Quick Note: The Hong Kong MTR is Profitable

There’s a pervasive myth that the Hong Kong MTR is profitable only because the company’s real estate investments subsidize the train operations. For a trivial refutation, go to the MTR’s 2008 financial statement. Operating income was HK$13,995 million, which breaks down as $4,670 million from real estate development and $9,325 million from railway operations. Net income was $8,280 million; the statements do not break down depreciation, amortization, and interest charges according to whether they come from transportation or real estate, but the operating profits from transportation would’ve been enough to cover everything.

For a comparable link to Japanese private railroads, another source of the myth of development-subsidized transportation, see this article from JRTR.


  1. Danny

    Thanks. I have often wondered about the origins of this myth. No doubt the US has its history of companies that subsidized transit as a loss leader for real estate development, or required government subsidy…but it also has its history of streetcar systems, rapid transit systems, and intercity rail systems that were a) not subsidized, and b) never had real estate investment as the core of the business model.

    • Alon Levy

      My guess is that the myth comes from the fact that the MTR does use real estate development as a) a source of extra profit, and b) a way to generate demand for the rail lines. This too exists in US history – owners of elevated rail lines developed the areas at their outer ends, often as amusement parks. Probably this mutated to “the MTR is only profitable because it can also develop real estate” (possible, but doubtful) and then to “the MTR’s real estate business subsidizes transportation” (wrong).

    • Wad

      Yes, but usually they were investor-owned utilities. New Orleans (NOPSI) was one of the most famous, and it kept the transit operation well into the publicly operated era.

      • Danny

        Of course they were investor owned. And there is nothing stopping investor ownership currently either, other than decades of historical political meddling, villainizing, and populist price regulation.

        And it wasn’t greed or any of the typical anti-capitalist bogeys that brought them down…it was the same forces that currently inhibit public transit growth: Our own government and their pro-auto policies.

        • Wad

          Danny, NOPSI is New Orleans’ electricity and natural gas provider. Until 1983, it also ran New Orleans’ bus and streetcar system. It was all too happy to be relieved of the burden.

          Ditto for South Carolina. There, SCE&G operated the bus systems in the state’s biggest cities well into the 2000s! Private, yes, but a major money-suck for the company and the state. Subsidy transfer payments and a utility bill surcharge accounted for most of the revenue.

          These were the last two in recent memory to be operated by private utility companies. Back when traction was commonplace, utilities as mass-transit providers were also a civic fixture.

          Private, yes. Money-making, no. If ledgers are still available for utilities’ transit operations, they may show that utilities were likely just as bad as modern publicly run transit systems in cost management and productivity. Why? Utilities were built up as corporate monopolies. In exchange for a monopoly franchise, communities were able to negotiate goodwill out of the utility company. One of them, commonly enough, was to provide some form of transit service.

          This sort of arrangement made sense for its time — namely when ubiquitous automobile ownership was not yet viable. Also, traction justified the capital cost of overbuilding a power plant, as there would be a constant demand for any slack electricity-generation capacity.

          So even under private management, these companies never put an emphasis on mass transit as a profit center. It was always seen as goodwill or a catalyst for other products.

          Danny, I could think of one good reason why investors wouldn’t currently want to own a transit system: No one goes into investing with the intent of losing money. You’ll know this going in. Pull up your favorite transit system in the National Transit Database and see why.

          I could tell you, here on this board for free, to not take this sucker bet. ConEd could devote its time, money and brains into a much more precise and accurate study and come to the same conclusion should it seek to take over MTA in New York City.

          • Danny

            Telling me that there aren’t any transit systems with profit potential because none are currently profitable is like telling someone that it is impossible to pass a sensible budget because congress can’t. The MTA has potential for profitability on the scale of systems found in Hong Kong, Taiwan, and Japan…because they have ridership levels on the same scale. Even apart from the MTA, there are plenty of systems that could be profitable with a combination of the right amount of investment and the right operating environment. There are some systems that could be profitable tomorrow with just one adjustment: bring wage levels in line with those of the private market.

            The truth is that historically there were plenty of non-utility, investor owned, and profitable transit systems that stayed profitable up until the point where local governments were given the incentive to take them over. They had decreasing profitability because they had decreasing ridership, a fault that can be placed almost completely on our own government…not the ability of the enterprises. Sure, plenty of them were happy to sell out to the government after years of trying to compete with subsidized competition…but that isn’t because they weren’t capable of making a profit.

            And sure, government subsidy of the automobile still exists which would be a problem for potential transit investors. But there is one major trend that exists today that didn’t exist in the mid 50’s-late 80’s. Increasing ridership.

  2. Wad

    I generally don’t hold out hope that transit systems run in tandem with development firms could be done in North America.

    If left up to the market, the spread between the money made by real estate and from the farebox would be so great that the assets would be stripped and the transit operations dumped.

    • Andre Lot

      That wouldn’t be necessarily the case. Where “light rail to nowhere” is viable, having one end in greenfield developments and other (or some point through the line) in meaningful destinations of the same metro area in a logic route, there is potential for TOD, at least, coming upfront with money for capital investments.

      We had a recent example of that in Salt Lake City, where the developers of Daybreak (www.daybreakutah.com), a “smart” new development aimed at the upscale market of single houses yet incorporating very interesting TOD design features, helped fund an extension of light rail.

      As for keeping fares low to attract residents, that became economic non-sense since cars replaced horse-powered or bicycles as the only alternative to streetcars (thus making the advantage of “cheap” streetcar fares much more meaningful to attract residents than today).

    • Danny

      This is ridiculous. There isn’t a single fortune 500 company that doesn’t have an unprofitable division, let alone an underperforming division. There are plenty of reasons to stick with underperformers, ranging from the hope of turning it around, to strategic positioning, to the hope for future growth.

      And when companies do decide to get rid of an underperforming business, they sell it. Even when it isn’t profitable.

      Dismantling a business only becomes an option when liquid asset value is greater than either the highest bidder or the NPV of expected future profits.

      • Wad

        Danny, if it were so ridiculous, why isn’t there any eagerness on the part of any Fortune 500 company, or any small upstart for that matter, to give public transit service a go?

        I don’t buy the ruthless enforcement of monopoly on the part of transit agencies or their unions. Even in New York and Miami, where Afro-Caribbean communities started up jitney services, there was an attempt to snuff them out until a broad community coalition of the left and the right rose to the defense of the jitneys and helped legitimize them. Also, there are end-runs around the monopoly enforcement. Charter bus companies can run peak-hour commuter services, or office buildings and suburban office campuses can run members-only shuttles.

        The problem comes in the workhorse urban bus lines. Even with high ridership and fabulous productivity, these still don’t come close to covering operating costs.

        Again, investors don’t go in with the prospect of losing money. Yet, looking at today’s profiles of transit agencies, investors are guaranteed to spend $1 to get back only 20 to 33 cents from a bus, and 30 to 60 cents from a train.

        And this leaves nothing for capital. Good luck finding anyone willing to lend to you if you lose 80 cents for every 20 you earn.

        A pure transportation concern would fail. There would need to be something else, like development endowments or a more profitable utilities unit, that would need to cover those losses. But subsidies don’t go away. Investors won’t tolerate those subsidies, especially since utilities and real estate can be profitable without a transit component.

        • Danny

          There are several companies that have given public transit a go, and some have been fantastically successful at it despite competing with the subsidized government. They ALL exist in areas where public transit monopolies do not exist. Jitneys for one, but intercity bus routes are also included.

          The monopoly issue is real. The idea that one could invest tens to hundreds of millions of dollars in capital in an illegal business, just because the government might not enforce the monopoly, is ludicrous. Food Trucks can get away with semi-illegal status, but they will speed away at the slightest hint of a cop to be able to protect their $10k investment. It is one thing to buy a used van to operate outside the law…another thing entirely to buy a fleet of modern buses and try to operate as a common carrier outside the law.

          And again, today’s profiles of transit agencies are completely irrelevant. It is like comparing apples and bulldozers.

      • Wad

        As for cost savings through labor, the single largest expenditure of a transit system, I ran some numbers through the Bureau of Labor Statistics database.

        I couldn’t copy a working link from the database, but here’s what I found for the three primary front-line occupations of a transit system and their 6-digit classification codes: Bus Drivers, Transit and Intercity(533021); Bus and Truck Mechanics and Diesel Engine Specialists(493031); and Cleaners of Vehicles and Equipment(537061). These figures are for the New York MSA (including New Jersey and Pennsylvania).

        The bus drivers’ mean wage is $21.90 an hour and $45,550 a year. The percentile scales are 12.61(10%), 15.53(25%), 23.78 (median), 28.01 (75%) and 30.72 (90%).

        The mechanics’ mean wage is $25.07 an hour and $52,150 a year. The percentiles are 15.84(10%), 20.39(25%), 25.95 (median), 29.69(75%) and 33.12 (90%).

        The cleaners’ mean wage is $13.85 an hour and $28,820 a year. The percentiles are 7.79(10%) 8.84(25%) 12.05 (median) 18.01(75%) and 23.59 (90%).

        This is the market wage table for the New York City area, encompassing public and private operators and union and nonunion shops. You do see a step pattern that reflects skills and labor supply.

        Cleaners would be the least difficult job among the three, and you would also have access to a broader working pool. This reflects the greatest public-to-private pay disparity out of all of the jobs. You can pretty much tell right off that MTA and New Jersey Transit contribute to the wages on the north end of the median. Statistically, half of your cleaners’ payroll will be between 8.84 and 18.01. That’s a 104% cost escalation. For the universal range from 10% to 90%, that’s more than 200% escalation.

        Bus drivers are in the middle. The educational requirements are the same as a cleaner, but the job is far more physically demanding (very high stress, fatigue from driving, and irregular start and stop times that can disrupt a driver’s body clock). The labor pool, though, is more limited because of the licensing requirements and federal and state time-of-work limits for commercial drivers. The spread for half of the drivers’ payroll is 80%. The universal spread is 144%. By inference, this means that public and private wages are somewhat closer together. Upon further inference, this means there’s less of an opportunity for savings than in the cleaners.

        Mechanics are the highest paid. This reflects the knowledge of the profession, as the job entails both heavy classroom and experience-based instruction. Note that the entry-level mechanic can command a higher wage than the bottom 25% of bus drivers. It’s north of the median when bus drivers come close to mechanics for wage parity. Here, the half-spread is only 46%. The universal spread is 109%. This would likely mean that the supply of mechanics is the thinnest, and this category has the least opportunity for cost savings.

        • Danny

          It is funny you bring up New York when throwing out your worthless mean/median (as in not accounting for the difference between public and private market) labor statistics. Funny because labor costs aren’t significantly lower in Japan or Hong Kong, where their systems with similar ridership are quite profitable. New York’s system should be profitable, in spite of its high labor costs.

          If you want to do some real research into how wages can affect profitability, look into a smaller city like Pittsburgh, Dallas, or Denver. In these locations the cost of living is much lower which means a higher supply of unskilled and low skilled labor.

          Furthermore, don’t just research mean/median wages for occupations. Research the public/private compensation discrepancy (including benefits). With public employees making 44% more than private sector employees on average, there is plenty of room for improvement. Furthermore, the public/private gap grows weaker as skill levels increase, which implicitly means that the gap is wider for unskilled and low skilled labor.


          • Wad

            The data are from the Bureau of Labor Statistics, not me, and are dated May 2010. if you think the stats are worthless, take it up with the BLS.

            You wanted knowledge of the wage market? Now you have it.

            (as in not accounting for the difference between public and private market)

            Irrelevant to the discussion. These wage scales are helpful because they don’t reveal the differences. In New York, there is a heavy market for transportation workers for private and public entities. If MTA and NJT are the major leagues of transit, school bus companies, Greyhound, Megabus/Coach USA, Peter Pan, and charter/tour companies must still find workers and give an incentive for workers to play in the minors, so to speak.

            If you’re running a transit property and you can’t find anyone to work with you, or you are deluged with applicants, these wage scales serve as useful guides.

            Funny because labor costs aren’t significantly lower in Japan or Hong Kong, where their systems with similar ridership are quite profitable.

            Then explain why. Don’t say it’s because they’re private for-profit companies.

            We could do that here and still see our lousy farebox recoveries. (Westchester County’s bus service is owned and operated by a private company, Liberty Lines. Its farebox recovery in 2009 was 33%).

            What do Japan and Hong Kong do that could be legally and practically imported to North America and Europe? Japan and Hong Kong both have transport workers’ unions, and Japan is known for its tight immigration policy, so we can rule out savings through a flood of surplus immigrant labor.

            Their mode shares are largely a function of density, but both have high densities most North Americans wouldn’t tolerate. Also, there’s a market for density because of a scarcity of land.

            What about Japan’s and Hong Kong’s car ownership policies and handling of employee welfare? The former can put transit systems on a better playing field, and the latter can help if the burden of health benefits and pensions are not put on the companies’ ledgers.

          • Danny

            The blog post I linked to, showing the difference between public and private wages, is using data from a regular report from the Bureau of Labor Statistics. If you think those stats are worthless, you can take it up with the BLS as well. They are relevant because they reveal one reason why publicly operated transit agencies perform so poorly.

            But it is just one reason.

            Union rules are a huge burden, in some cases tripling the needed cost per operating hour. This isn’t just staffing level requirements. It is sick day rules, vacation rules, and working hours. Inflexible working hour requirements require hiring many more operators than is needed, driving up wages in addition to costs. Capital purchasing costs increase operating costs as well, which is most visible in the faster depreciating bus fleets.

            My point is this: 1) Wages/Benefits are out of hand in the US, contributing to low FRRs…in addition to a myriad of other costs that are seemingly exclusive to our government culture. 2) If private companies aren’t subject to the excessive costs of government operation, then profitability, while not a given, is MUCH more likely.

            The NYC comparison was just to show that wages aren’t everything. It is still an appropriate comparison even when discussing them in terms of density which isn’t all that different. Ridership levels, mode share, frequencies…also very similar. Even if high wages stayed, the subway system could still be operated quite profitably.

  3. ant6n

    One reason why the MTR and other Asian transit agencies are profitable is probably density – good density near any rail throughout the whole metropolitan region.

    The question is to what extend the MTR is profitable because its real estate branch can help bring up density.

    • Danny

      They aren’t creating demand any more than any other real estate developer. They just know that increased land values accompany their transit developments and they use their inside knowledge of future expansions as an opportunity to be the first mover on the real estate.

      If there is any advantage to the MTR’s situation compared to those found in the US, it comes in the form of more liberal laws regarding development.

      • ant6n

        Well, presumably their developments near MTR stations are dense, which is something they can control. I am not sure many American transit agencies have much control against transit-adjacent low density or parking (although some deliberately build parking…).

        • Danny

          It is true that they can control the density level near their stations. But density level has far more to do with demand than it does with the choice of a developer. If MTR built skyscrapers with room for 3000 residents right next to their station, but only 20 people wanted to live there, the development density doesn’t matter.

          My previous point is this though: If 3000 people want to live right next to the transit station, but only 20 people do, the first culprit is likely zoning/development laws. And that is something that a real estate division can’t remedy.

          • ant6n

            The private company independent of the transit agency might decide (possibly even wrongfully) that they will get the highest margin by developing detached single family housing; the MTR might decide that building the skyscrapers will not maximize their profit on the real estate, but it will make their transit system profitable throughout the longterm.

            The MTR might have the clout to get zoning exceptions to build without any parking.

            I’m saying there might be synergy effects going on for the MTR, which help making the transit operations profitable. And that without their real estate portion of their business, station-adjacent development may not be maximized for the best transit outcomes.

          • Danny

            Well there are an awful lot of “might”s in your conjecture, but sure they “might” be possible. The real question is probability.

            I don’t think your situation is very probable at all, and historical precedence in economic planning would suggest that the effect goes in the opposite direction.

            There are several reasons for this.
            1) There are established best practices for market research in property development, and they are mature enough to be used by nearly all property developers.

            2) The developers that are most likely to win a bid on highly valuable land near a brand new rapid transit station are most likely to be very experienced developers, and very unlikely to deviate from marketing best practices.

            3) Those best practices for market research uniformly lead development criteria toward uses that maximize the capabilities of the land and environment area. There aren’t any tried and true best practices that would lead a developer to place a property next to a transit station that would not maximize the use of the transit station…unless there were a significantly greater-value local asset in the same location. It would be difficult to conceive of such an asset, although I think a first order international link for a fiber optic Point-Of-Presence might do the trick, rare as they are.

            4) The ability to plan economic outcomes has historically proven to have diseconomies of scale. While it might be possible for developers to get pretty close to the efficiency of a very decentralized market, I can’t imagine them actually doing any better. Never in the history of city planning or in the history of megadevelopers have we ever been able to design an environment as effective at maximizing transit use as we did in the midst of the industrial revolution with its lack of planning and large real estate developers.

            The fourth point is the reason that the effect could go in the opposite direction. MTR could be building mistakes that other developers wouldn’t make, much in the same way that BART tried to maximize transit use in Pleasanton by building parking lots.

        • Danny

          Oh, and the parking lot thing can get ridiculous sometimes. There is a BART station in Pleasanton CA that has several mid rise apartments surrounding the station, with a bunch of 3 story but closely packed apartments for about a mile radius after that. But every single person that walks to that station has to walk through a gigantic flat BART parking lot to get to the station.

          Parking lots are for transit systems/agencies that have no faith in the value they provide.

          • ant6n

            “Parking lots are for transit systems/agencies that have no faith in the value they provide.”

            Nice way of putting it.

          • Jym Dyer

            @Danny – BART was designed to coenable white-flight suburban sprawl, which is why it provided tons of free parking for decades. Even now their notion of “transit-oriented development” means a long walkway from a freeway median and past a gigantic parking garage to a mall.

          • Wad

            I took a photo of the TOD Danny describes when I was there.

            Transit oriented development near Dublin/Pleasanton BART station

            This would be the closest complex to the actual station. The full lot of cars should give some idea as to the distance of the homes to the station.

            Riders from the TOD may need to take a bus, but the local bus service in the Dublin/Pleasanton area is awful. Hourly service or worse is the norm.

          • Miles Bader

            @Wad — otoh, it looks like they’ve got lots of nice open space right next to the station they can develop … :]

          • Wad

            @Miles, there’s land a plenty but the Bay Area makes it notoriously difficult to get any kind of project, good or bad, done. That’s partly why the San Joaquin Valley is now part of the Bay Area’s commute shed. Not too far from here, there’s a commuter train that runs from Stockton to San Jose, and there are commuter buses from Stockton and Modesto that serve this BART station.

            This is one project that shows how to do Transit Oriented Development wrong.

            The term itself is a whole made up of three interconnected parts. The development is there. There is transit for those who can acclimate themselves to trips within the constellation of BART stations, but who wants to rely on what passes for bus service provided by County Connection or Wheels? (I got to BART via a County Connection route from San Ramon. There were 5 passengers the whole way, with a 40-foot bus on the route.)

            Most of all, though, there’s nothing that orients the transit with the development. Those homes are separated by an asphalt moat from the BART station, and living in one of those units does not make transit, walking or biking any more attractive since the rest of the area is so pedestrian-hostile.

            If you see development and transit, and you have to say to yourself that something doesn’t add up, then the orientation fails.

  4. Krzys

    Since I live in HK I may add a few points to this discussion:
    1) MTR is majority owned by HK Government – it was fully owned and it was later partly listed to raise money for more development of the network.
    2) MTR typically buys land leasehold from Government (all land in HK is owned by Government which sells it as 99-year leases). They pay market rates but not through competitive auction (i.e. other developers are normally not allowed in and sometimes do JV with MTR to get into projects) – the price they pay for land reflects market value BEFORE station development – hence MTR then captures increase in real estate value due to building of the station and not just the normal real estate profit. This is probably the biggest difference from the west where increase in value of real estate due to public transit projects accrues to private owners and not to the public. In HK it is captured by MTR subsidizing the economics of building new lines. Btw, MTR uses its stations as ad platforms to promote new developments.
    3) As to costs – MTR is operated to minimize the operating cost – While unions do exist, collective bargaining is not allowed. They pay market rates for labor – i.e. cleaners will be a fraction of US rates, while for mechanics might be similar. As a for profit entity they are also very good in controlling construction costs.
    4) Revenues – the ticket prices are regulated and need to be approved by Government – there is annual fight in HK – last year MTR was actually denied a rate increase.
    5) Car ownership – it is not encouraged. There are zoning rules that BAN excessive parking lots – I recall that there is regulation that sets a maximum number of spots for a development – for example max 20 spots for 100 apartments (not sure but likely to be around this number) – hence monthly parking is typically US$200-500/mth. There is 100% tax on new cars, gas is $8/gallon, annual registration fee is US$1,000 – all tunnels have tolls – around US$4-5/crossing.

  5. Krzys

    Two more points I forgot to make on operation of MTR that might be relevant:
    1) They are extremely good at interconnecting lines – minimizing walking distance – for example it is common that 2 lines will stop on opposite sides of a platform meaning you can change a line by walking across the platform – making it very convenient to cross HK by transit
    2) Where labor rules probably matter most is in train frequency variability – they run 30-second frequency on main lines during peak (morning and afternoon) and say 5 minute frequency off-peak. That might be difficult to replicate in US given work rules – not sure if you can have part time workforce on such scale which I suspect is what MTR does. They maximise profit but also offer convenience.

    As a side point – the fares are actually quite low – around $0.50-$1.5 per ride.

  6. Pingback: The Option of Profitable Transit | Pedestrian Observations
  7. Pingback: "Transit" Doesn't Lose Money. State Government Kneecaps PA's Local Transit Authorities. - Keystone Politics
  8. Pingback: Meme Weeding: Land Value Capture | Pedestrian Observations

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.