How to Build a Subway — Let Me Count the Ways

As everyone knows, the public sector wastes huge amounts of money on overblown, out-of-control projects, or at least that’s the prevailing view in some quarters. One major side-effect of Ottawa’s participation in funding the York subway extension (or TYSSE: Toronto-York Spadina Subway Extension), in an insistence on value-for-money and the maximum participation by the private sector.

Leaving aside that the TTC doesn’t actually build anything itself, and contracts out a vast amount of the design/engineering work, there is a perception that (a) subway construction projects could be better run and (b) there is (even more) opportunity for private sector participation.

A report on this week’s TTC agenda discusses the various options for delivering this project and includes some revealing information about the pros and cons of various approaches. I will leave it to those who are interested to read the full text, but a few things caught my eye.

First off, there are many ways and degrees in which a private sector company or consortium can be involved in large projects like this all the way from complete design, finance, construction and operation down to a basic contractor who bids on a job, builds a box and leaves.

In the review of options, the variants where operation rested with a private entity were rejected outright because the TYSSE is part of an existing, operating, public-sector subway line. Experience elsewhere suggests that private operation tends to occur only when a new, free-standing line is built such as the Canada line in Vancouver.

Financing options are not discussed at length, but this issue always turns on the question of whether a private partner can provide capital at a lower cost than a public agency. This also involves some creative accounting. A privately owned line (and its associated capital debt) does not appear on the public books. This scheme is commonly used to hide debts (think Enron), although generally accepted accounting practices for governments make this more difficult to pull off. Even if a government is not technically exposed to the debt, the last thing any (well almost any) provincial government will do is to let a subway line close because its owner is bankrupt.

We have heard a lot lately about borrowing from large investment pools such as pension funds. Whether this is done on a government basis or by an arm’s length agency, somehow the interest and debt must be paid. Either this is a direct charge against current operations, or it is transferred to a government through a subsidy arrangement (no doubt with an appropriate management fee).

As for construction, the difference between the two main options depends on whether the TTC designs it all and contractors just build it, or if the TTC says “build me a subway station” and the contractors design to a general set of specs and deliver a finished product. A working group from the construction industry reported that their preference is to leave the design to the TTC for contracts under $100-million, or where there will be multiple contractors (possibly including the TTC itself) onsite. This relieves the contractor of having to manage (and assume risk of) portions of the work not under his control.

For large contracts, it may be worth a contractor’s while to bring the design work in house provided that the job is fairly generic and does not require skills in special systems peculiar to transit. For example, building an empty box that will become a subway station or tunnel structure is a fairly straightforward task while design and co-ordination of the many subsystems fitted within the structure are complex and outside of their regular scope of work.

One point not mentioned here is that the industry’s skill base depends strongly on what they do most of the time. An important observation about Madrid in recent studies is that their continuous program of system expansion allows the industry to develop expertise and continuity of staff that would not otherwise be possible. In Toronto, there are only so many subways to go around among the major players, and we don’t build them very often.

Finally, there is a breakdown of the TYSSE’s estimated cost — $2.09-billion in 2006 dollars, or $2.633-billion assuming completion by 2015. This date could slip if various governments spend time squabbling about whether the project passes a private sector sniff test, adding to the cost.

For all you readers who have wanted a subway station of your very own, we see station costs of $73-to-$100-million depending on the complexity. Of particular note is a 26% contingency. This rather generous $400-million slush fund (a cool half-billion with inflation) will allow considerable overruns while keeping the project “on budget”. At this point, I will be generous and hope that this is for “things we haven’t thought of in the preliminary design”, but at some point this needs to be nailed down. I doubt we will ever know how much contingency everyone builds into their estimates (public or private partner) and yet there is probably more money on the table in this one line than in any savings, real or imaginary, from increased private sector participation.

I should mention the rolling stock. The estimate shows 56 cars, although it is physically impossible to buy them in this quantity as they now come in married sets of 6. Even with these cars, the TTC will not have enough “Toronto Rockets” to completely replace the existing fleet on the Yonge-University-Spadina line, and another car order will be needed. Thunder Bay will be churning out cars for years.

The TYSSE will be interesting if, for nothing else, showing us whether there is money to be saved on subway projects with greater private sector know-how at work. Alas, we won’t have an answer to the question for about 7 years, and we will have spent $2.5-billion finding out.