Updated February 6, 2012 at 11:45pm:  The Chong report is now available online (linked below) together with a report from KPMG on financing the Sheppard subway.  Large chunks of the KPMG report are reproduced in the Chong report.

Original post from Feb 5 follows:

Gordon Chong’s report

Toronto Transit Back on Track
Sheppard Subway Development and Financing Study

will be released sometime this week, but a copy has already found its way to me.  This report was commissioned by Mayor Ford to explore the viability of his proposal, as stated in the purported Memorandum of Understanding with Queen’s Park, for the City to go it alone on the Sheppard Subway project.  The report is close to 200 pages long including its appendices, and I am not going to review it in a single article.

There is no question Chong’s mandate was to substantiate the need for and viability of the subway line, and to that end his report is coloured with sections intended to denigrate LRT alternatives and the political process that led to the Transit City proposal.  I will turn to this material in due course, but any decision on the subway project must stand on its own.  History is worth reviewing insofar as it provides technical background and shows the evolution of transit planning in Toronto.  Fighting old battles may score political points, but the subway must be justified on its own merits.

The cost and financing model are central to the thesis that the Sheppard line, and by extension a network of subways, is an appropriate goal, indeed the only goal, for Toronto.  The common wisdom is that “everybody wants subways”, but as with many aspects of public spending, what people want is not always what they will get.  Recent events in Toronto’s budget process are littered with lectures by Ford’s followers about fiscal responsibility and the need to make do with less.  We are told that the city and its taxpayers cannot afford to pay more.  We must, therefore, examine claims that major new public works are affordable with suspicion.

What is the estimated cost of any new project?  Are the numbers we are using credible?  Are subways actually cheaper than we have been told, and could a lower cost bring them within financial constraints of potential revenue?  Are public agencies the appropriate developers of such projects?  Are their costs (historical and projected) greater than might be achieved through another delivery mechanism?

While it may be a common Toronto sport to poke holes in TTC budgets, management practices and operations (a not uncommon thread on this blog), such criticism must be backed with a standard of accuracy and care.

How Much Will the Sheppard Subway Cost?

A central premise in this debate is that the TTC, and by implication the public sector generally, is unable to deliver this project at a reasonable cost.  An oft-quoted figure puts the TTC’s estimate of the project at $4.7-billion while an estimate from Metrolinx sits at $3.7-billion.  These numbers first appear in Table 2 in the Executive Summary, and they are routinely repeated as gospel.  One must read all the way down to page 147 and Table 38 to see the details. Here we see component costs that are generally higher for the TTC estimate than the Metrolinx one, but the details reveal that the billion dollar difference is not all that it seems.

Metrolinx estimates the cost of maintenance facilities at $138-million based on a per-car value of $2.66m.  A footnote on the table clearly states that the TTC estimate of $500-million is based on a new facility larger than is needed to hold just the fleet for Sheppard.  Why such a big difference?  Metrolinx assumes an expansion of the yard at Wilson and therefore a marginal increase in system capacity whereas the TTC makes provision for future fleet growth for demand and for system extension.

Wilson Yard has a looming problem with its size because there are limits on how fast trains can be pushed out for service buildup in the AM peak.  Already there is discussion of shortening the hours of subway service to retain an overnight maintenance window between the end of one day’s operations and the start-up of the next.  TTC plans include proposal for an underground storage yard north of Finch Station and, eventually, to a new carhouse somewhere in York Region.  We cannot simply keep stuffing more and more trains into Wilson.

This aspect of the cost difference cannot be counted as a penalty against the TTC because it addresses a completely different model of what would be built (and why), not some inherent flaw or inflation in TTC costing.

“Operating Systems” covers a range of items listed in the comparison.  For this, the TTC’s value is 4.5 times the Metrolinx value ($329m vs $73m).  This amount cannot be explained simply by claiming inefficiency at the TTC, and it is wildly out of scale with the differences in other items.  At the very least, anyone purporting to compare estimates would flag such a difference and explain it in their report rather than simply using the numbers without question.

“Contingency” is a catch-all allowance in any project budget to allow for unexpected events and costs during construction.  Both the TTC and Metrolinx estimates allow about 26% over and above the component costs, and with the TTC’s costs being higher, so is the contingency in their estimate.

Sales tax is included in the TTC estimates, but it is not in the Metrolinx version.  This shows up by virtue of an HST Rebate in the TTC section of the table which has no equivalent on the Metrolinx side.  The HST is included in the component costs including the contingency factor, and the TTC unit costs are not presented on an equal, untaxed footing with the Metrolinx costs.  Again, this sort of adjustment is a basic requirement of financial analysis, but it is absent from Table 38.

There are likely other areas where differences between Metrolinx and TTC figures would bear scrutiny, but as the TTC numbers are not detailed here, nor are the assumptions on which they are based, it is impossible to dig further.

Taking what we can see into account covers about three quarters of the difference between the Metrolinx and TTC estimates.  Before we can believe the Metrolinx $3.7b estimate, the inconsistencies with the TTC numbers must be explained.  Both values may be legitimate given the underlying assumptions used in each case, but these are demonstrably different.  Saying that there is a $1-billion spread between the two is an apples-to-oranges argument.

Here, it suits Gordon Chong’s thesis that the TTC is an inherently poor steward of public funds and that the project could be delivered at lower cost through another agency or mechanism.

A January 2008 Metrolinx report on a study tour to the United Kingdom and Spain is included as an appendix to the Chong report.  Even a cursory reading of this document shows that there are significant differences in the environment in which large-scale projects were undertaken in these jurisdictions compared with Toronto.  A major source of savings lies in the scale and continuity of construction projects, a general agreement that the projects should go forward (possibly with less up-front review such as our Environmental Assessments), and a regulatory environment that reduces contention between proponent agencies and the companies actually building their projects.  (Buried in the report, by the way, are references to “Tram” (LRT) components of the Madrid system which are considerably cheaper per kilometre than their subways.)

The degree to which each difference between the European cities and Toronto contributes to differences in costs is not explored, and yet this is essential to any comparison.  The scale of their projects and longevity of their construction plans are not directly transferable to a single Toronto subway extension.  It is not enough to say “look at how cheaply Madrid builds subways” without also understanding why they can do it.

In my next article I will turn to the question of how we will pay for this project.