High-Speed Rail

On Political Will and High-Speed Rail

I wrote yesterday on U.S. Secretary of Transportation Ray LaHood’s trip to Europe to study high-speed rail there, and asserted that Spain’s mission to put ninety percent of the country’s citizens within 50 km of a rail station was a good model for the United States to follow. One commenter argued that the effort was ultimately going to pull the Spanish government far into debt because of the cost inefficiencies of building lines to relatively small cities; the result would be proof to conservatives that high-speed rail isn’t an acceptable use of government funds and therefore shouldn’t be pursued by Washington. I’d like to address those issues here.

The Tokyo-Osaka Shinkansen and the Paris-Lyon TGV lines are said to be the only high-speed corridors in the world that have paid off their initial construction debts. They were also the first of their kind and have had decades to make profit on services along their respective corridors, so there is reason to believe that other highly frequented routes, such as Lyon-Marseille, Madrid-Barcelona, and Paris-Brussels, will be equally profitable over time.

But what about those lesser-performing lines? Spain’s investment in new corridors between small cities — such as the 90 km line planned between Ourense and Santiago de Compostela, both with only about 100,000 inhabitants — raises questions about whether the government should be prioritizing projects that affect so few people.

Undoubtedly, it’s true that we need a standardized system in the United States to ensure that the most qualified projects are those that are funded. But I disagree with the idea that the ambitions of a country such as Spain are unrealistic; its goals for rail transport are both universal in reach and achievable in scope. With a population twice that of Texas on a land area 2/3 the size, Spain is small enough and sufficiently dense to mean that connecting every part of the country to the fast rail network is a reasonable goal. That may not be true in the United States, with the huge emptiness of the Great Plains and Rocky Mountain regions constituting an unmanageably large gap, but strong intercity connections on the Eastern and Western seaboards are certainly worth envisioning.

Indeed, the Spanish mission to ensure strong train connections to every part of the country differs very little from that of the U.S. Interstate Highway program, whose links now extend to virtually every metropolitan area — no matter how minor. My simple point is that Americans deserve excellent rail connections just as they have come to expect the same from our road network.

Perhaps, then, one might argue that new rail connections are acceptable just about everywhere, but that high-speed rail is overkill for Springfield, Missouri or Boise, Idaho, for example. I think that this is largely a disagreement about definition. The federal government now defines HSR-Regional as train service operating at speeds of over 110 mph; I think that it would be acceptable for the federal government to guarantee such service to every metropolitan area of more than 100,000 inhabitants. Perhaps a similar goal would guarantee HSR-Express service, at more than 150 mph, to every metropolitan area of more than 500,000 inhabitants.

It is true that many of these lines would take decades, if not centuries, to break even on their initial construction costs. Yet, in terms of gas tax revenues, will the segment of Interstate I-90 constructed across South Dakota ever do so, either? Is it reasonable to affix profit-oriented aims to national infrastructure policy?

Ultimately, national rail construction policy should serve two primary goals: one, to improve mobility between the nation’s most populous urban centers within 600 miles of one another; and two, to connect the entire nation through a more sustainable mode of travel than currently offfered by either automobiles or airplanes. To fulfill the first objective, we must we must invest in expensive new lines for very fast trains in significant corridors like California or the Northeast. These lines are likely to be profitable, and, if managed correctly, would eventually make up their construction costs. To fulfill the second, we have a national incentive to take a loss on infrastructure building to ensure rail traffic to smaller metropolitan areas throughout the country.

The government has finite resources, and it cannot commit to a rail line to every municipality. But to dismiss the goal of a truly national rail network as unprofitable and therefore not acceptable to American conservatives would be succumbing to failure before the game’s even begun. Rather, we need to start with a national goal and move on from there, working in phases to accomplish universal coverage just as Spain has. It may take fifty years to finish, but long-term planning is the best kind of planning.

31 replies on “On Political Will and High-Speed Rail”

A fine article, Yonah; one hopes that it will be widely distributed and read.
Keep up the good work.

This post asks the excellent (and interesting) question:
“Is it reasonable to affix profit-oriented aims to national infrastructure policy?”
Locke’s social contract invoked by the Consitution’s Preamble actually authorizes us to say “No, profit is not a necessary criterion for government investment.”
But, if profit must remain a criterion, the full benefits of rail–beyond ridership fare dollars–will exceed the costs.
Just as transit fares don’t recoup operating costs, but the broader benefits–which are indeed measurable and have been so measured–far exceed the gap.

To continue on what TAS said, that’s something I’ve often thought about: you will often hear conservatives go on about how there is no money to be made in HSR, or other infrastructure investments, or universal health coverage. To which I’d reply: yes. That is the point. Profitability should not be the goal of all our endeavors. The profit motive is great for making consumer goods, it works well in real estate, finance, and the vast majority of the economy. But some things – infrastructure, health care, national defense, education – are public goods. If we can find a way to make money off them, great. But the primary goal of policy in those areas should be to ensure that the greatest possible quality of that good reaches the highest possible proportion of the population. Of course, there is a balance – the government should not break the bank to run 220-mph HSR through Montana. But if you build a line, say, from New York to Chicago via Pittsburgh and Cleveland, and it loses money, but serves hundreds of thousands of people, then you’ve achieved your goal.

Remember that 80% of the country lives in urbanized areas.

Also it’s entirely reasonable that a high speed train run from Chicago to Springfield MO. It could run at high speeds to St. Louis and then at 110 mph from St.Louis to Springfield. Likewise a “high speed train” could run at high speeds from Seattle to Portland, then continue it’s run at 79 mph to Boise.

But that’s way down the line. At this point it’s just a nice discussion.

The Interstate program is exactly what national HSR shouldn’t look like. It was a boondoggle, going over budget by a factor of four and over schedule by a factor of three. It was so overbuilt that the feds have to raid the general fund just for maintaining existing highways; when they can’t do that, bridges collapse and kill people. Outside the imaginations of Randall O’Toole and Wendell Cox, it’s a huge money drain even in normal times, all so that over-subsidized ranchers in South Dakota get easy drives to Chicago.

A better model for HSR building is each country’s first HSR line. The Tokaido Shinkansen and LGV Sud-Est were both built to relieve congested regular-speed routes, over the objections of those who wanted to four-track the existing line instead. This requires some ROI formula, which Boston-NY-DC and LA-SF should pass, but Boston-Martha’s Vineyard, New York-Montauk, and Philly-Scranton, should not. All of the lines in the second category lie entirely within states that are denser than Spain, but none will attract much ridership and none will ever pay off its construction costs.

You seem to set up a straw man argument in your list of HSR lines that should not be built. Straw man because nobody is talking about a Boston-Martha’s Vineyard or Philly-Scranton high speed line (although that Philly Scranton line might actually be worthwhile considering the huge ridership levels on the near HSR line from Harrisburg to Philly).

I’m not sure you’re arguing the merits of a position at this point or simply throwing bombs. It doesn’t come across as helpful though and I think whatever point you’re trying to make is getting lost amongst some of the more bombastic statements.

Adam: of course nobody’s talking about those lines – they’re self-evidently ridiculous, to an American. But to a Spaniard, they’re not. Many cities with a planned HSR connection in Spain are no bigger than Scranton or the East End of Long Island, and lie at the end of the long lines in planning or under construction instead of between two bigger cities. At the most extreme, the AVE is planning a dedicated 90-km spur line serving just the city of Soria, whose metro area has 42,000 people. Another spur of about 170 km, under construction, is to serve Lugo, metro population 99,000, connecting it to A Coruña (population 376,000) and Ourense (population 137,000).

High speed rail pays itself off in MANY other ways as well.
All the new development that occurs along the routes and at stops eventually raise tax revenues, so there’s another plus. THose things aren’t always taken into account. I do agree for the UNited States that, at least for now, we shouldn’t go HSR crazy. We need to focus on the BIG projects, and perfect that, and see how it goes.
The Interstate System is just as expensive in the long run as a HSR system is. Once the initial costs of high-speed rail are in, it’s much easier. I don’t think that you need to refill the potholes on a rail system every spring, do you? hahaha

Travel demand between two cities is not just a function of the size of the cities. It’s also about the distance between them. The ‘gravity model’ approximation of travel demand, for example, suggests that the demand between two cities is proportional to their size but also inverse-squarely proportional to the distance between them. Any federal policy would need to include this factor to explain why isolated cities in the Mountain West and Plains states can’t expect what a city of the same size would get if it were near a lot of other cities.

US Trans Sec LaHood posted a parallel argument on his blog for today on the benefits of public transit. Substantial economic, social, and environmental benefits accrue from transit investment. Sure, fares don’t cover the costs, but it’s not like public education pays for itself either, yet you won’t hear from credible sources an argument against that govt enterprise because it delivers something the nation has decided it values. HSR done well can also enact what the nation has decided it values.

Jarrett: HSR demand is highest at medium distance, rather than low distance, when cars’ point-to-point nature more than compensates for their low speed. HSR demand seems to peak at a travel time of 2-3 hours. When you get far below that, for example in the LA-San Diego market, HSR will not see much ridership.

Alon, demand may be highest at medium distances, but Spain apparently doesn’t know this. Their goal is to have HSR stations within 30km of 90% of the population by 2020. If we’re looking at HSR from SanFran to LA, extending the line to SanDiego could make sense. Then, the question becomes will San Diegans ride ti to LA? If we’re talking downtown-to-downtown, even at the SD-LA distance, HSR will beat the car in door-todoor travel time.

Alon … Agreed. My point is merely that rational policy on what cities are entitled to rail must take into account not just a city’s population, as Yonah proposes, but also its distance from other cities. A city in California is likely to justify HSR sooner than a city of the same size in Montana on this principle.

Any exclusively population-based standard such as Yonah suggests would, if taken literally, lead to wildly disproprtionate spending to cover huge distances to reach the hardest-to-reach cities that met the population standard. The structure of the Senate already steers enough transportation funding into rural states.

Yes, Eisenhower spent a fortune grade-separating every farming road in South Dakota so that you could drive from Seattle to Miami without hitting cross traffic (it was all about defense, remember?) Surely we can make better choices this time.

90% of the population by 2020 is admirable and ambitious, but entirely unrealistic.

Rather than aiming too high, it’s better to start small. A few major corridors of ideal distance could cover a large number of the largest metropolitan areas in the country. Boston, New York, and Washington: 450 miles, and includes the Baltimore and Philadelphia areas. Seattle and Portland are both desperately seeking high speed upgrades to its existing Cascades line (at least between those cities), which covers Vancouver, BC to Eugene, OR: 428 miles. Detroit, Chicago, and St. Louis seems like an ideal candidate, too: 581 miles. California HSR between the Bay Area and Southern California will be a relief to many: 382 miles.

And just like that, you’ve given HSR access to over 60 million people, and covered a fair portion of the largest metro areas in the country. Even with massive influxes of funding, that’s a stretch for 2020. 2030 definitely, possibly 2025.

But once the foundation is laid, only then can the system begin to mature. The Cal HSR system can make the connection north, linking the entire west coast from Canada to Mexico. Las Vegas and Phoenix could also be built into the network. The northeast corridor can extend west at multiple points: New York City to Albany and Buffalo (and Toronto); Philadelphia to Pittsburgh and perhaps Cleveland.

This is building a real rail network, but there is a modern catch: plenty of places want HSR, but only a select few actually require it. There will be much political noise about towns allowing rails through their town without a station. This argument, while valid, is selfish and inefficient. (I realize that this is not exactly the most kosher of political stances on the issue, and is not fully representative of my views, but I digress.) A 747 going from New York City to LAX does not stop in Amarillo, Texas, simply because it’s on the way. (With apologies to the good people of Amarillo.)

The conclusion: keep it real. Aim high, start low. 30% by 2025, 60% by 2040: Add 10% every 5 years.

Going back to the LA – San Diego example . . .

Only a fraction of the market is actually going between LA and San Diego. So if you run a non-stop it will compete well in that market, yet still get low numbers. People from San Diego go to Santa Anna, to Riverside, to Santa Barbara, San Bernadino, Ventura, etc.

The beauty of a train is you can capture all of those markets. The San Diego trains can (as they do now) go to LA and then on to Oxnard, Ventura, Santa Barbara and San Louis Obisbo. Others can go San Diego – Las Vegas with stops at Riverside and San Bernadino. Run them at high speed out of San Diego, then put them on BNSF to make the cut-off over toward Victorville, back to high speed again to Las Vegas. Others will got to LA at high speed then onwards to Santa Barbara, etc, at 79 mph. You still get a very competitive Ventura – San Diego trip time.

According to the estimates of the California HSR authority, only a fraction of the ridership will be between Greater San Diego and Greater LA. The bulk of the ridership will be between the LA Basin and the Bay Area, followed by LA Basin-Sacramento, and then intermediate markets like LA-Fresno, San Diego-LA, San Diego-Bakersfield, etc. Although right now Amtrak lives on intermediate markets, especially outside the Northeast Corridor, HSR tends to be dominated by end to end riders; most Shinkansen and TGV trains run express from one major city to another, with only a few locals serving the smaller stations.

Because HSR is electrified, it’s impossible to run high-speed trains on low-speed lines unless they’re electrified as well. Unless the FRA reforms its safety standards, it’ll be impossible to run on low-speed lines, period. This makes TGV-style solutions, like San Diego-Cajon Pass-Las Vegas, impractical. For service to Las Vegas, the current plan is to connect via Palmdale, which is geologically simpler than going through Cajon Pass. It will screw over the Inland Empire and San Diego, but benefit the Central Valley and Bay Area.

As for which lines to start from, the formula should include not only population and distance, but also ROW availability. The Northeast Corridor offers a fairly straight ROW between New York and Washington and between Providence and Boston, making it far cheaper per mile to rebuild it to HSR standards than is usual. Other lines with good existing ROW include Chicago-St. Louis and Chicago-Cleveland-Buffalo-Albany-NY. On the other hand, in Pennsylvania, Texas, California, the Upper Midwest, and the Southeast there is no way around brand new track mostly along new ROW, increasing costs, and in the cases of Pennsylvania and California the new ROW will involve many tunnels. With the exception of California and maybe Texas, those lines would be better funded out of the profits of lower-hanging HSR fruit.

Alon Levy said: “Spain is a bad example to learn from. Most of those HSR lines have no economic justification (…) It’s better to focus on HSR connecting medium and large cities, rather than ensure every small town has an HSR connecting it to the capital.”

I wonder what is a small town for you. If you connect Madrid (pop. 6 million) with Seville (pop. 1,5 million), it is logical to build a couple of stations along the way. As there are no big cities in between, smallish Puertollano (pop. 50,000) and Cordoba (pop. 300,000) got connected; but the line was not built for Puertollano and Cordoba, they just happened to be along the way.

The same happens on almost any other line. While there are plans to connect a couple of small towns laying out of economically justified lines, those lines just won’t be built. In fact, as a Spaniard, I’d take the plans to build 20,000km by 2020 with a grain of salt. It will probably end up being 10,000km by 2025 (which will be the largest in the world, anyway, as the Chinese have no spending power to buy HSR tickets; existing HSR in China have been utter failures).

By the way, I’d bet the US has not a single operational HSR line by 2025.

(I wrote this comment on the previous post, but I thought it was convenient to re-post it here. I’d add Soria is the perfect example of a line which will never be built; especially since they already have a fairly good rail connection).

Sorry, my numbers were wrong. I should have said: “I’d take the plans to build 10,000km by 2020 with a grain of salt. It will probably end up being 7,000km by 2025 (which will be the largest in the world, anyway…”

Regarding Orense’s connection to Lugo, plans are exactly two weeks old, after an “AVE to Orense” populist platform allowed the new regional leader to win a key recent election. Previous plans considered upgrading a single-track line for speeds up to 160km/h, which is not true high speed by any measure. The fact that the new Transport Minister, which has never visited a university nor its surroundings and can’t even speak proper Spanish, came from Lugo, may have played a role in that complete mistake.

Anyway, it won’t be built. Take my word for it. Despite all our shortcomings, reason usually prevails in the end in this country. And there are no hard plans as of today on a HSR connection from Lugo to A Coruña.

Yonah said: “Spain’s investment in new corridors between small cities — such as the 90 km line planned between Ourense and Santiago de Compostela, both with only about 100,000 inhabitants — raises questions about whether the government should be prioritizing projects that affect so few people.”

I think this deserves some explanation. The aim is to connect the two biggest metro areas in the northwestern region of Galicia, namely Vigo (pop: 670,000) and A Coruña (pop: 410,000) with Madrid (pop: 6 million) and the Portuguese metro areas of Porto (pop: 1,7 million) and Lisbon (pop: 2,8 million).

Both Orense and Santiago de Compostela are the most important towns located just on the “natural” track to connect those metro areas. Santiago is Galicia’s capital city, has its own international airport and is the third most important pilgrimage destination for Catholics, after the Vatican and Jerusalem. Orense just happens to be the only town in a mostly uninhabited area. Most fast trains won’t even stop here, so having an AVE station in Orense is basically for free, and can provide a huge stimulus to the local economy.

So Spain is not really investing in corridors between small towns, but between big and medium-sized metro areas. Some small towns just happen to be located on the right, lucky track (pun intended!).

Even Vigo and A Coruña are marginal propositions at best. Akita, with a metro population of 450,000, only got a medium-speed Mini-Shinkansen Line. Larger cities in northern Japan, like Niigata and Sendai, with metro populations of 1.3 and 2.1 million respectively got full Shinkansen lines to Tokyo, but those lines were not very successful, not nearly so much as the Tokaido and Sanyo Shinkansen in the Tokyo-Osaka-Fukuoka megalopolis.

Alon Levy,

Spain is not looking up to Japan for an example. We haven’t even gotten started and we already have nearly three times more km in high-speed track than Japan on a per-capita basis.

I don’t really know Japan, but the pattern of Spanish movements is surely different. Akita has no manufacturing basis, while Vigo and La Coruña manufacture cars, windmills, ships and design electronic devices and clothes. Moreover, tourism is far more developed in Spain than in Japan, and the pattern of family visits is different, too.

Especially innovation-intensive activities, which in Japan would typically happen in the 3-4 main cities, need a deep labour pool that only occasional visits from professionals living away can make up for. That’s why tiny Orense and smallish Albacete have set up regional science parks near their AVE stations.

In other words: if you can have your own Tokyo (or Silicon Valley), try to turn the whole country into Tokyo (or Silicon Valley) through high-speed rail.

In other words: if you can’t have your own Tokyo (or Silicon Valley), try to turn the whole country into Tokyo (or Silicon Valley) through high-speed rail.

Spain may have more HSR track length than Japan per capita, but it only has about one quarter the per capita ridership. Even South Korea, whose HSR system consists of one incomplete line opened in 2005, has more ridership than Spain. So looking up to other countries’ examples is still instructive. Some of its lines were expected to have the same dynamics as Spain expects. Niigata in particular is a center for ski resorts; however, local businesses have complained that instead of promoting tourism, the Shinkansen made it possible to take day trips from Tokyo, so that nowadays tourists don’t stay overnight.

The idea of turning the entire country into Tokyo (or Paris, or Madrid, or Seoul) has failed elsewhere. Despite the Shinkansen, all parts of Japan except Greater Tokyo and Greater Nagoya are losing population, due to both low birth rates and migration to Tokyo. And the TGV has done nothing to revitalize the economy of Marseille, which remains poor and less developed than the TGV-less Riviera cities to its east.

Alon Levy,

ridership in Spain is still lower than in Korea and Japan, but that’s because it is brand new and end stations are not connected yet. E.g. the line to Barcelona is less than 18 months old, and Seoul is more populated than Madrid and Barcelona combined.

However, the economic effects are already being felt, especially in middle-sized cities: much more business tourism, innovation-related industries are being attracted (e.g. IT companies in Ciudad Real and Lleida, etc.) and people are starting to move more around. Japan and Korea are highly centralised countries with extremely low entrepreneurship and an underveloped tourism industry, but Spain is reaping the gains from this higher mobility fast.

On the population front, Spain’s has grown from 39 million 12 years ago to 47 million people now (immigration accounting for almost all this growth). However, population has decreased in those areas with bad rail access and has increased the most in new smallish AVE cities. So, just as you can’t compare the effects of building airports and highways in a long-run declining, depopulating economy (Japan), you probably can’t compare high-speed rail effects in a long-run economically and population-wise booming country (Spain).

Japan’s decline is a fairly recent phenomenon. Before 1990, it was growing very quickly, complete with a housing bubble of the same kind that caught Spain in this decade. The Joetsu Shinkansen between Tokyo and Niigata opened in 1982, at which time Japan was growing at 4% a year and American bookstores were filled with “How to Think Like a Japanese” guides. Back then, to say that Japan had extremely low entrepreneurship was unthinkable. For what it’s worth, in the next few years Japan may well grow faster than Spain, which was hit really badly by the current crisis.

The examples you give about revitalization of intermediate cities on the AVE are true for the TGV, too. If I’m not mistaken, there’s a big office park in Aix-en-Provence near the TGV station (but the local economy is still dominated by tourism). Nord-Pas de Calais has managed to slow its decline because of the LGV Nord. But in neither case has the line prevented decline, and Marseille and Nord-Pas de Calais are still poor and deindustrialized. More instructively, Lille has a metro population of more than 1 million, but not enough travel to Paris to justify the construction of the LGV Nord by itself, without connections to London and Brussels. This suggests that even a city of about 1 million people should not be a terminus for HSR, unless it offers good connections to larger cities.

Japan has always had extremely low entrepreneurship. Innovation has always been driven by conglomerates of big corporations, big banks and political interest (which, to be fair, worked fairly well until the 90s, as you said) and was centred on huge metropolises.

The concepts of university spin-offs, start-ups and entrepreneurship are foreign to the Japanese, as are basic economic principles and freedoms. That makes people on outer regions depend on immigration or corporations setting shop on their home town for a living; while in Spain they could just found their own business (they don’t need to anyway, since others did before them).

Spain’s economic geography, complemented with modern services (including modern tourism and banking, which Japan lacks), is greatly helped by mobility, for reasons known to everybody. Let me show you a couple of stats to make clear what I mean by modern services:

1) Spain attracts 60 million foreign visitors every year, more than the whole USA, compared to just 7 million in Japan. You could wrongly say it is all because of climate, beautiful beaches or whatever; the fact is it requires good services, including a multi-lingual, specialized workforce (e.g. lots of university graduates in Tourism, Japanese-language guides, etc.). These foreign visitors can be divided into two groups: low-class Germans and Nordic young backpackers coming for the sun, getting beer at the supermarket and costing us more money than they leave in the country; and businesspeople in fairs and conferences and high-class couples in their 50s, who demand excellent services and can afford them.

As you may imagine, it is the second group that spends the most of the $50bn we get yearly from foreign visitors. For this group, the choice between 7-hous bus trips and 2-hour comfortable rides in the AVE is a no-brainer, as is the choice between countries with high-speed rail network and those lacking it. Now, if you do the math, the complete planned Spanish AVE network will cost around $150bn; i.e. three years of foreign high-class tourists’ spending. Let’s suppose having an excellent high-speed network means 10% more foreign tourists’ spending in the long run: then the AVE would pay for itself in 30 years even if nobody but foreign tourists would use it!

2) Services require more mobility than manufacturing. The Spanish service sector (including banking, telecoms, consultancies, engineering services, etc.) is more developed than Japan. That’s why both Spain and Japan have each around 4% of global market share on service exports, despite the Japanese being 3 times more populous.

Spain, as I have showed, has the kind of economic geography and service infrastructure to fully reap the gains of high-speed rail Japan lacks.

Now, you talk about the housing bubble. This has led many commenters to rightly question the content of last 8 years’ economic growth in the US, the UK and Spain, among others. Both the US and the UK got heavily in debt to fuel consumption, allowed their financial sectors to overgrow and lost global market share for manufactured goods (as did most of Western Europe and even Japan!).

Spain, meanwhile, got in debt to fuel investment (30% GDP against 17% in UK and 19% in the US), not consumption. Banking growth was prudently restricted and now we have a healthy banking system we didn’t need to bail out. Moreover, Spain and Germany were the only two Western European countries to maintain its global market share for manufactured exports in the last 8 years (something, I repeat, not even Japan could do). R&D investment, though low, has nearly doubled.

In this period, the US has not developed a single source of future income (if anything, barring Facebook). On the contrary, Spain, which didn’t lead the world in any industry a decade ago, now boasts the first telco in the world by number of subscribers (Telefónica), the first bank in the Eurozone (Santander), the first company in renewable energies in the world (Iberdrola Renovables), the second largest windmill manufacturer (Gamesa), the first clothes company in Europe (Inditex-Zara), etc. etc. Look at stats for pharma R&D growth, high-tech manufacturing, etc. or simply look at infrastructure improvements and you’ll clearly see the content of Spanish growth these last years, unlike in the US case (which, admittedly, started from a wealthier position).

Regarding Japan: In the last year, Japan’s economy has contracted over 9%, compared to 3% in Spain (the best y-o-y number in Western Europe). This is consistent with my explanation.

The reason Spain has more foreign visitors than Japan is that it’s smaller. Japan and the US are both large and isolated from most other developed countries, so their tourism markets are mostly internal. Japan may have very few foreign visitors, but Niigata has plenty of internal tourism. Spain is in the opposite position; however, the AVE-to-the-beach projects may have the same effect, as foreign tourists choose to stay in their hotels in Madrid and Barcelona and take day trips to Malaga.

The other issues you mention don’t point out to high ridership on midsize town to midsize town HSR lines. Why is local entrepreneurship more conducive to HSR than large corporate development? Why is Malaga inherently likelier to make good use of HSR than Lille? Why do services require more HSR ridership than manufacturing (note that France, with a strong service economy, has no more HSR ridership based on what you’d expect from population than manufacturing-oriented South Korea)?

Alon Levy,

“The reason Spain has more foreign visitors than Japan is that it’s smaller.”

You obviously don’t know what you are talking about. If that were the case, it would show on the stats as less Japanese spending on tourism abroad. Tourism’s “current-account deficit” (foreign spending in the country minus locals’ spending abroad) is:

Japan: $7bn – $38bn = – $31bn
Spain: $51bn – $15bn = + $36bn
France (number 1 by arrivals): $43bn – $33bn = + $10bn

Spain’s data are exceptionally high, especially when corrected for the size of the economy. Not even France’s data can be said to be similar. If you’re still not convinced: Spaniards rank first in complaints at foreign hotels, while Japanese rank last, which is consistent with a service-quality-conscious culture in Spain.

But foreign tourism is not nearly as important as domestic tourism. Málaga is experiencing a boom in weekend and business tourism as it became the nearest beach city from Madrid thanks to the AVE. Niigata should stop complaining and reflect on what’s doing wrong.

Now, I think we can agree that mobility helps a modern tourism industry more than an underdeveloped tourism sector. Let’s look at other services: are IT consultancies, architect bureaus, marketing services, lawyers, etc. mobility-sensible? Yes, since you can set up shop in cheaper Ciudad Real and serve Madrid’s market with bi-weekly visits. Are universities and science parks mobility-sensible? Yes, since you can have experts come and help you in specialized R&D.

But that means you have good universities distributed all around the country and some entrepreneurship in the first place. If everything is centralized in corporate headquarters in Tokyo-Osaka or, to a lesser degree, in Paris, it doesn’t work anymore.

Alan, actually the Los Angeles-San Francisco route is projected to be the highest ridership out of ALL the other networks.

Leave a Reply