Prophylactic planning: Rapid transit in the era of cost disease.

A couple years back, Slate Star Codex published a widely-discussed article about what the author called “cost disease” – not the disease identified and explained by William Baumol, but a distinct and poorly-understood condition where prices in certain industries rise at a rate much faster than inflation.

This condition is particularly acute in the case of rapid transit infrastructure.

Since Slate Star Codex had limited itself to American data, I wondered whether Canadian rapid transit projects might be suffering from the same ailment. It’s difficult to compare infrastructure costs over time, as no two projects are identical. But looking at projects in Toronto and Vancouver, I found a noticeable upward tick since the 1990s, suggesting the presence of cost disease in Canada.

(Click image for data and sources.)

(To emphasize, these are per-kilometre, inflation-adjusted prices that appear to be rising.)

Accepting that this wasn’t just a mirage created by staring too hard at a limited data set, I went on to argue that we should hurry up and start digging subways now, while it’s only shockingly expensive, instead of putting it off into the future, when it’s likely to be cripplingly so.

(But one could also argue that we’ve already missed our window, and that instead of blowing money on huge infrastructure projects that are unlikely to prove cost-effective, we should invest in a better bus network, or fleets of autonomous cars, or the latest Elon Musk fever dream, or what-have-you.)

For now, let’s assume what I can’t prove: that at least some rapid transit megaprojects are still worth the inflated price. If we want that to continue being true in the future, we need to smarten up the way we plan our rapid transit network.

In my earlier essay I mentioned four factors contributing to the spread of cost disease:

1. In the big cities where rapid transit gets built, land values have been rising at a rate far exceeding inflation, leading to higher property acquisition costs.

(Click image for data and sources.)

2. The accumulation of buildings, pipes, and wires – what I called infrastructure clutter – around potential rapid transit corridors makes construction ever more complicated.

3. We’re more safety-conscious these days, which means we build more slowly, carefully, and expensively than they did in the rough-and-ready 20th century.

4. We’re willing to spend more to protect our natural and architectural heritage from the negative side effects of construction, and to ensure full access for handicapped people.

Since then, other factors have occurred to me:

5. Population growth means there are ever more residents and business owners to object to the inconvenience of construction, the noise of passing trains, lowlife transit riders invading their fancy neighbourhoods, and other blights of public transportation. The internet has made it cheaper and easier for obstructionists to organize and demand pricey compromises.

6. We’ve gradually used up or sold off the most promising rights-of-way that came available in last century’s shift from rail to road as the primary transportation mode, leaving us no choice but to dig tunnels.

7. Related to point 6, for political reasons earlier city planners prioritized easy-(and therefore cheap)-to-build projects, leaving the most challenging (and pricey) pieces of the network to be dealt with by future generations – i.e., us. Look at Toronto, which has spent the last half-century pushing its subway ever further into the lightly-built suburbs, ignoring the pressing need for a new line downtown.

8. Related to points 6 and 7, modern planners may be more choosy than their predecessors about where to place their routes. Vancouver’s original 1986 Expo Line was built on an abandoned rail right-of-way four or five blocks from Kingsway, the busy road it parallels. Thirty years later, many Kingsway commuters continue to take the bus. Compare the current plan to extend the Millennium Line, which disregards the out-of-service rail line a few blocks to the north for a brand-new tunnel directly under Broadway.

Not coincidentally, the per-kilometre cost of the Millennium tunnel is expected to be over five times higher than the Expo Line.

Millennium Line extension (6 stations, 5.5 km.) estimated at $2.57 billion (2018 Canadian dollars), completion date 2025. Source.

Of the above list – which I fear is far from comprehensive – most of the factors are driven by population growth, which means they’ll only get worse.

We could conceivably save money by skimping on factors 3 and 4 – by building more recklessly, noisily, and uglily. As an example, Surrey mayor Doug McCallum has suggested that the proposed Langley extension of the Expo Line could be built more cheaply if crews worked round-the-clock. Maybe so, but I suspect they’d only run up against factor 5: angry opposition from residents living near the construction zone.

This doesn’t mean we’re helpless against cost disease. Since the choices of today’s planners determine where future residents will live, we can predict where demand for transit should grow. And by thinking about why costs go up, we can predict which routes will be most expensive to build in the future and should therefore be prioritized, and which can be affordably postponed.

Some basic principles of cost prophylaxis. All else being equal…

1. Extending existing lines is cheaper than building new ones.

2. Building all in one go is cheaper than building in fits and starts.

3. Building where there’s nothing is cheaper than building where there’s something.

4. Coordinating with other infrastructure projects lowers costs. (Toronto’s 1966 Bloor-Danforth subway saved a bundle thanks to the farsighted inclusion of a lower deck on the Bloor viaduct – fifty years earlier.)

…But as I stated in my earlier post, planners are already well aware of the above strategies. If they fail to implement them, it’s only because it’s difficult: to anticipate which projects future politicians might prioritize; to coordinate between multiple agencies, levels of government, and private entities; to resist political pressure to cut corners and push expenses into the future.

I doubt I can offer any insights that haven’t occurred already to the experts. But I think they might give points 5 and 6 a little more weight:

5. Every project will stir up opposition. But well-off residents, because they rely on transit less, and because they tend to own their homes, have less to gain and more to lose from transit expansion. Their money, education, and well-groomed spokespeople make them more effective obstructionists. Therefore, try to put rapid transit into a neighbourhood before it fills up with yuppies.

6. If you’re hoping to preserve a corridor for future use, you’d be better off building now: to prevent future politicians from selling it off piecemeal; to prevent infrastructure clutter along the route; and to prevent residents from getting emotionally attached to it in its virgin state.

(Here I’m thinking of Vancouver’s Arbutus corridor – the rail line shown on the map above – stretches of which were colonized by gardeners from neighbouring properties during its years of disuse. When the city, after acquiring it last year from CP Rail after much haggling, tried to convert it to a paved bike path, nearby residents protested the despoliation of what they viewed as their private rambling grounds. I can just see the outcry in the future when the city attempts to pursue its vision of running a streetcar down the line…)

Vancouver’s new mayor, Kennedy Stewart, has been advocating for the Millennium Line extension – currently set to terminate at Arbutus Street – to continue down Broadway and West 10th Avenue all the way to UBC.

…Which, don’t get me wrong, would be terrific: I’d use it a lot.

However, looking at it from a cost-prophylactic perspective:

1. The corridor is already densely built-up: infrastructure clutter is therefore unlikely to worsen.

2. The properties for future stations, if they haven’t already, can be acquired now, and held onto until needed. (Astonishingly, if this Daily Hive article is to be believed, as late as March of last year the site of the Millennium Line’s planned Broadway & Granville Street station hadn’t yet been sewn up.)

3. The West Side can’t be any further yuppified. The locals have already secured all the compromises they’re likely to dream up: instead of a noisy elevated train like their poor cousins in East Van, they’re getting a tunnel; and not a cut-and-cover tunnel, like the one that enraged Cambie Street merchants during the construction of the Canada Line a decade back, but a fully bored tunnel.

If the Millennium Line can be extended all the way to UBC in a single go, then it absolutely should: it would be far more cost-effective.

If (as seems likely) it can’t, this corridor presents only a moderate inflationary risk. The UBC extension can affordably be postponed.

Meanwhile, down at the southern end of the Arbutus corridor the homely neighbourhood of Marpole, containing one of the few concentrations of affordable apartments still left within the city’s borders, is seeing a surge of new construction. Would it make sense to move the Arbutus streetcar plan forward, while there are still a few students and working class people living along its route who might benefit from it? [1]

The Arbutus corridor is wide enough that property acquisition costs for future streetcar stops should be minimal. But have they figured out yet how to bridge the 2-kilometre distance between the end of the corridor and Marine Drive station on the Canada Line? That could be an expensive gap to fill.

And if the city is seriously contemplating rapid transit along 41st Avenue, even in the “extremely long term”, they’d better start locking things down now. With the lower-middle-class enclave near Joyce-Collingwood station doomed by encroaching condo towers, and the futuristic “micro city” about to begin construction at Oakridge, 41st Ave. is extremely susceptible to cost disease.

I’d nominate Marine Drive and East Hastings as two other yuppifying streets where rapid transit would make sense, and where the risk of cost disease is acute. I hope to have more to say about the latter corridor in a follow-up post.

[Update, Dec. 25 2019: As promised above, The Hastings SkyTrain alternate reality.]

M.

1. This isn’t an endorsement of the Arbutus streetcar scheme, by the way. To me it seems as misbegotten as the now-abandoned Surrey LRT plan I discussed last month: why spend a bajillion dollars laying rail for “rapid transit” that’s not significantly faster than a bus? You might as well make a bigger up-front investment in a tunnel or elevated tracks and enjoy the benefits of higher speed and driverless operation.

If I’d been benevolent dictator, on the day the Arbutus corridor was acquired I would have turned it into a trench two SkyTrain cars wide, put a roof over it, and left the resultant tunnel until it was needed. Instead, the city elected to fancy up the corridor with walking paths and flowerbeds from one end to the other, guaranteeing an infestation of sign-waving old ladies in sunhats whenever they attempt to alter it.