Hacker News Reader

Comment by londons_explore | Hacker News Reader

> requires an annual subsidy of around €3bn

I suspect the key is to find ways to run railways cheaper. It has been a long time since railways were under severe cost competition encouraging people to look for efficiencies.

I think the main one would be to find a way for railways to operate with no staff. Just like a typical road operates most of the time with no staff.

That means you need to redesign everything that currently uses people to not need people. Sure - there would still be occasional maintenance - but nobody always on duty driving trains (automated), nobody selling tickets (online), nobody cleaning stuff (automated cleaning of trains inside and out) etc.

View on HN

In Japan, rail companies also own the land around the station. There is often a mall or office right on top of the station. So they capture a more of the value of having public transport.

In western countries land lords benefit a lot from nearby public transport for free and maybe only contribute in taxes.

You see the same thing in Hong Kong - the rail company practically creates new towns when they add a new station or extend a rail line.

I was under the impression Hong Kong was one big urban area already, is there really empty land new towns can be built on?

Most of the new territories is not highly developed. Kowloon and hk island are both entirely urban.

Yes. https://en.wikipedia.org/wiki/Hong_Kong#Geography, https://en.wikipedia.org/wiki/New_towns_of_Hong_Kong.

You can be smack in the middle of Central, and then walk up a path behind the Cathedral of the Immaculate Conception and be on Old Peak Road hiking up to Victoria Peak. Or be in Wan Chai, take the stairs next to the Rolls Royce car dealer, and be on Bowen Road Fitness Trail and in Aberdeen Park within minutes.

HK island, New Territories, the outer islands and even large parts of Kowloon are predominantly green, parks and forests.

As GP mentioned, occasionally the subway operator MTR will build a new station, build a shopping mall on top, and housing for 50,000 people.

Originally, rail worked kind of like this in the US too. Alternating squares of land adjacent to tracks were ceded to railroads, so you'd have sections of public land and private land beside each other.

I learned about this when someone was trying to make a test case of "corner crossing" from one patch of public land to another a while back.

A Georgist Land Value Tax would fix this!

Opening the railways to competition has utterly ruined the UK's rail network. Ticket price has skyrocketed while service quality has fallen.

In the 30s, French rail companies begged and lobbied the government to nationalize them, so they could exit the burden that was maintaining a rail network.

Rail just isn't profitable, but is vital to society, and will become even more so as gas becomes increasingly expensive/lacking. Some things should just never be opened to competition.

According to GPT, the price per kilometre for rail travel is between €0.23-0.46 in the UK and €0.13-0.20 in Germany. I'm not able to verify those numbers but from my own personal experience I wouldn't doubt it.

Rail in the UK is just so bad that I stopped using it to go to the airport. Everything is constantly delayed or cancelled like I can plan a journey to arrive 2hrs before my flight and the train will randomly pause for 1hr when I'm half way... also the luggage situation is extremely stressful as the racks are out of sight and generally full, like where else am I meant to put my large suitcase after that?

It seems like trains are much larger in Europe with a lot of double deckers.

If there's two of us it's easier and sometimes cheaper to pay £100 for a taxi to the airport door to door (we're fairly rural)

There is very little true competition on UK railways. Most services are run on Government-granted franchises which give a monopoly or at best a duopoly on specific routes.

There are a handful of open-access train operators who operate outside the franchises, but they can't just decide to compete on a route if that route is already covered by a franchise.

To create true competition, the infrastructure and stations would need to be taken into public ownership and the train operators would run whatever services they see fit.

> To create true competition, the infrastructure and stations would need to be taken into public ownership and the train operators would run whatever services they see fit.

The problem with that is that on a mixed-traffic railway it's very easy to run out of capacity once you have a mixture of trains with differing stopping patterns (and perhaps some freight on top) running along the same line. Once you reach that point, competition stops being directly about passengers and starts being about train paths, which can have rather annoying side effects for passengers.

Plus connections don't really work well with competition, either, because a) that'd require coordination between possibly differing operators, which is anathema to unfettered competition, and b) where a regional feeder route might only run hourly or half-hourly, it's impossible to have good connections with more than one or at most two sets of connecting trains anyway.

(Also c) through-ticketing with sufficient passenger rights in case you miss a connection enroute – at least the UK kept a national ticketing system even throughout privatisation and including all open-access operators, whereas in Germany it's an incompatible free-for-all, despite a legal obligation for railway companies to ostensibly offer through-tickets.)

When installed and kept up properly with good policymaking railways are always a positive for the economy. They cause a huge amount of cheap movement that increases business activity. Little towns that become railway stops develop much faster. The bigger stations are great for shops and food. The access to education and high-value driving extracurricular activities increase for younger people too. They overall make the economy resilient against all sorts of crisis (energy, markets)

The more railways are used for commuting, the less people are on the road. So it increases the road efficiency too and reduces degradation. Railways are great drivers of innovation and the technology they generate can be backported to cars, they are initial investment drivers.

However, the first if/when is a big one. You can half ass roads. You pay compensation for a pothole every now and then and make small improvement to get votes. You cannot half ass railways. They require constant maintenance and a whole mindset built around them.

Japan does railways correctly. China is getting there. The Netherlands is nearly a paradise of bike and railways. Germany isn't. The countries with almost the same culture, Austria and Switzerland, care much more about their railways and invest them properly however their government aligns. Germans keep electing the right wing party with their ministers of ~BMW~ transport and then complain about 50% of the non-cancelled trains are late and the maintenance cost of the falling apart rail system is quadrupled.

As someone who's been to both China and Japan I'd say railway is far ahead in China than the rest of the world.

Again, according to GPT, the price per kilometre between them is €0.03-0.06 in China vs €0.15-0.26 in Japan.

I know Japan has a lot of high speed rail coverage but so does China and the difference in price is absolutely insane.

Rail between Beijing and Shanghai for example has an average speed of 300kmh. I believe their high speed network is vaster than Japan's.

Most of these numbers are probably due to differences in wages and the size of the respective countries.

If they aren't completely fabricated.

That is so the wrong approach. If I’m very reductive, I could say that if we automate everything there would be no one to commute left. Staffing is good for people: for safety, information, help for the vulnerable, expertise, robustness of the system, etc.

Perhaps some public services can’t turn a profit but are still necessary. Perhaps they can turn a profit if even more people use them. Perhaps other economic models are needed for these companies, see for example Singapore and Hong Kong where the rail company owns the land at stations and are allowed to develop there which funds the network and incentivizes buildout.

I doubt that moves the needle a lot unless you're making _everything_ fully automatic, i.e. including the infrastructure creation & maintenance and streamline that extremely well.

If you take a ride from Hamburg to Berlin with the ICE, there's maybe 8 staff on board, 200-300 passengers, it takes ~2 hours and the average ticket price is 77€ according to some travel app.

Even if you get rid of all 8 of them, the price per ticket isn't going to be lowered significantly.

But there is also the ~80 staff manning stations at each end and along the route, and more staff in the control room, the cleaning staff, the people who manually do maintenance checks every single night, etc.

You can get rid of the cleaning staff right now. The terminals will become disgusting and people will stop using them. There is no way around this type of overhead. If you have larger numbers of people passing through, you need cleaning and maintenance.

Your original point isn’t even true - in California there is constant maintenance of the road network and crews of community service workers doing their sentence cleaning trash on the highway shoulder.

Roads still have far fewer staff per mile than railways, however you measure.

In turn, that makes them cheaper.

Is there a source for this? Is this source heavily context-dependant?

Lol, that depends heavily what you consider staff. E.g. look at freight, if you move cargo by trucks you need 1 staff member (sometimes called drivers) per truck. Train on the other hand needs only 2 people per hundreds of meters of train...

Even worse for individual car transport. You need one staff (driver) per vehicle...

Why would there be 8 staff on board? Isn't there typically one train driver and one attendant?

I don't know. They have the restaurant car on some trains so that's another 2-3 people, and I really have no idea about their operation so I added some buffer. I wouldn't expect there to be more than 8 on a train unless it's some special event.

For some high speed routes there’s a limit on how many carriages one attendant can supervise (maybe to ensure timely evacuation?) so a 12 carriage ICE train needs 3 attendants to fully operate.

The staff on board also contribute to keeping the train safe and pleasant - if there’s an unruly passenger, or a washroom malfunctions, I’m happy to have paid a bit extra for my ticket for a couple of people to be on hand to deal with the problem instead of being trapped for two hours.

Yeah, it seems that "self-driving trains" are a much, much more tractable problem than self-driving cars. On the other hand, the cost of the driver is amortized over many passengers, and much of the labor isn't driving but rather serving as conductor, etc, so it may not even matter too much.

> automated cleaning of trains inside and out)

For the outside, you can imagine a carwash.

For the inside, my brain goes to scary dystopian places. Like, "what if we make the inside out of chemically-inert glass-based materials, and clean it by immersion in pirhana solution?" One would just need to recycle the solution, and recharge it with hydrogen peroxide. This would rule out the use of plastics in the interior, however.

Maybe something more like a dishwasher could also work, but I'm not sure it'd be Strong Enough for Tough Stains. It could even just make a mess. I've heard stories of Roombas that encounter dog poop; they say it goes badly...

> That means you need to redesign everything that currently uses people to not need people

Simplest redesign is running longer trains - they still need just 1 driver.

Britain’s trains are sometimes comically small (3 carriages) and overcrowded. French trains are 2-3x longer, Russia/China even more so.

Second is standardisation - all of Uks train companies run different stock and it’s ovsoleye

At some point the train is longer than the station platform. Then you need to extend the platform, or passengers have to know that they need to be in specific cars to be able to exit.

When you rent each carriage from a rolling stock company you’re incentivized to run as short trains as possible to keep them as full as a sardine can. Renationalize the whole lot.

Longer isn’t always better for trains. Longer trains need bigger platforms and stations are often the most expensive parts of the network. Big trains also weigh more, so they accelerate slower and have longer trip times.

Weight savings can also translate into cheaper bridges and viaducts, though that’s only true for dedicated track. If you’re sharing with freight they are always going to be far heavier.

Much better is upgrading signalling, automating the train, and running far more service because frequency is king for public transit. Many short trains at high headways vs less frequent long trains has the same total capacity but the short, frequent trains provide far more value. When people don’t have to think about the schedule and can just show up and ride public transit is great.

There are also limits to the value of standardizing rolling stock. You don’t want every train set to be bespoke, but there is also danger in every train being exactly the same. If there is a parts shortage or identified design flaw your whole fleet can be grounded all at once. If you have some diversity you can limp along if one design has to be sidelined.

Reply on news.ycombinator.com