Finance Hong Kong Metro Rail

Hong Kong’s Expanding Metro a Model of Development-Funded Transit

» New extensions in central Hong Kong will be mostly funded by development around future stations.

As one of the densest places on earth, it is no surprise that Hong Kong is a transit city, with over 90% of journeys made by public transportation. This concentration ensures high ridership on the city’s extensive MTR metro system. Yet it also increases the cost of line expansions, since building new projects in places that are already packed with people necessitates the creation of tunnels and requires expensive environmental and social remediation.

Hong Kong, however, is going about constructing new projects at a steady pace. At the end of last month, the MTR transit agency approved the construction of the 4.3-mile South Island Line, which will provide service to Aberdeen and other sections of the south side of Hong Kong Island. In addition, the agency will complete a 1.6-mile extension of the Kwun Tong Line to Whampoa. Both will open for operations by 2015, with ground breaking scheduled for next year. The city already has a 2-mile metro extension to the West Island under construction.

All this is being built in the context of relatively high construction prices: The Kwun Tong Line is being constructed at a total cost of HK$5.6 billion, or roughly US$450 million per mile. Though that’s less expensive than what a similar project would cost in New York City, where the Second Avenue Subway is being built for more than $2 billion a mile, it’s on par with most other Western cities building new subways, unsurprising considering Hong Kong’s wealth. So what is this city’s secret? How is it able to continue building new metro expansions — and plan for more — when other cities are being forced to postpone their transit projects due to the recession and the resulting government cutbacks?

The answer is that the MTR, in association with the local government, has become one of the city’s major property developers. It has used profits from those new housing, commercial, and retail schemes to pay for part of the cost of constructing new subway lines. Along the urban rail lines, the MTR has funded dozens of new housing projects with 300 to 7,000 apartments each. The operations of the subway are entirely unsubsidized by the local government.

This approach — called “Rail+Property” by the MTR — does not involve the city simply handing over development land to the transit agency at no costs (land in Hong Kong is all owned by the government, though it is leased out to private individuals and corporations for long-term periods). Rather, it is expected to pay the government the land costs estimated based on a no-rail scenario. Thus the MTR is not forced to deal with the problems many agencies face when they use eminent domain to take land, such as escalating values in anticipation of the new transit service. Rather, it is rewarded for the added value it will produce once its new transportation project is completed.

For the South Island Line, the government has agreed to provide MTR development rights to a site at the former Wong Chuk Hang estate. In turn, the transit agency expects the government to pay for less than half of metro construction costs. The rest of the tab will be picked up by the
MTR. This process, which directly associates transit operator with transit-oriented developer, makes the financing and construction of new underground transportation links far more simple than the typical approach, which requires governments to use public tax funds to pay for most of the cost of transit projects. The latter funding mechanism, common in the U.S. and Europe, is politically difficult and financially troublesome, especially in times of increasing budget deficits.

Of course, the MTR is not alone in using this method to assume some of the costs of construction. In Paris, the government’s grand plans for a network of 96 miles of new lines, costing more than €20 billion, would be partially paid for by the creation of development districts around each station whose own existence is only possible thanks to the creation of the transit line. This is a reasonable way to fund a transit line, as it is paid for by the benefits it produces. In a typical U.S. city, the transit agency provides a significant boost to local property values when it invests in a new rail project and yet it is able to capture very little of that extra wealth created — and when it does, only indirectly through tax revenue increases. This slows down the development of many cities.

One of the difficulties in proposing a similar finance method in the U.S. is that there is a residual fear of letting the public sector (or even the only partially public sector, in the case of MTR) engage in land development. The memories of urban renewal and public housing remain fresh enough on the minds of enough Americans to dissuade them from wanting to repeat the process today. Thus the closest we’re able to get is a financing method like that being used for New York’s 7 Subway extension. There, the city has created what is essentially a tax-increment financing district in the area to be served by the line, and it will use those tax revenues to pay back the initial costs of the project.

Yet this indirect approach, which does not allow the transit agency to reap the full benefits of the value increase it has sowed, leaves too much room for massive developer profits and too little room for actually funding new transit lines. It is worth considering whether it would be worthwhile to allow a transit agency to engage directly in the creation of new developments around the stations it has built. If properly managed, such a system could result in better transportation for all and less of a constraint on the public purse.

I should note that Hong Kong may be a unique case. At the aerial map at the top of this article shows, it is hemmed in from all sides by natural features — the ocean, mountains, and parks — that force all new development to be quite dense. Similarly, most housing and commercial activities are stuck in the relatively narrow strip of land between the ocean and the mountains. That density and linear concentration removes space for potential transportation infrastructure and limits the amount of walking necessary for anyone who lives along a transit line that follows the rough line of urban landscape; this makes transit work better here than most place. All together, you have conditions almost ideal for transit-oriented development. You can’t just transpose the Hong Kong model on any place.*

* This paragraph was added after the initial publishing of this article.

31 replies on “Hong Kong’s Expanding Metro a Model of Development-Funded Transit”

This is already being done with the 7 extension: Bloomberg claims that the $2.1 billion will be financed through future real estate sales and property taxes.

The problem with this model is the same as with anything else involving government in the US: there’s no guarantee the projected revenue stream will hold. Bloomberg came up with the idea that real estate would prop up his subway extension at the peak of the real estate market, when a $1 billion redevelopment deal was soon to be signed; the collapse of both the deal and the housing bubble did not lead him to reconsider.

There’s potentially another way of doing this. Around here, when a large development goes in, it frequently has to pay transportation mitigation fees. Given our auto dependency, this is taken to mean road expansions. Is there any reason it couldn’t be used to support rail-based transit?

It depends on where “around here” is, but it would probably be politically impossible as there’s a strong highway lobby but no “rail-based transit” lobby to speak of. Maybe in NYC or some other larger, denser city this may happen, but it would be an uphill battle.

The main problem with rail in the U.S. is that auto so completely dominates 95% of the landscape that efficiencies of scale make it that much more expensive for the 5% of the built environment where it makes very good sense.

This logic also explains why merely removing zoning restrictions won’t lead to dense, mixed use development. Developers just do a “copy and paste” job from the greenfield developments that they’re used to (and banks comfortable with financing) and you get more of the same, because sprawl and highways are the “default”.

Around here is specifically King County, Washington. I think we have gotten a few transit-related benefits from transportation mitigation. Some examples are: a transit center, a contribution to a P&R with structured parking, some added bus service, potentially an upgraded light rail station.

Tax-increment financing as currently understood doesn’t work for a couple reasons. First, politicians generally don’t have the courage/independence from the developers who finance their campaigns to actually raise tax rates in the benefitted area. Since property tax rates are low, the increase in property value is captured almost entirely by landowners.

Second, the details of bonding limit the amount that can be financed in this way. The TIF market is small and illiquid and usually there are restrictions surrounding the amount of the anticipated tax increase that can be bonded. It would be better to apply the TIF proceeds into the project’s general account, then finance wholly at the total project level. Breaking out elements of the revenue stream and financing them separately usually works only when there is a particularly active market for bonding that specific class of revenue.

In areas where the existing zoning is for single use, single residence per lot, possibly two story height limits, certainly residential and commercial parking minimums in place, one TOD strategy is an easement within a quarter mile of the transit connection ~ eg, 3 stories, waiving parking minimums or pooled parking, up to two professional/commercial uses per lot and up to four uses total, etc.

Where the auction of transferable easement raised below comes in is rather than trying to capture developer profits before the fact, capture a portion of them up front by organizing the easement as a transferable permit within the easement zone, auctioning in blocks with a reserve price.

That may fit in most comfortably with a transit infrastructure development bank structure, where the transferable easements stretch the capital base of the infrastructure development bank.

For the operating funding component, a VIF rather than a TIF, with the transit service provider receiving a fixed share (say, 20%) of the property tax revenue on the increment in value due to the easement.

NY’s MTA couldn’t run a profitable property development or management company if their lives depended on it. It would just be one more thing they would need taxes for.

Hong Kong MTR’s model is what streetcar companies did in the US 100 years ago. The private sector approach did have a certain downside though where the company had an incentive to skimp on service once all the property near a new station got sold.

Nowadays, here in the Bay Area there are private residential and office developments that provide shuttle bus service to downtowns or transit stations. Also, we are getting a new infill station on one of the BART lines at West Dublin, where part of the cost is paid for by a development company putting up some buildings nearby.

Also the same way the US Intercontinental railroad were constructed in the late 1800s. The Govt gave land away (or sold them cheaply) to railroad companies that promises to lay tracks. For 60 some odd years, companies like Union Pacific made more money developing property than actually moving passengers.

That’s a fine method for building infrastructure into a completely untapped area, such as the transcontinental railroad or many of the old US streetcar suburbs – it doesn’t work as well when you’re trying to build a new line in the core of an already developed city. The builder will need some other method of value capture.

When transit service is profitable, there is very little incentive to skimp on it. In fact the incentive goes the other way: provide as much as possible.

True, and in Hong Kong specifically, the real estate is a source of both independent profits and TOD increasing transit profits. So far I don’t think there’s much risk of Hong Kong (or Tokyo, which has a similar model) doing what the American companies did and treating the transit as merely a loss leader for real estate sales.

In the U.S., one of the big dangers of relying on the private sector to provide both the development and the associated transit service connecting it all is that as soon as one (the development) provides a higher value, investors will insist on stripping the company into discrete components.

The most recent example is Motorola, which was forced to break apart into two companies come next year: One for consumer electronics and the other for corporate and government technologies.

The American market can’t be trusted to let investors asset-strip.

I think the difference is in America we still have a large amount of greenfield space for development. Japan and most of Europe are pretty much maxed out when it comes to space for urban expansion. New development is mainly going to be infill (densification) and done vertically with taller buildings. A developer there isn’t going to have too many opportunities to drop one project for a more profitable one.

That may also be why congestion pricing works over there. People don’t have much choice. If you implemented it in an America city people would just go elsewhere.

Not necessarily and not necessarily.

We have a pretty long history of corporate conglomerates, and we have more of a history of our justice department breaking up companies than we do of investors doing the same.

Furthermore, while developments have been more profitable in the past (growing population and public policy that supports it), it is transit that is growing in demand and development that is stagnant in demand right now, and we are likely to see a continuation of those trends.

I don’t see how Hong Kong’s geography makes this particular model more viable. All geography does is increase the demand for these lines, it has nothing to do with the political conditions that make it possible for Hong Kong to use this system.
Looking at how mass transit systems are built in Asia and Europe makes me wonder if New York will ever be able to build systems as efficiently. Looking at our history it seems like we’re doomed to always see most of our transit spending siphoned off by corruption.

Have you been to Hong Kong? There’s a lack of flat land (even after a century of dumping more sand into the harbor). That makes land very expensive, and makes it possible for the developer to reap huge profits on any developmens. Because the transport company MTR is the actual developer (and it’s one of the biggest developers in the city) it can use these huge profits to invest in its system.
This is in addition to the huge profits it makes on its existing system (because in the eye of the consumer, the train right outside the door is a much better option than a congested highway (which is congested even more so because of the high density)).
Basically, the perfect conditions for a transit system exist in Hong Kong, and only in Hong Kong.

One approach for a suburban setting is a variation of a zoning easement approach:

(1) Within a quarter mile of the transit station or stop, the zoning easement allows multiple use, multiple residence construction, an increased height allowance, and waiver of parking minimums;

(2) The easements are vested with the transit authority responsible for construction, which auctions them to prospective developers.

(3) Easement auctions are reserved for capital works, while 20% of the property tax increment due to the easement is reserved for operating funding.

It should be easier to work a managable TIF-type approach (or any form of recovery on property improvement) in a suburban neighborhood, where you’re less likely to be dealing with huge developers that have great political power and less competition. One nice thing about the approach you mention (easement auctions) is that it puts some of the equity risk of the ultimate success of the project onto developers, rather than retaining all of it for the state.

My job kept me in Hong Kong for long periods of time so I can speak with knowledge of the Island and new Territories. HK has always had good transportation with an exploding population. Subways are air conditioned with each station color and material coded for easy ID. Small Public Light Busses brought the people to the subways – hot, crowded but needed. There is little parking for expensive private autos – and driving between the people in certain areas is challenging. In the build-able areas many of the buildings used for housing need to be either fully renovated or torn down and totally rebuilt. This leads into exactly what is being done per ur article.
Here in Los Angeles, people bought homes expecting that the Community Plans were created with forethought – the new transit district designation is causing SF homeowners property value losses and the inability to sell their property near/above projected subway tunnels. We need subways and adequate transportation, but at what cost.
The subways were first built under the British rule, now even though HK has not fully reverted back to Chinese rule – the Governing policies and methods are quite different from those found in the USA.
I agree developers should bear the cost of adequate transportation and also bear the cost of repairing and resurfacing roads leading to these developments but they should create their developments with a goal of fitting into the area in reasonable ways profitable to both the developer and to the community instead of looking at the maximum that they can build by law or code and shooting for that.

Maybe the US answer to this is summed up by NJ Transit’s Montclair connection. Here’s this crappy spur in a north Newark suburb…gee, it could work VERY well for funnelling trains from Morristown into NYC…but Montclair says yes to weekday trains, no to weekends. Even in NIMBY-saturated Germany, nobody would be this grossly impractical. In Germany, it’d be like saying local/regional trains can’t run through Potsdam, or Spandau, or Höchst, or Neuss, just because it’s the weekend and we don’t want a train every 30 minutes. In the US, a minor Jersey suburb and its insanely disproportionate role have been matched by Atherton, California’s lawsuit regarding rail alignment (specifically HSR as I understand it, but it also gets into Dumbarton corridor and its potential use for HSR). Property rights are one thing, but turning a rail line into a $500 mil/mile (mis-)adventure (or worse) is insanely wasteful.

It’s interesting that while this blog is admiring Hong Kong for its fast and efficient way of building public transportation, the British-derived system of planning makes Hong Kong appear glacial compared to its northern cousin of Shenzhen. It had no metro system at all until 2004, and Shenzhen’s system is also financed by and modeled upon MTR. At the same time proposals to expand the MTR are beset by NIMBYs, public consultations, and legislative debates.

It wasn’t so long ago that Hong Kongers looked down upon Shenzhen as being a cheap knockoff version of themselves. Now Shenzhen will have access to China’s new high speed rail system five years before Hong Kong does, given the delays inherent with Hong Kong’s way of planning. It’s a cruel twist of fate.

Seeing that many HKers weren’t too keen on the HSR extension, I don’t think it’s a given that they’d be upset over the delay. Having an HSR connection will not be a deciding factor in making Shenzhen a nicer place to live.

I don’t know about Shenzhen specifically, but many Chinese cities are embracing the car. Shanghai isn’t, but it’s an exception; others, like Beijing, are building massive expressways and subsidizing parking for the car-owning minority. In contrast, Hong Kong does nothing of this sort, resulting in the lowest car ownership rate and highest transit use in the developed world. Once much of the rest urban China turns into a traffic-jammed cloud of smog, Hong Kong may turn out to have the last laugh.

Hong Kong doesn’t consider itself part of China (and its citizens are actually wary of Communists). What makes you think they would WANT a HSR link to Beijing?
Also, the planning process in Hong Kong prevents stupid projects from going forward (like a full fledged, two-track 350 km/h railway to Urumqi, a small city in Western China).
Also, J B’s got a point. HSR doesn’t do anything for livability – it doesn’t make local schools better, lower crime rates, or provide better healthcare. People in Hong Kong are satisfied with what they have – and if they don’t like it there, they usually move to Singapore.

Travel and commerce between Hong Kong and the mainland is very important–while many HKers are distrustful of Beijing; the mainland is not viewed as an “enemy” to be avoided, or anything like that. The number of connections between the SAR and the mainland has been growing over the years, with a new bridge crossing the Sham Chun River into Shenzhen, and a new 32km (!) bridge in the works crossing the mouth of the Pearl River to Zhuhai and Macau.

HSR isn’t important for “livability”, but it’s very immportant for intermediate-to-long distance travel. HSR to Beijing might not make sense as a direct route (the distance is about 2000km, outside the “sweet spot” for rail–in other words, even accounting for airport delays and hassles its more convenient to fly, about a 3-4 hour journey in the air); but as part of a system connecting China’s eastern cities, it makes sense. HK to Shanghai (about 1200 km) is well within the HSR sweet spot, and it’s a similar distance from Shanghai to Beijing.

Certainly, the Chinese government likes to spend money on “show projects”, such as the ridiculous HSR line between downtown Shanghai and the airport, but your example is not one of them. Urumqi, with a population of 2.5 million, is not a “small city”. Do you think that Interstate 90, a 5000km freeway from Boston to Seattle (a “small city” in the Pacific Northwest), was also a stupid thing to do?

You’re right that below the elite level, Hong Kong isn’t all that supportive of HSR to China. But the line to Urumqi isn’t as stupid as you think it is. The terrain is so remote that the construction cost is very low, a little more than $20 billion for 1,700 km of route. Like many of the Spanish HSR lines, it looks better once you consider the fact that it doesn’t cost the same as your average American boondoggle.

> I think the difference is in America we still have a large amount of greenfield space for development. Japan and most of Europe are pretty much maxed out when it comes to space for urban expansion. New development is mainly going to be infill (densification) and done vertically with taller buildings. A developer there isn’t going to have too many opportunities to drop one project for a more profitable one.

In both America and Europe there is plenty of greenfield land that *could* be developed, but America allows pretty much unlimited development on greenfield land while Europe severely limits it. However, in big cities there is a limit to how much urban sprawl there can be before commute times become intolerable, and this limit has already been reached in many North American cities (Los Angeles, New York, Chicago, Bay Area, Washington, Toronto, etc.) Therefore high density development will have to become the reality in the US.

Its not just commute times, its also the total residence cost in terms of direct cost of residence and associated cost of transport. We experience the boundary where cheaper land is offset by more expensive commutes squeeze in during the 2006-2008 oil price shock, and that boundary will swing in by a greater or smaller amount in each of the oil price shocks that we will experience in the coming decade.

Leave a Reply