I don't quite get how you're coming to 2/3 of revenue between MTC and SCAG regions (Its not adding up to that for me). It also isn't very clear: MTC has a projection for ridership to SCAG, but SCAG doesn't have projections for ridership to MTC but they do for the SJV?
I'm also not sure where this newer forecast model came from, but it seems highly subject to politicized biases: they basically asked individual regions to forecast their ridership and revenue to other regions, and then added it all up.
The original forecasts [0] were built off of SNCF forecast modeling expertise, which is one of the best in the world and also from a country with very similar geodemographic characteristics as California. For just the subsegment of LA-SF, less than 50% of revenue is between SF and LA, and the ridership between SF-SJV and LA-SJV adds up to approximately the same amount as LA-SF. In other words, going through the valley doubles revenue potential.
This isn't a new phenomenon. It has been noted by and (if necessary) relearned by railroad capitalists going all the way back to the foundation of railroads. For an example, read chapters 16-20 of this economic history of Railway Mania [1]. Railway demand between termini has almost always been overforecasted by railroad capitalists, and they either figure it out and start serving intermediate destinations or they go bankrupt. If they build the line from the very beginning by bypassing intermediate population centers, their overestimates of ridership will doom them to a fate they can't recover from, barring major track realignment.
A background in Operations Research gives me a professional social network of people across pretty much every industry, all of whom understand some of the hardest business and technical problems of their industry. Some of my contacts work on schedule optimization in Germany and Spain. One of the quotes that constantly gets spouted in their conferences is that HSR is politically successful because of its top speed, but it is commercially successful because of the first derivative of its speed. Top speed gets votes, but the ability to start and stop very quickly allows them to obtain very significant ridership boosts with very little degradation in average speed. This is a very distinct competitive advantage...they can do something that no airline could ever do. Between major termini, there will always be some form of competition with airlines, but intermediate stations will always be a veritable monopoly for rail.
>I don't quite get how you're coming to 2/3 of revenue between MTC and SCAG regions
I said "to either", not "between". I also didn't suggest cutting out intermediate stops, but I did allow them to move outside of the town centers. In other words, most travelers from SJV are going to either MTC or SCAG, not within SJV, and in particular this is not directly analogous to the 19th century because of the extensive road network and the popularity of cars. Similarly, putting the rail corridor 30-60 minutes from the town is probably OK since the majority of residents of those towns will probably be driving to the train anyway, while doing that in a horse-and-buggy economy will of course doom ridership.
From page 5 of your link:
LA-Bay = $735M
LA-SJV = $355
SJV-Bay = $346
...
Within SJV = $29
total = $2355
to [LA or Bay] = 735 + 355 + 346 = $1436 or 62% of total.
So over 90% of ridership from SJV is going to either SCAG or MTC or further. The data seem to comport with my estimate that in order to start recouping costs you need, at the very least, service to one endpoint. This doesn't mean you should remove intermediate stops but it does support the idea that SJV HSR customers are mostly traveling long distances and might not mind leaving town to get to the train. This is particularly true if parking fees can be reduced by moving the train station out of town, since the current projection is $12 which I'm afraid might be "per day" (nearly as bad as the airport!) and people hate that.
The report I linked notes that the projection is for cars to continue accounting for 93% of intra-regional travel demand: i.e. while "the world is local" as your link states I have a hard time believing that market insights from a world without cars will translate readily to one with cars. Additionally, while some New Urbanists have started to try to build car-free cities, this goes against the wishes of the general public (including me) and I doubt it'll be popular in the politically conservative Central Valley. In this light, driving the train through the middle of Bakersfield as opposed to the outskirts looks like an attempt by the state to implement a vision of Bakersfield's future that the residents of Bakersfield did not really ask for. The majority of Central Valley customers will be driving, rather than walking/bussing, to the train even under the current plans. I'd also like to point out that, while there is an environmental motive for reducing our dependence on gasoline via cars, I personally expect electric cars to arrive sooner than a restructuring of the infrastructure of Central Valley cities that deprioritizes cars in favor of public transit. It could happen, but I'm not betting on it. So this:
>intermediate stations will always be a veritable monopoly for rail.
doesn't seem as clear to me as it does to you. And while as you mention the projections may have underestimated ridership in the (less-infrastructure-keen) non-city regions, they also assume a 50% real increase in the cost of auto travel, which as noted may be softened or reversed by the adoption of battery-powered vehicles. Therefore, Dionysus Lardner's observation that the median trip in 1846 was about 15 miles long probably would not apply in 2025.
As you've probably inferred, most of my background in this area is political and personal rather than technical. So, for example, I don't really know if moving the train line out of Tulare actually results in a significant savings on land and/or a reduction in construction time (especially this!), without which my whole point is moot. But I do not think that Fresno is analogous to a city in Germany, nor do I have a desire to turn it into one. Americans like cars, despite the powerful people with professional social networks who want to control us.'
edit: agh this is a bit of a mess, I hope it isn't too confrontational, and I appreciate you sharing your expertise.
>I did allow them to move outside of the town centers.
Which is the case with a number of the shin- stations on the shinkansen lines. (Although, being Japan, those stations are still well-connected to the city transit systems and they're not that far out of the traditional town centers.)
I would agree that having two major city termini is a huge determinant of demand. I definitely wouldn't suggest starting with one city, although branch lines can definitely be built at lower risk levels once you have a strong backbone.
I've seen overviews of the LGV models to know that they aren't making decisions on where to stop based purely on politics (I'm sure there are some politics involved, but the schedule optimization models are very objective). Those optimization models take into account things like cannibalization, where most riders from two close small towns can be served by one station in between them, only losing a small percent over a situation with two stations. But mostly the model is an LP of the total revenue lost from removing a station vs the total revenue gained from adding a station, and the revenue projections obtained from a forecast model that predicts the cartesian set of all possible orig-dest pairs of a line. And these models have resulted in stops in tiny towns like Loche (pop 35k), Macon (pop 35k), and Le Creusot (pop 22k).
> It could happen, but I'm not betting on it. So this: [...] doesn't seem as clear to me as it does to you.
In France, the outlying towns are not much different from California. They may not be as car-centric as places like Madera or Merced, but they are at least as car-dependent. Density is low enough that cars are still a necessity. And while some of these stations avoid going directly through the towns they serve, very few are more than a 3-5 km from the town center.
From the perspective of someone who grew up in Stockton, I'd say that concerns about car dependency are still overblown outside of intra-SJV trips in general. If the destination is a line terminus, transit/taxis are abundant enough to use easily, and that leaves the trip to the station as the only deterrent. But getting a ride isn't that big of a deal. Living there, you almost always have access to someone that can give you a ride to the airport 45 minutes away if you need it...a trip across town should be pretty easy. Taxis are available on a scheduled ride basis, and most businesses with regular travelers maintain a list of van services that handle trips to and from airports. The concern is real, but I think most valley-dwellers have adapted to their situation quite well.
I'm not an expert on property values in the valley (haven't lived there for 2 decades), but I can imagine the possibility of it being cheaper to build through the cities simply because they already have freight ROWs running through them. Getting a ROW easement on an existing un-partitioned property tends to be fraught with peril. It is easy for a large land owner to hold a line hostage for higher prices if they know that one "no" ruins an entire plan. Buying up existing freight ROW (which I believe is part of the plan) would definitely be easier and faster...it might also end up being cheaper.
And no worries about being confrontational. I wasn't trying to, and I didn't interpret your response as confrontational either.
I'm also not sure where this newer forecast model came from, but it seems highly subject to politicized biases: they basically asked individual regions to forecast their ridership and revenue to other regions, and then added it all up.
The original forecasts [0] were built off of SNCF forecast modeling expertise, which is one of the best in the world and also from a country with very similar geodemographic characteristics as California. For just the subsegment of LA-SF, less than 50% of revenue is between SF and LA, and the ridership between SF-SJV and LA-SJV adds up to approximately the same amount as LA-SF. In other words, going through the valley doubles revenue potential.
This isn't a new phenomenon. It has been noted by and (if necessary) relearned by railroad capitalists going all the way back to the foundation of railroads. For an example, read chapters 16-20 of this economic history of Railway Mania [1]. Railway demand between termini has almost always been overforecasted by railroad capitalists, and they either figure it out and start serving intermediate destinations or they go bankrupt. If they build the line from the very beginning by bypassing intermediate population centers, their overestimates of ridership will doom them to a fate they can't recover from, barring major track realignment.
A background in Operations Research gives me a professional social network of people across pretty much every industry, all of whom understand some of the hardest business and technical problems of their industry. Some of my contacts work on schedule optimization in Germany and Spain. One of the quotes that constantly gets spouted in their conferences is that HSR is politically successful because of its top speed, but it is commercially successful because of the first derivative of its speed. Top speed gets votes, but the ability to start and stop very quickly allows them to obtain very significant ridership boosts with very little degradation in average speed. This is a very distinct competitive advantage...they can do something that no airline could ever do. Between major termini, there will always be some form of competition with airlines, but intermediate stations will always be a veritable monopoly for rail.
[0] http://www.hsr.ca.gov/docs/about/ridership/ridership_revenue...
[1] http://www.dtc.umn.edu/~odlyzko/doc/hallucinations.pdf