) when viewed in totality.

 

[Trollopian]

I've been looking at efficiency studies (car v train) and have always found them lacking in that they are one dimensional. In other words, solely focusing on energy consumption ignores an important variable (time). Time has a cost as does energy. Ignoring time as a cost factor when determining efficiency invalidates whatever "study" is presented.

Economist here, and I completely agree that time has a cost (an "opportunity cost" in econspeak). But even then the analysis isn't clear-cut. A merit of train travel and public transit is that the traveler can do other things. Productive things or leisure things, whichever. People regularly work on commuter rail or Acela. On Washington's Metro I've seen lawyers reviewing briefs and civil servants marking up drafts and fellow economists eyeing reams of very dull-looking statistics. Admittedly, not so much on (e.g.) the Zephyr. But driving takes, or should take, constant attention and mental energy. That's a time drain.

In very dense cities like New York and London that are hives of high-paid and modestly-paid employees, a train commute will typically beat the same trip by auto.

(That actually is a major reason I worry about self-driving cars. They will upheave the economics of driving and of urban sprawl, once a car commute no longer requires a driver.)

 

[JamesWhitcombRiley]

I wish self-driving cars only self-drove on highways with guide lines, as the original plan was.

OK, back to the original post.

A private jet is pretty time-efficient. Yeah I think as a practical matter, all of us have chosen driving over a train depending on circumstances. It's been fascinating reading the posts about India. That's a middle income country, but in truly developing places like Lagos people try to do as much business as possible by text or video, due to unbelievable road traffic. Before the smart phone even, a guy from there was telling me about 7 second voice billing, and message rates. Meanwhile we do have this massive investment in highways, so it can be pretty efficient. But say, Norfolk VA to Alexandria/Washington, people take the train because of the massive clouds of uncertainty and stress and cost trying to drive it (about 4 hours) during business or weekend hours, really any time the sun is up or has been recently, or will be soon. Might apply especially to people who do the trip often. Ridership is way up, anyway.

Also, the environment.

Another thing about time, the quickest the railroads will get freight from Norfolk to a place like Washington or Hagerstown is one day. For bulk, a train carries 4 times a truck. A barge on a tow on the Mississippi carries 10 times a train car. A barge is really slow, but cheap. The Port of Virginia runs a barge service between Richmond and Norfolk/Newport News, perhaps not wildly popular. Tacoma to Alaska, of course, it's the major mode. Not that different than what you're saying perhaps.

 

[MARC Rider]

May commute from Baltimore to Washington back when I worked was about 2 hours, including a drive to the station, an hour ride on MARC, the Metro from Union Station and a short walk. A couple of times I drove down for various reasons, and the drive time was about an hour and a half. I still continued to take the longer public transit commute because, though faster, driving was much more stressful. If you weren't stuck in stop and go traffic, you were in bumper to bumper traffic going 70 mph. With big trucks and cars changing lanes in fron of you without warning. In the dark.

Maybe if I had a nice plush limousine with a chauffeur driving it might be different.

Also, the 40 mile trip by car had operating expenses of about 60 cents per mile or $24. Plus parking in downtown DC is not cheap. A MARC one way ticket is $8, though most regular commuters buy monthly and weekly passes that come out to less per ride. I hardly paid anything because I had a transit subsidy from my employer.

 

[slasher-fun]

I've been looking at efficiency studies (car v train) and have always found them lacking in that they are one dimensional. In other words, solely focusing on energy consumption ignores an important variable (time). Time has a cost as does energy. Ignoring time as a cost factor when determining efficiency invalidates whatever "study" is presented.

Except that for commuting, time is a fixed factor, not variable: people will on average live a certain time away from work (usually ~30 min), not a certain distance away.
That's why widening highways to solve congestion never works: what used to be 30 min away turns into 20 min away, so people are more willing to make that was-30-now-20-min trip, so there's more traffic, so... congestion increases, and it quickly goes back to still-30-min-away.

Meanwhile we do have this massive investment in highways, so it can be pretty efficient.

Highways cannot be efficient. Or actually no: highways with buses are efficient, highways with cars are not.

Let's take for example a perfect single lane 30 mph street. No intersection, no traffic lights, no pedestrian crossing, perfect drivers.
On that street, you have capacity for about either:
- 1500 cars per hour. There are usually 1.5 occupant per car, so that's 2250 people per hour, but assuming all cars have 5 seats, that could be 7500 people per hour if they're fully used
- 1200 buses per hour. Not sure about the average occupancy, but peak occupancy means 125 000 people per hour. Yup, as many as if you had 17 lanes for cars
- 20 000 pedestrians per hour (assuming here that's a one-way pedestrian street with perfect pedestrians of course)

If you want to have as many people as possible travelling as fast as possible, you make it so that cars are only used as a last resort solution, when all other modes of transportation aren't a plausible solution for that very trip you're trying to make.
But if you want to have as many people as possible travelling by car, you'll only have a small number of people travelling vs what you could have had with more efficient modes of transportation.

 

[Tlcooper44]

There isn’t any accurate or meaningful evidence to suggest a car-dependent, or even car-centric society is efficient by any factor, especially in urban areas.
In North America, we have somehow convinced ourselves otherwise, and most DOT’s will be happy to explain some convoluted reason as to why their highway boondoggle is the right course. This doesn’t make it true.

A car accommodating society (the Netherlands, Germany, Singapore, etc..) is even better for those who want to drive.

 

[Literalman]

A perfect street has no pedestrian crossings: I know it was a hypothetical example, but I know the mindset. I've had drivers yell at me and swear at me for crossing a street when they wanted to run a red light or stop sign but I was in the way. I want to write something soon about how pedestrians subsidize drivers. Where I live, driving is fast because drivers don't have to obey speed limits or obey red lights or stop signs, and most drivers don't (yes, I've counted). Driving is convenient because drivers don't have to park legally; they can park in crosswalks and on sidewalks rather than farther away.

Certainly in Philadelphia, where I go about once a week, the subway is much faster than driving. A couple of times we've taken a cab from 30th St. to Drexel Hill because of the weather, the hour, luggage, or all three. I noticed how long it took in the city. I would think, "We're only up to 52nd Street?" And regional rail is faster than the subway.

Wait time is a factor too. With the subway, it's usually only a few minutes. With regional rail, I have to plan the trip to minimize wait time; otherwise it could be an hour or more. With the cab at 30th St., wait time was basically zero, but one time we called for a cab to go to an emergency room, we waited all night, and when I called to cancel the request I was told that the cab company had canceled it after three hours but hadn't told us.

When I commuted on Virginia Railway Express, my final commute involved a 9-mile drive, a walk to the station, an hour and a half on the train, and a bus trip. It took five to six hours a day. The Washington, DC, area has lots of long-distance commuters because the cost of housing near the city was too high for many workers. Also, my first job in Virginia was in the next county, not 50 miles away, but when I had to find a new job, I ended up working in Alexandria. We couldn't afford to move, even to a much smaller hours. When VRE didn't run, I sometimes drove. It was about 50 miles door to door, the office overlooked an exit from 395, and I lived about three miles from I-95. So 47 of the 50 miles were on Interstates, but it never took less than 2 hours.

Finally, consistency is a factor (not that VRE was consistent; I got really tired of VRE telling us to call after 4 am to see whether VRE would run any trains the next day). One morning in Fredericksburg I was talking to a guy who worked in Maryland on the other side of Washington. He said he took the train even though the whole trip was 3 hours one way because when he drove, sometimes it took 2 hours and sometimes 4.

 

[Ziv]

Multiple guesses ahead warning:
Public Transit works best where it is ubiquitous, cheap and fast, or so I would guess. Which pretty much rules out buses unless they have a dedicated lane or ground level trams for that matter. What would work best, it seems, would be an underground system with a dense network of tubes operating driverless trains at speeds of at least 50 mph, with at least 20 trains per hour but 30 tph would be even better during rush hour.
The main problem in building such a system is that to get Manhattan/Paris/London/Moscow levels of ubiquity of service you would need to spend an incredible amount of money today.
UNLESS the Boring Company ever gets to the point where they can build a 15' bore tunnel at prices close to the ones Musk has spoken of. But Prufrock-2 is capable of building just a 14' bore tunnel at this point (or so I have read) which gives you just over 12' of finished tunnel (IIRC) which means even the low profile 1996 Stock trains (New Tube for London?) like those on the Jubilee Line will not (?) fit because they are 9'6" tall. [Prufrock-3 is in testing now, is it a 14' bore?]
If you are going to build a cheap, fast and dense public transit service it would be best to use existing rolling stock with minimal modifications, i.e. contract to have Alstom build a slightly updated version of their 1996 stock here in the US. The one thing that would make it more popular is to make it taller, even 3" would help. Americans may not be Dutch levels of tall, but we are fairly tall, and wide. I could see a "Beltway" line being really useful in DC, by building a ring line tying all the metro lines together at stations that are just outside the Beltway.
It really comes down to how successful TBC is with their new Las Vegas Loop.
I have to admit that I do not know as much as I would like about TBC, existing metro systems or state of the art driverless trains, for that matter, but it is fun to think about American cities with state of the art driverless train subway systems being built over the next decade. Think DC, Chicago, Houston, Phoenix, LA, San Diego, Denver, Miami, etc... all with 2 or 3 new subway lines being built all at the same time. Some of these would be building additions to their existing subways/elevated trains and some would be building their first subway systems but the lower cost of the building process and operating the trains would transform commuting in American cities.
Ok, I have come to the end of my wish list and will show myself to the door.

 

[Tlcooper44]

Multiple guesses ahead warning:

Public Transit works best where it is ubiquitous, cheap and fast, or so I would guess. Which pretty much rules out buses unless they have a dedicated lane or ground level trams for that matter. What would work best, it seems, would be an underground system with a dense network of tubes operating driverless trains at speeds of at least 50 mph, with at least 20 trains per hour but 30 tph would be even better during rush hour.
The main problem in building such a system is that to get Manhattan/Paris/London/Moscow levels of ubiquity of service you would need to spend an incredible amount of money today.
UNLESS the Boring Company ever gets to the point where they can build a 15' bore tunnel at prices close to the ones Musk has spoken of. But Prufrock-2 is capable of building just a 14' bore tunnel at this point (or so I have read) which gives you just over 12' of finished tunnel (IIRC) which means even the low profile 1996 Stock trains like those on the Jubilee Line will not (?) fit because they are 9'6" tall. [Prufrock-3 is in testing now, is it a 14' bore?]
If you are going to build a cheap, fast and dense public transit service it would be best to use existing rolling stock with minimal modifications, i.e. contract to have Alstom build a slightly updated version of their 1996 stock here in the US. The one thing that would make it more popular is to make it taller, even 3" would help. Americans may not be Dutch levels of tall, but we are fairly tall, and wide. I could see a "Beltway" line being really useful in DC, by building a ring line tying all the metro lines together at stations that are just outside the Beltway.
It really comes down to how successful TBC is with their new Las Vegas Loop.
I have to admit that I do not know as much as I would like about TBC, existing metro systems or state of the art driverless trains, for that matter, but it is fun to think about American cities with state of the art driverless train subway systems being built over the next decade. Think DC, Chicago, Houston, Phoenix, LA, San Diego, Denver, Miami, etc... all with 2 or 3 new subway lines being built all at the same time. Some of these would be building additions to their existing subways/elevated trains and some would be building their first subway systems but the lower cost of the building process and operating the trains would transform commuting in American cities.
Ok, I have come to the end of my wish list and will show myself to the door.

While this stuff matters, public transit is most successful when there is good land use around stations - the stations are destination in and of themselves, oftentimes owned by the transit company. Hong Kong, Tokyo, and Singapore understand this.

As long as most American transit stations are parking lots; as long as density is essentially illegal in zoning codes; as long as minimum parking requirements rule; public transit will always struggle here (there are other reasons too, I just view these as the most important ones).

 

[AmtrakMaineiac]

A perfect street has no pedestrian crossings: I know it was a hypothetical example, but I know the mindset. I've had drivers yell at me and swear at me for crossing a street when they wanted to run a red light or stop sign but I was in the way. I want to write something soon about how pedestrians subsidize drivers. Where I live, driving is fast because drivers don't have to obey speed limits or obey red lights or stop signs, and most drivers don't (yes, I've counted). Driving is convenient because drivers don't have to park legally; they can park in crosswalks and on sidewalks rather than farther away.

Certainly in Philadelphia, where I go about once a week, the subway is much faster than driving. A couple of times we've taken a cab from 30th St. to Drexel Hill because of the weather, the hour, luggage, or all three. I noticed how long it took in the city. I would think, "We're only up to 52nd Street?" And regional rail is faster than the subway.

Wait time is a factor too. With the subway, it's usually only a few minutes. With regional rail, I have to plan the trip to minimize wait time; otherwise it could be an hour or more. With the cab at 30th St., wait time was basically zero, but one time we called for a cab to go to an emergency room, we waited all night, and when I called to cancel the request I was told that the cab company had canceled it after three hours but hadn't told us.

When I commuted on Virginia Railway Express, my final commute involved a 9-mile drive, a walk to the station, an hour and a half on the train, and a bus trip. It took five to six hours a day. The Washington, DC, area has lots of long-distance commuters because the cost of housing near the city was too high for many workers. Also, my first job in Virginia was in the next county, not 50 miles away, but when I had to find a new job, I ended up working in Alexandria. We couldn't afford to move, even to a much smaller hours. When VRE didn't run, I sometimes drove. It was about 50 miles door to door, the office overlooked an exit from 395, and I lived about three miles from I-95. So 47 of the 50 miles were on Interstates, but it never took less than 2 hours.

Finally, consistency is a factor (not that VRE was consistent; I got really tired of VRE telling us to call after 4 am to see whether VRE would run any trains the next day). One morning in Fredericksburg I was talking to a guy who worked in Maryland on the other side of Washington. He said he took the train even though the whole trip was 3 hours one way because when he drove, sometimes it took 2 hours and sometimes 4.

Your experience in commuting in the Philadelphia area is interesting as I did some experimentation with commuting options when I lived there. We lived in Swarthmore within a longish walk or short drive of the Media/Elwyn line station. My wife for a while commuted daily on the train to a job in center city. My job however was in North Wales to the north of the city and a little over an hour's drive on a good day. Once the Center City Commuter Tunnel opened I tried using the regional rail, riding into suburban station on the Media/Elwyn line and changing to a Lansdale/Doylestown train usually at Suburban Station. I found this to be over 2 hours commuting time due to the wait between trains which didn't run all that frequently even at rush hour, and just the time it took especially on the line from Center City to North Wales. Eventually the opening of the "blue route" (I-476) between Chester and Plymouth Meeting meant I could drive in a little less than an hour, unless of course it was snowing, or there was an accident, or whatever. Coming home there was also the regular jam where the road stupidly went from 6 to 4 lanes, which was done as some sort of compromise with the various NIMBYs in the area to allow the road to be built at all .

 

[fdaley]

While this stuff matters, public transit is most successful when there is good land use around stations - the stations are destination in and of themselves, oftentimes owned by the transit company. Hong Kong, Tokyo, and Singapore understand this.

As long as most American transit stations are parking lots; as long as density is essentially illegal in zoning codes; as long as minimum parking requirements rule; public transit will always struggle here (there are other reasons too, I just view these as the most important ones).

Yeah, the assumption among American planners -- and Canadian ones too, given how the new Montreal REM is set up -- seems to be that almost everyone will continue to travel by car in perpetuity beyond the dense urban cores of major cities. So you wind up with transit and commuter rail systems where the outer-end stations are basically huge park-and-ride facilities. And while in many cases that serves the current development pattern, over time you'd expect these suburban transit hubs to foster denser, more walkable neighborhoods. But as you point out, the zoning laws in lots of suburban communities actually prohibit this level of density. And that keeps most of us who live outside the urban core stuck in our cars for a great deal of our travels.

 

[Literalman]

Just to clarify one thing about my comments: I don't commute in Pennsylvania. I retired before moving, and we live car-free, so I ride Septa a lot: trolleys, buses, regional rail, and subway. I commuted for 26 years in Virginia, sometimes into Washington.

 

[allanorn]

I haven't read the paper in its entirety yet but it appears they've come to similar conclusions as other active travel studies: public transit works great at short distances and when road capacity is constrained during peak times. I'll read more to see if what angle are they trying to cover given that it's in Nature magazine, but greenhouse gas emissions reductions seem to be the hook.

At the distances they're mentioning where public transit works (3km/2 miles or less), my guess is that it's true there are no transfers to other transit that increase total trip times. Even then, at 3km I would think bicycle can be faster than transit unless your public transit line runs frequent enough that schedules don't matter. I am really curious if built environment and land use is covered in the study to any significant degree.

They may be missing the same argument that @Trollopian has presented: time-efficiency (e.g., shortest trip time) vs. time-effectiveness (what you do with that time). I get a lot of arguments against night trains and California's HSR program because they're "too slow" relative to airplanes or day trains. Usually the argument centers around the short trip time, so I have to bring up that night trains are moving hotels that aren't desired to go fast, or HSR allows you overall more productive work time with a potential tradeoff in trip time.

Just by reading the abstract, I don't think the efficiency argument could win out unless we're discussing longer-distance commutes such as regional and intercity rail. I haven't seen their proposals to solve an issue if they're going to dive into that direction.