I am proud of my first piece published in Bloomberg. I didn't pick the title. It is costly to build high speed rail (HSR) but these costs decline if we use technology built in China, Japan, Korea and France. The rest of the world has plenty of practice in building these trains. While I do not support a San Francisco to LA train, I do support trying smaller experiments in HSR connecting major productivity hubs such as San Jose to smaller cities such as Fresno or Merced. In the Northeast, Baltimore's growth would be accelerated by having access to a real fast train to travel to Philly and NYC. Such trains would allow middle class people to live in one area and work in another area. This would solve the co-location problem for many married couples and would unbundle the current challenge that people now must live and work in the same metro area.
Of course I understand that it would be cheaper if local zoning laws were repealed so that higher density housing would be built in productive city. If such cities made this policy change and introduced road pricing then HSR would not be a good project. Despite the good news in Manhattan recently --- I don't believe these policy changes are on the horizon. Given these realities, I think it is important to run some experiments to see how the introduction of high speed trains for short rides (20 to 100 miles) affects urban growth in satellite cities close to major productivity hubs. I do think that most of the funds to pay for these should be at the state level. If Hsieh and Moretti are right about the aggregate productivity benefits of encouraging agglomeration, then the federal government will gain more tax revenue from such investments and this suggests that federal $ could be used to partially subsidize the costs.
The U.S is not China but China's HSR network highlights how trade between cities in terms of tourism and commuting and interaction is facilitated by HSR. Uber and other ride sharing programs will prosper due to the last mile problem when people are dropped off at HSR stations. In China, the HSR feeds into the downtowns of cities. As crime falls in U.S cities, downtown locations are more desirable and HSR would further enhance this effect.
We need more speed both within cities and across nearby cities. China's case shows how the development of nearby 2nd tier cities makes the overall system of cities stronger.
I do realize that if a real fast train is built from San Jose to a place like Merced that this is taking an irreversible bet on the "field of dreams" effect. Co-ordinated plans would need to be made with real estate developers to build desired housing at the destination. In an existing city such as Baltimore, such risk would be lower. In China, there is the discussion of "ghost cities". In the U.S, there is a discussion of housing affordability.