[niftich]

Despite several flaws and uncharitable misrepresentations in his policy analysis, I actually agree with most of O'Toole's conclusions.

I post a lot on here about rail, and I disagree with a lot of output from the Cato Institute, but the infrastructure cost of long-distance High Speed Rail in the US would be immense, and the geography of settlement and commuting patterns in the country are too car-oriented to take advantage of passenger-only HSR. Americans already do most travel by car and long-distance travel by air, so High Speed Rail would be a slower and costlier substitute for long-distance flights, and a maybe-faster but drastically less flexible alternative to <300 mile travel.

It's no accident France runs the TGV like an airline, because they work the same way: you get to your destination station, and now what? You need to hop on public transit or rent a car like you would at an airport. But in Europe, a big town is far more likely to have public transit of acceptable quality, frequency, and coverage to solve the last mile problem; all but the most transit-webbed US cities do not.

This article exposes some of the flaws and misrepresentations in his analysis, but then contributes its own flaws in turn. One example: in truth intercity buses are very widespread in Europe, and not only do they fill in gaps left by the rail system, but thanks to expressways on some routes they be as fast as "moderate-speed trains" too. Truly High Speed Rail only runs in a dozen corridors in Europe, and the rest of the passenger rail on the continent runs the gamut from decent to atrocious. Buses are flexible, because they can go where the roads already go.

One way to get around the last-mile problem is to ensure your destination is likely to be transit-webbed town, or your destination station is very close the location to which you actually want to go. This sounds a lot like commuter rail -- the speed depends on how much you want to spend on infrastructure. In the US, this would mean that lines radiate out from NYC, Boston, DC, Philly, Chicago, Atlanta, LA, SF, Miami, Seattle, Portland... but not any further than an hour or two of travel. The Northeast Corridor is a lucky exception because you have some of the most interconnected cities in the US all in a convenient line.

What's most frustrating is that many of both the opponents and supporters of HSR in the US miss the point: the point is to both invest in and subsidize infrastructure and programs that are societally useful and unlock productivity and opportunities. The Cato Institute would prefer a world without subsidies, but that's not appropriate for the sorts of high-cost functions that offer a major benefit to society. Pedestrian Observations would prefer more mobility and transit, but sometimes that transit actually looks like an airplane or bus or subsidized taxis, because it's what makes most immediate use out of the current infrastructure in a way that balances opex with capex.

[PaulDavisThe1st]

> so High Speed Rail would be a slower and costlier substitute for long-distance flights, and a maybe-faster but drastically less flexible alternative to <300 mile travel.

Lots of people in the US fly distances in the 300-800km range cited in the article. So equating long-distance and flight doesn't seem right. Trains are no less flexible than flights (or busses) over any distance (obviously they can be slower); they only lose flexibility when compared with cars.

> One example: in truth intercity buses are very widespread in Europe, and not only do they fill in gaps left by the rail system, but thanks to expressways on some routes they be as fast as "moderate-speed trains" too.

The article's author says more or less precisely the same thing as you've said here. In addition, the flexibility of busses is not the same as the flexibility of a bus. Yes, bus services can range from inter-city routes moving at 80mph to local ones servicing small villages. But these are never the same busses.

[niftich]

In my comparison of HSR vs. cars for <300 mile travel, the rail being 'drastically less flexible' means that it's subject to the same last-mile problem as planes are, whereas cars do not have this problem. Therefore the advantages of cars for trips like this are difficult for HSR to overcome.

As for buses, the article's author diminishes the significance of intercity buses in Europe by making it sound like private intra-national intercity bus service isn't competitive with HSR on travel time, as if HSR were widespread. HSR is only present along a dozen or so corridors in Europe, and while within France such premium bus services are a relatively new phenomenon, that isn't true elsewhere on the continent; so across the whole of Europe intercity buses are both much more common than he initially suggests, and much more competitive vs. rail than he suggests. After this, he does say buses thrive in the gaps between the train network, complement it, and have historically been important for international travel because of rail fare structures, and on those points I agree.

My wording of 'buses are flexible' does refer to the ease of introducing new routes (i.e. not having to build lots of rail), as a sibling comment correctly identified.

[PaulDavisThe1st]

>In my comparison of HSR vs. cars for <300 mile travel, the rail being 'drastically less flexible' means that it's subject to the same last-mile problem as planes are, whereas cars do not have this problem. Therefore the advantages of cars for trips like this are difficult for HSR to overcome.

For personal travel, I think I'd generally agree. However, lots of US business travellers (pre-COVID anyway) would fly distances of 100,200,300 miles (the shuttles from Phila to NYC were almost always full, and that's just 90 miles!). These journeys have the no-car-last-mile problem, but that doesn't seem to have stopped the wide use of flight for those journeys.

Granted, post-COVID, it's no longer clear how many of these short-haul journeys business travellers will be making in the next 2-10 years.

[niftich]

Business travelers take taxis from their arrival airport to their destination and then get reimbursed by their company later.

They are among the least price-sensitive travelers and are the ones least inconvenienced by the last-mile problem, so their decision-making differs from those traveling for other reasons. (On average, they are less constrained by price and switching of modes, but are more constrained by idiosyncratic company procedures around travel.)

[etrabroline]

>they only lose flexibility when compared with cars

Planes fly in more or less straight lines from source to destination, and adding a new route is just some paperwork and renting the terminal space. No multi-billion dollar investment to connect a new nearby or distant city.

>flexibility of busses is not the same as the flexibility of a bus

Again, he means the flexibility of adding new routes, not having the bus pick you up at your hotel.

[wahern]

> Planes fly in more or less straight lines from source to destination, and adding a new route is just some paperwork and renting the terminal space. No multi-billion dollar investment to connect a new nearby or distant city.

One of the justifications for California HSR was that existing airport slots are almost saturated. Studies showed that airport expansion was going to cost at least as much as HSR. Runway and terminal expansions actually do cost billions of dollars.

And while HSR costs have ballooned, so too would airport expansion costs. Perhaps even more so because much of the HSR cost increases are related to intransigent farm owners in rural areas, whereas the majority of airport capacity expansion work (at least on a cost basis) involves much more developed areas, where NIMBYs tend to be at least as ornery.

The studies also projected greater cost burdens with highway expansion alternatives. Nobody disputes that building additional lanes on I-5 in the Central Valley is cheaper than building HSR in those areas. But that's beside the point because the dilemma isn't about increasing throughway capacity in the Central Valley, but expanding capacity into and out of the SF and LA metro regions, where highway expansions are insanely expensive. Trains (and planes) let you offload people at various points within the metro region, so there's less of a highway bottleneck[2].

Maybe those studies were biased. Certainly many critics believed so. But the relative costs are much closer than people tend to believe.

[2] Especially considering the geography.

[etrabroline]

https://en.wikipedia.org/wiki/California_High-Speed_Rail#Con...

>In March 2018, the Authority revised its estimate to $77.3 billion and up to $98.1 billion, pushing initial service to 2029 and services from Los Angeles to San Francisco to 2033

The cost of Ca HSR has gone up to $98 billion. There is no way building even 6 airports could possibly cost that much, and you can build the airports when you need the capacity one at a time and airports support travel to anywhere in the world, not just the handful of other cities in California the rails connect to.

https://www.therichest.com/luxury-architecture/10-most-expen...

> #1: Kansai International Airport - $20 billion

So California could build 5 huge earthquake proof airports in the middle of the ocean including sea wall and bridge for the cost of HSR. It just makes no sense. I'm not opposed to rail or high-speed rail. But the places where better public transportation are needed are mostly in the US North East where density is much higher. Existing rail infrastructure is crumbling in the North East while California asks for Federal money for white elephant boondoggles like Ca HRS. It's infuriating.

[wahern]

Much of the cost increases you see in figures like those quoted for CAHSR are related to timelines.

First, projected infrastructure project costs in the United States are usually quoted (by agencies) in year of expenditure dollars, which means projected expenditures for each particular year are literally inflated. If you push a project back 15 years, you can easily see double-digit percentage increases in quoted total expenditures. I don't know the exact figure but AFAIU this is very much the case with CAHSR--a large fraction, and perhaps even a majority, of nominal cost increases is related to inflation adjustments as the timeline keeps getting pushed back and stretched. (Of course, that still leaves a large component that isn't.)

Secondly, pushing back timelines incurs all sorts of additional real costs. For example, you have to extend supply and labor contracts, which often means you end up getting less for your money because more people and equipment will end up sitting around idle for longer. You can also incur greater financing costs.

Also, I don't think any of the figures in that therichest.com link are inflation adjusted. AFAICT, they seem to just be tallying the year-of-expenditure or perhaps year-of-completion costs. This difference compounds inflation discrepancies even more. (There are other problems w/ those comparisons, too, such as that most of those airports were built far away from developed areas, but it's not worth going down that rabbit hole.)

Time is the the most critical component in all large construction projects, public or private. Time is money isn't just a catchphrase. If there are any hand-wavy, magical solutions to the cost problem that can be easily applied, one of the simplest and most obvious is to finish projects as quickly as possible. Don't let them linger and get stretched out.

Unfortunately, there are often political and regulatory pressures that push things in the wrong direction. If political blowback causes timelines to be stretched out, directly or indirectly, cost increases become a self-fulfilling prophecy. After learning about the accounting methods for how projected costs are quoted, and in particular digging into CAHSR costs, I vowed as a voting citizen to never oppose a project once it got off the ground, even if I initially opposed it.

[PaulDavisThe1st]

I think given the variation in costs for HSR worldwide, we need to consider the first point of discussion for CA (or US) HSR to be "why are the costs so much higher here?"

It's true that the costs are crazy high, but that's not true worldwide. That means that objections to CA/US HSR are initially (at least) really objections to the ridiculous costs, not the actual idea itself.

[PaulDavisThe1st]

> Adding a new route is just some paperwork and renting the terminal space. No multi-billion dollar investment to connect a new nearby or distant city.

Adding a new route, sure. Adding a new destination, not so much. Are airports cheaper than train stations? I don't know.

[etrabroline]

Train stations are cheaper than airports. California High-Speed rail line plus a few stations is most certainly not.

[bluGill]

> infrastructure cost of long-distance High Speed Rail in the US

Is 4 to 7 times higher than anywhere else in the world. We need to bring it down to reasonable levels.

>commuting patterns in the country are too car-oriented

HSR isn't for commuting. It is for trips longer than a normal commute. The US is car-oriented - but I disagree with the word "too". HSR won't do as well (at least for the first dozen years) as other countries, but there are a lot of places where it should do well enough if we ever build it for a reasonable cost.

> intercity buses are very widespread in Europe

They also are in the US.

[DrBazza]

HSR is for commuting in additional to longer travel. I say this as someone that uses the HS1 line in the UK to commute to London at 140mph.

The key is the commute is under an hour, one stop, and it’s into London which a massive source of jobs in South east England.

I agree with your comment that HSR doesn’t do well for a decade. My train runs on the Eurostar line. That took a decade to draw people from planes. And the domestic train that I catch is now standing room only into London.