High-Speed Rail Urbanism

The Sprawling Effects of High-Speed Rail

» Faster travel times result in larger commute markets. Whether that means more suburban sprawl, however, is a decision left to municipal planners.

We’re told that French and Spanish villages, once too far from major cities to have any national consequence in an information age, have seen their fates reverse thanks to high-speed rail systems that have put them far closer than ever before to the centers of politics and finance. The result, apparently, has been a re-population of formerly dying towns with long-distance commuters working in the big city but interested in living at the rural edge.

For American urban planners, this prospect puts a blot on the otherwise shiny reputation of fast train service, which many see as a key to encouraging the growth of downtowns and promoting a new era of urban living. But if high-speed rail also allows people easier commutes across ever-increasing distances, couldn’t it encourage sprawl? And if it does so, is it worth constructing?

Jason Kambitsis reviewed the issue this week on Wired, suggesting that U.S. land use patterns would likely mean the construction of huge, sprawling subdivisions in places like California’s Central Valley, which would suddenly be within striking distance of downtown Los Angeles thanks to a proposed high-speed rail system. Though stations themselves are planned to be in central cities, people might drive from elsewhere nearby to get onto the train.

For the most part, though, fears of rapid population shifts to the exurbs as a result of the construction of a high-speed line seem unnecessary. In France and Spain, despite seeing redevelopment around some stations, there has been no clear correlation between growth and fast trains. The French region that has benefited most from TGV service — arguably Nord-Pas de Calais, in the north of the country — has also had some of the country’s slowest growth rates (mostly because of the continued effects of deindustrialization).

In places like Avignon (pictured above), where high-speed stations have been constructed outside of city centers, some suburban growth has followed, but this is more likely a result of the preexisting economic dynamism of the area than the damaging effects of train service.

The concerns over a migration into the countryside thus seem exaggerated. There are clear reasons for this: Even if tickets on fast trains can be offered relatively cheaply, they will still be far more expensive than the average cost of commuting by car or public transportation within a metropolitan area. Even if cheaper housing can be found thanks to a long-distance commute, transportation costs on high-speed rail will be so high as to make the service unaffordable for most users. These are very clear economic reasons, then, why few people will choose to ride the fast train to and from work every day — and that will limit the demand for increasing sprawl creation.

Nonetheless, some increased development far from traditional center cities does seem likely. The few people who will be able to afford paying daily round-trip tickets may see the benefit of living far into the suburbs, but the European example people like to cite in stoking fears of sprawl is not particularly relevant to the United States. Wealthy people in France and Spain may choose a home that takes advantage of romantic village life and then travel quickly by train into business centers in Paris or Madrid, but Americans have no similar choice, since the village is basically a non-existent entity in most of the U.S. and the only real alternative to urban life is the low-density cul-de-sac, found almost everywhere.

In fact, people who can afford so many train tickets have no real incentive to choose the same old suburban sprawl far away when they can get it close to their work as well. There just doesn’t seem much of a market for such commutes.

Some suburban areas, however, do seem likely to be built up thanks to the increased growth of medium-sized cities which are suddenly connected to major metros thanks to the rail system. The cities of Fresno and Bakersfield, for instance, are likely to see increased development interest because they’re both set to receive stations on the proposed California line. If their downtowns grow up around stations, their suburbs may see further habitation by residents who work downtown, but not significantly of people who plan to take the high-speed trains somewhere else. The train may encourage urban growth in today’s forgotten areas, but it won’t do so by developing a whole new class of long-distance commuters.

In the Wired article, Kambitsis argues that this type of sprawl could be prevented through the creation of transit-oriented development zones in the affected cities, encouraging people to live in denser areas near station sites. And he argues that improved regional transportation choices could play an essential role in promoting the use of transit by even the people commuting long distances.

Indeed, there should be some effort by cities with stations to leverage the benefits of station construction to spur smart growth, rather than more monotonous landscapes of single-family homes. Engaging in an effort to orient new businesses and housing in the most efficient way possible has to be an essential element of any new high-speed system’s construction, and it is something Vision California has attempted to undertake at the statewide level. Yet fast trains will not be the primary cause of sprawl in any place: most people living in each metropolitan area will continue to work in that metropolitan area, so high-speed rail is not so much the culprit as a mechanism to evaluate whether existing growth paradigms are sustainable.

It remains to be seen whether cities like Bakersfield or Fresno will find themselves compelled to fight for a better, different type of urban form, or whether they’ll simply pass back into their old assumptions about how housing should be constructed. Perhaps the construction of new stations in their respective center cities will prove to be a watershed event in their thinking about how to plan for growth.

Image above: Avignon TGV Station, by Flickr user jean-louis zimmermann (cc)

42 replies on “The Sprawling Effects of High-Speed Rail”

I remember reading somewhere that Japanese officials claimed that building the Shinkansen would lead to a country less centralized in Tokyo. What really happened was a country with much greater access to the capital than ever before and more centralization.

Also, if dense growth happens on cheaper land, it allows people of lower incomes to live in denser, more environmentally friendly areas without paying the high costs of living in the existing city center. Small towns in Europe don’t feel like isolated blips of suburbia off the highway, they have a real sense of place.

People using HST to commute don’t have a lower income. It’s quite exepensive to commute by HST and not everybody can pull this off.

From what I’ve read they mostly are well-off urbanites that have traded the thrill of the very big city for the relative tranquility of a medium sized city.

It’s expensive to commute long distances but the lower cost of construction and much lower cost of land more than make up for increased transportation costs. Where do poor people in Europe live? In the suburbs and small towns.

Alex, the poor people of Europe live in Eastern Europe, or in very poor Western regions like Southern Italy.

The poor people of France live away from the core regions (Ile-de-France, Rhone-Alpes, PACA), and to some extent the suburbs in the core regions. The provinces are all littered with small picturesque towns that actually have some money, including the towns with TGV stops, but otherwise they’re poor.

Japanese politicians say that everything will lead to less centralization (and it usually doesn’t)

The centralization in Tokyo has more to do with the economy than with transportation (most Japanese cities were former manufacturing centers; think Detroit or Cleveland).

Also, about all the confusion about Europe – there’s the inner city (expensive), the inner city suburbs (poor), the outer suburbs (more expensive than the inner city suburbs, but less expensive than the city itself) and then the poor provincial towns.

In Northern CA, it is not uncommon for people to already move out to the exburbs for cheaper homes in exchange for longer commutes. 1-2 hour commutes by car are the norm already.

The problem will not be in Fresno or Bakersfield, but rather, Los Banos (plenty of potential for sprawl) and possibly Palmdale in So. Cal.

Given the extreme in balance in housing costs between the inner Bay Area and the Central Valley this may be a real issue. CA is not France.

Elizabeth Deakin addressed the potential for sprawl in the Central Valley at the UC Berkeley symposium last October on high speed rail: . The Central Valley cities will need good transit feeder systems that connect to the high speed rail stations. There was a very detailed analysis (I forget which city– I think it was Modesto) of the potential to build around the planned high speed rail station.

Fortunately, there will be no Los Banos station on the California HSR system. The current plans do not include that station, and I believe the law that put the bond measure on the ballot actually excludes a station in Los Banos.

There may be increased development in Merced, but the University of California campus was already going to cause growth. And Fresno and Bakersfield could really use some pull towards development of their central business districts.

The Palmdale station is the most likely candidate to cause unwanted development, but it already has a 1 hour 15 minute train to downtown Los Angeles, and I haven’t heard anyone worry about Metrolink causing more sprawl. As Yohan wrote, the results will depend on local government’s plans for the area.


I remember reading a column in the Sacramento Bee several years ago suggesting that building the UC campus in Merced rather than Fresno was a terrible decision. It is largely a gift to developers there so they can build more single-family homes, much as UC Irvine was planned and built.

I’ve seen towns with commuter trains in New Jersery that where at least 50 to a 100 miles from New York City and everything was still based around the major highways and freeways. The train station did have a small group of stores and parking spaces around it but not really that big. I think the highways are still going to be the future sprawl generators.

Those commuter towns’ trains are almost never faster than driving, and never much faster. So once the train station had seeded the town, and once cars became widespread, it was easy to drive to most destinations.

The issue is that short drive-times to a secondary HSR rail station (Waterloo, WI; Modesto, CA) will open up vast swaths of land to sprawl. Instead of living within the Chicago or Milwaukee drive-sheds, commuters will have the option of a short drive + a pleasant, amenity-laden train ride in to work. How convenient. Expense is not a deterrent.

So Los Banos’ lack of a rail station may not be relevant to the sprawl question; people will drive to HSR stations as long as land use, time constraints and highway access allows that to emerge as a habitual pattern.

That’s what happened in New Jersey. Park-&-Ride facilities are customary and expected; there was no attempt to install sane land use controls near stations—and no desire to do so. So the default outcome is minor-league town centers near stations, with growth & commercial districts exported to other towns, to NYC or Philly, and to auto-accessible campuses on the periphery. Everyone expects to drive everywhere, and has to because of a fealty to single-family homes & much hand-holding by the self-interested highway-building industry.

That disinvestment is what killed Newark.

By adopting structurally innovative TODs & strong land use regs, HSR stations in other states can beat sprawl—as long as they design forward-looking places people actually want to live in. And that’s an open question.

MTA refused to extend BART service to outlying stations until they adopted specific TOD standards. Not a bad ‘stick’ to secure land use-transp coordination.

What happened in Jersey is the exact opposite of what you say happened there. First, there was pervasive auto-oriented suburbanization, opened up by projects like the Holland Tunnel and the Turnpike. This was in the 1920s and 30s. It was only in the 1960s, in response to that trend, that the Pennsylvania Railroad started building park and rides.

This is not what killed Newark, by the way. The same trend happened on the other side of the river, and New York managed to do okay. What killed Newark was the city leaders’ inability to annex, and their eagerness to level every neighborhood to build highways and project superblocks.

For large commuter volumes, none of the proposed stations is close enough for anything, except Palmdale and Gilroy. That’s why the Sierra Club was so insistent on dropping Los Banos. The other stations are probably going to produce some sprawl, but to a smaller degree.

AlexB: That’s true, but the suburbs they live in tend to be within metropolitan areas. An HSR commute is far more expensive than a commuter-train commute. Also, in most of Europe HSR fares are cheaper than they probably would be in the U.S.

The sort of easy answer to this – don’t build stations where you don’t want growth. Plus it will just make transport between two points faster by eliminating stops.

Ben is right to point to Elizabeth Deakin’s detailed, fascinating proposals for Transit Oriented Development in the Central Valley. More details are here:

Some of the CV cities could use a lot of help in becoming welcoming places to live. Still, I wouldn’t be surprised if whatever new core development happens there generates a round of sprawl around them, especially since local policymakers are unlikely to see this as a problem at first.

Who cares if HSR creates sprawl if the sprawl that is created is not of the same character as current exurban sprawl. People will drive less if they have access to more forms of transportation that connect them to the urban cores. People’s lifestyles will require fewer carbon inputs and we will be achieving what is ultimately the goal of all of these projects.

Our lifestyle as it stands is not sustainable and HSR allows people begin to drive less and live in slightly denser communities then we it will have accomplished a great thing for this country.

In the case of California’s high-speed rail, what would likely happen is the Central Valley cities would become giant park-and-ride lots.

High-speed rail would produce a labor/commute shed where 85% of the state’s population lives.

We don’t know what the fares are going to be yet, but chances are there will be some commuter rate available. Some may see the sprawl threat and demand fares that would deter Central Valley to Bay Area or Southern California commutes.

It depends on how high.

On Metrolink, the most expensive station pair option is from Lancaster to Oceanside. That’s a distance of 155 miles, and it’s pretty much a practical impossibility to pull off. However, the prices listed for it are $26.50 for a daily round trip and $402 monthly.

So on the one hand, HSR would be sprawl on steroids. California would have a giant Bay Area-Central Valley-Southern California commute and labor shed. This is, though, where 80% of the state’s population lives.

On the other hand, Central Valley sprawl came first. This is a primarily agricultural region with a population of about 5 million. Worse, the Central Valley gained infamy for being one of the worst-hit housing crash areas in the entire U.S. So there’s a glut of housing.

With high-speed rail, the Central Valley might have a shot of becoming the megalopolitan area it aspires to by being a labor shed to the Bay Area and Southern California. While the cities may hate becoming bedroom communities, the Central Valley would see a transformation similar to Orange County’s, which first became a bedroom community of Los Angeles and San Diego then gradually developed an urban economy.

I don’t think it will be that bad. First, one can safely assume that HSR will bring new bussinesses in vicinity of stations, most of their employees will be local. Second, connecting from train to already established destinations like universities by anything else than transit sucks. Do it the right way and local will start to use it to get to station too.

Dejv, the issue isn’t so much locals finding a way to use stations. For Amtrak California, people find the stations just fine.

I also do expect the transfer of some job functions that would make the Central Valley a true job center. Its problem right now is that it has all of California’s high costs and very little of the relative economic attractiveness of the high-cost coastal areas. Putting the Central Valley within commuting distance could help correct that.

The Central Valley has cheap land, universities and a large labor pool, but that’s about it. This describes a vast cohort of American regions that are either fast-growing agricultural exurbs or desperate suburban and industrial regions in decline. The Central Valley cities are a mixture of both.

High-speed rail, despite the onerous capital costs, at least gives California one advantage in that it connects three populous state regions together. While the business plan seems to rely on day trips between Southern California and the Bay Area, this implies the Central Valley will be within commuting distance of both regions.

This locational advantage makes it feasible for businesses in high-cost coastal California to shift lower-value segments to lower-cost Central Valley while still retaining a connection to the higher-value area.

This would help businesses that have lower-value functions that for logistical reasons cannot shift the business unit out of state, and/or jobs that require frequent training or other constant contact between workers. Sectors that could see gains include healthcare and biosciences, law, finance and government.

I did remember talking to someone who lives in Ashland Virginia and drives up to Fredricksburg to get on the Virginia Express. They said, “They would love to take Amtrak the whole way if it had better departing and ariving times. I think the high speed rail can’t do as much damage as a major highway. In that a highway can have as many openings and exits as builders want. But a high speed rail line would only have a small number of tran stations for stops so it would be far more limited.

Wouldn’t the HSR fares be too high for commuting purposes? I don’t think they were going to give away tickets at, say, Metrolink or Caltrain prices….

I suppose, if you’re making enough money, HSR might let you live in Fresno and commute to San Jose. Heck, there are people commuting on *airlines* now. Again, though, these are mostly specialized, high paying jobs.

Another market for these types of trips are college students, who may wish to go to a particular school, but, for whatever reason, do not want to move/live closer to campus. (Near the end of my college career, I was taking Greyhound once or twice a week between LA and Santa Barbara, to finish up a few pesky last units. But I felt no need to go rent an apartment for that….)

It might be too high, or it might not. It depends on the pricing – at Shinkansen non-reserved seat all-stops prices, it would be feasible, in small volumes. Fresno is probably too far away for it to be time-competitive, but Palmdale and Gilroy aren’t.

I’ve actually often cited “sprawl” as a reason not to build high-speed rail; not to mention that HSR is typically not a “commuter” service.

In the Pacific Northwest, Oregon is planning an HSR route that would actually remove trains off of the current Union Pacific (formerly Southern Pacific) route that serves Portland, Oregon City, Salem, Albany and Eugene – and replace it with the former Oregon Electric route (once an interurban line) that extends off of the current WES Commuter Rail route from Wilsonville (itself a sprawling “exurb” of large, widely spaced industrial sites and warehouses, devoid of a city center, and whose WES station is located a walking mile off of the main west-east route and nudged up between warehouses and a protected wetland that cuts off access to the nearest tract housing development) south through miles of farmland until reaching Salem – skipping past communities like Canby and Woodburn, where the Union Pacific rail line (which hosts Amtrak) run through through the center of those communities’ downtowns.

HSR would devestate much of Salem by eliminating hundreds of homes in a lower-income area; would wipe out an industrial area on the north side of downtown, and would require significant land to build a station facility on the western edge of downtown. (The current Amtrak station is located on the east side of the core, near Willamette University, the Capitol Mall, and the Salem Hospital – all major employment centers, as opposed to the more retail oriented area to the west.) From Salem to Albany, HSR would again pass through undeveloped areas and farmland until reaching Albany.

HSR is being seen as a “solution to all problems” when in fact it solves virtually no problem. What HSR did in Europe wasn’t reduce traffic congestion on local roads, it reduced or eliminated short-distance air travel. Which isn’t a bad idea per se but Americans are being sold a bill of goods that doesn’t meet the needs it is sold as doing. HSR would eliminate the Horizon Shuttle and United Express flights between Portland and Seattle (31 flights in each direction for the 160 mile route). But there are only seven flights between Portland and Eugene – NONE between Portland and Salem (or Salem and Eugene), and current Amtrak Cascades ridership is barely enough to fill two buses – and the trains on average get a 25% load factor.

HSR in France removed cars from the intercity highways. To see the result, look at how congested they get whenever SNCF goes on strike.

And as for Portland-Eugene, that corridor is low-performing. The only reason it’s seriously considered as an HSR corridor is because the Amtrak route that serves it is the same one that serves Portland-Seattle, which is a good potential corridor.

> HSR in France removed cars from the
> intercity highways.

hmmm… I don’t get that feeling when on the A6/A7 between Friday and Sunday in July and August…

(A6/A7 is the Autoroute from Paris to Marseille via Lyon)

Well, yes, it did remove some cars between Paris and Lille or Lyon … on the other hand, anyone driving into Paris deserves it .

I just read somewhere that when the Intercity trains were introduced in the UK house prices went up in a lot of cities served by it as they were now within commuting distance of London, and I’d expect similar to happen with HSR in the US as well. Of course, as someone said elsewhere, we don’t have the country charm quite like France, but plenty of small east coast towns and midwestern towns have their advantages and fans. I’m thinking the big old manufacturers houses from 1900-1960 or so which are far cheaper than urban suburbs.

You have to remember how small the UK actually is. London to Birmingham is less than 150 miles, well within the distance people in the US commute BY CAR routinely.

Yes, rail lines of that sort of distance will create increased home values as commutes will be possible by train instead of car.

I’m just going to question your claim of a “large station facility” – platforms aren’t that wide, and there would be a few things like stairs and a ticket booth, but not much else.

Your theorizing about current Amtrak ridership is also flawed – it’s only very low because the service is total crap, not because there is no demand for that sort of travel.

Personally, I would prefer that there be no HSR in Oregon, not because of the “sprawl”, but because money is tight and could be better spent elsewhere. The Administration needs to set its priorities straight instead of flooding every state with cash.

Nobody is proposing HSR in Oregon. They are proposing upgrading the almost conventional rail.

Mr. Freemark is way off base on this one. For “sprawl” one needs cars. And Gen Y won’t drive (is not currently driving), favoring trains and light rail. YES, one can commute long distances with high speed rail. YES, small and medium sized towns with a train station will grow and prosper. But people will live near the train station. No sprawl, just livable communities. Try again to find a downside to trains. The sprawl idea doesn’t fly.

HSR will provide access, and help create more efficient land use both in the big city destination and the stops in between where people can live in smaller towns.

To the authors point – Will people choose to live 1 – outside the major centers as a result, 2-live in the big cities, or 3 – stay-in-place? The answer is an ‘all of the above’ and more, as we continue to grow (even in a recession). But for each of these trends we need to be building more efficient:

1 – outside the major centers = more clusters of village at much high dwellings/acre than traditional sprawl
2- live in the big cities = more multifamily housing
3 – stay-in-place = retrofit housing for more residents per unit

And all trends toward less car dependant and parking garage-less transportation systems).

Look for major redevelopment in big cities near stations. Look for more dense commuter towns to be developed in the areas with good locals services. The other suburban and exurban areas without new transportation investment (and other community amenities) will likely decline.

We have already sprawled (for the past 50 years) and paid a handsome price, but now we tax-averse Americans, although car loving we may be, will not want, not can afford to continue to pay for the increasing costs of a ‘surburban nation’.

Now the challenge is to infill both in the urban centers, and connect suburban villages. In short, doing more with less funding, land space, energy, and personal vehicle mobility.. just like the rest of the world.

“1 – outside the major centers = more clusters of village at much high dwellings/acre than traditional sprawl”

In other words, we get more beautiful streetcar suburbs and commuter villages, and fewer asphalt wastelands? :-)

Yes, seems like an improvement!

Not to mention that cities connected to a large high-speed rail network will start to focus more on the education, beauty and transit-convienice in the city rather than the workforce. You see, for example: if you get a job in Seattle, and there is a bullet-train connecting the Eugene-Portland-Tacoma-Seattle-Vancouver Megalopolis, then that person is able to choose which city would be best to live in rather than focusing on how close he should move.

Basically, which HSRail, cities will try to appeal to those moving into the area.

Theoretically any mode of transportation can lead to sprawl. Of all options for long-medium distance travel, HSR is the least likely to contribute to sprawl. This is of course because high speed trains can be built to run on modified existing railroads at lesser speeds, thus making it possible for trains to run to and from city center. However stations are often built outside of cities, so to not have noise, property destruction, etc… in densely populated areas.

“…huge, sprawling subdivisions in places like California’s Central Valley, which would suddenly be within striking distance of downtown Los Angeles”

If downtown Los Angeles had any jobs, maybe I’d be worried.

There’s just one problem with this sprawl theory: like every other projection that exists, it assumes that the economy is a giant, happy ball of sunshine.

Unfortunately, the economy is in awful shape. The housing market is total crap, especially in California, the state where HSR will (hopefully) be finished. A significant amount of existing homes would need to be sold before developers could show interest in (and make a profit off of) more housing around the HSR stations. Sprawl is something to worry about when the economy is wonderful, not in times like these.

Leave a Reply