We all think we know the Canada Line inside out and we take it for granted. But it turns out others don’t.
This American organization asked me to write about it, which made me realize, as I was doing my research, how preoccupied we had been with other parts of the story during the years it was being debated and built: the crazy back and forth votes, the massive disruption on Cambie Street, the torquing of the region’s transit planning because of the Olympics, and more.
In the end, none of us looked that closely at the public-private partnership and the details of how it worked. This time, I got to and learned some new things from Jane Bird, the lawyer who oversaw it (now in London overseeing renos of Canada House), and a Toronto prof who has looked closely at P3s in Canada and whether they really save all the money that proponents like to claim they do. (Seems like we do P3s in a very cautious, Canadian way, if I’m reading his paper right.)
Here’s the series I did, most of which also appeared in Atlantic Cities.
Matti (the Toronto prof) sent me this additional note after it was published. (Warning: Only transit nerds may understand/enjoy this.)
An interesting feature of this project is that the design put forward by the winning bidder was estimated to deliver considerably higher ridership levels and thus more revenue (148M) for Translink over the life of the project than the comparable design that would have been built traditionally. If you look at the value for money report in further detail on the bottom of page 17 and at the top of page 21, you get a sense of the innovations that were being brought forward by the PPP concessionaire, and how these were expected to impact on ridership levels.
Some of these innovations have proven beneficial, while others could potentially be short or longer-term hindrances. You may also consider how these innovations would have saved money during construction for the PPP as compared to a traditional build project, even though the PPP is reported to have higher base construction costs. The key innovations in the PPP were reportedly:
– higher proposed midday train frequencies which would attract around a third more riders during the day, and contributed to higher revenue projections for the PPP. The report acknowledges that this change in service levels could have been made on a traditionally built project too (see bottom of P17).
– more accessible underground station designs that were closer to the surface and had fewer stairs, which would decrease total travel time and enhance convenience. This may be related to the use of cut and cover construction, though it is not explicitly stated in the report.
– single tracking in Richmond (this likely would have lowered construction costs and possibly has community benefits as the guideway is thinner and less imposing on the street below; it also limits train frequencies on that segment of the line which will effect long-term capacity)
– the elimination of a station at the airport (faster travel times may attract more riders, lowers overall construction costs)
Finally, even after factoring in the revenue from higher ridership that the PPP would deliver, the PPP project was still estimated to have a total lifecycle cost that was 141M above the comparable traditionally procured project. It was only after considering risk on the project, and thus adding a risk premium to the traditional cost estimate, that the report shows that the PPP was likely to be the option that provided the best value. The risk premium added on this project was not particularly large compared to some of the ones I?ve seen on projects here in Ontario, and the risk of significant cost escalations is very real and does need to be factored into assessments of the best procurement option. Nevertheless it still had a defining role in the outcome of the value for money assessment.
So what to make of all this? To me it says that a premium is being paid through the PPP model to realize cost and performance predictability. Private sector concessionaires can bring forward innovative design and service delivery plans that improve the quality of service for users, but these need to be closely vetted to understand their broader community impacts. And risk transfer to achieve cost certainty continues to be a key variable that drives the value for money of PPPs. However this risk transfer comes at a significant cost for government and as such we need to continue to examine whether there are other less expensive ways for government to manage rather than necessarily transfer project risk.
26 responses so far ↓
1 tf // Mar 13, 2014 at 6:32 pm
The last sentence – “However this risk transfer comes at a significant cost for government and as such we need to continue to examine whether there are other less expensive ways for government to manage rather than necessarily transfer project risk.”
Thanks Frances!
2 Silly Season // Mar 14, 2014 at 2:23 am
I am happy to see that Jane Bird has a new career that suits her skills set.
‘Meet the new boss,
Same as the old boss.’
3 Rico // Mar 14, 2014 at 5:30 am
Two of the ‘benifits’ are clearly bullshit and are major negatives. Frequency is clearly more constrained in this model compared to the other Skytrain lines and the single track section in Richmond will probably require significant double tracking in less than 20 years. The elimination of another station at the airport probably is a benifit as I doubt another airport station would generate any extra ridership but the elimination of stations at 33rd and 57th hurts ridership in the long term. Cut and cover does make a significant savings and make stations more accessible….which leads to why was it not considered for the ‘traditional’ option? Next up and not covered why would anyone include financing in a P3? Government interest rates are way below what Translink is now forced to pay for this. Stupid.
4 Randy Chatterjee // Mar 15, 2014 at 1:35 pm
What galls me still about ALL of the media reporting on the RAV Canada Line from 2005 to today is the complete lack of information on cost. The Translink Board in fact never publicly authorized the full cost of the capital allocation for the construction of the line. This is evidence of contracting fraud, and civil and criminal investigations should have been launched years ago.
Three words: Provincial Performance Payments. Those three words are just about the only three left unredacted in several FOIs from 2006. These “performance payments” were and are an explicit backdoor, hidden kickback to the P3 which obligates the government to pay the consortium well over a billion dollars more than the maximum authorized contractual amount for the construction of the line. These payments are buried in the 35-year “operating” agreement and payments, which currently have Translink paying well over $120 million per year to keep the line going. It obviously costs nowhere near that amount actually to run the RAV line, and the vast majority of this payment is an unapproved construction kickback.
Translink is near bankruptcy and cutting bus routes everywhere for one reason only: the RV Line. This piece of P3 infrastructure is sucking it, and the taxpayer, dry. Illegally.
5 Jon Petrie // Mar 15, 2014 at 4:51 pm
Restating to some degree Randy above in different language:
The annual fee paid to the consortium to run the line is higher than what would have been paid if running the line and construction had been put out to bid separeately. (Think of getting a “free” phone when signing a two year lease with a service provider.)
Politically it was desirable to have the line be perceived to have cost under two billion.
Also of course PPP construction does not show up on the government’s books thus reducing the appearance of deficit spending (or increasing apparent surpluses).
6 Jay // Mar 15, 2014 at 5:20 pm
@5
I always thought the Canada Line was almost a miracle compared to other transit projects in Canada. I wonder what the actual cost of the C Line is.
The Evergreen Line, which I presume is more upfront about cost, is still a lot less expensive than Calgary’s West LRT. Calgary’s line runs mostly at grade, while the Evergreen Line is elevated with a short tunnel section. The price difference is 170 million/km vs 127 million/km. Vancouver, it appears anyways, does rapid transit far better than any Canadian city.
7 Silly Season // Mar 15, 2014 at 5:23 pm
@Jay and @Randy Chatterjee…as if by magic, the Sun’s Don Cayo brings up the same questions, just a few days ago.
Perhaps Jane Bird could help answer them?
http://www.vancouversun.com/columnists/index.html
8 Silly Season // Mar 15, 2014 at 5:25 pm
Sorry! Listed the Sun’s Columnists page…
Here’s the Cayo story link on P3’s and costs.
http://www.vancouversun.com/business/Cayo+numbers+need+closer+look/9612284/story.html
9 Tessa // Mar 16, 2014 at 3:51 am
@Rico: Agreed. I really don’t see any benefit for this arrangement, especially when it means Translink has far less control over how this line is now operated. Let’s hope all future expansions are simple Skytrain extensions like the Evergreen or else they’re likely to get sucked up by the P3 model.
10 Randy Chatterjee // Mar 17, 2014 at 7:33 pm
Jay #6: There is no final and comprehensive contract for the Evergreen Line; there has never been enough money to fully tender it. So, any price comparison with the West C Train–also a much more urban route on an all-new ROW–isn’t possible or fair anyway. While the Expo line could be argued to run an operating surplus, the Millennium drags our Skytrain transit bundle down. I’d say Calgary, with its wind-powered C-Train, beats us hands down in the transit department, and there are hundreds of cities worldwide ahead of us on transit access, reliability, schedule, cost, speed, and public accountability.
11 Voony // Mar 17, 2014 at 9:16 pm
Randy@10
First,
Not sure why you say
There is no final and comprehensive contract for the Evergreen Line;
when the Province awarded a $900M contract to the EGRT consortium, to build the whole Evergreen line….
Secondly
Translink (or its subsidiary BCRTC) doesn’t publish separate numbers for the Expo and Millenium line.
The sytrain network operated by BCRTC (Millenium + expo line) break-even as far as operating cost are considered.
numbers are compiled here:
http://voony.wordpress.com/2011/11/28/some-translink-statistic/
12 Jay // Mar 17, 2014 at 11:00 pm
I’m always skeptical, but I’m inclined to believe the 1.4 billion dollar figure for the Evergreen Line.
I’m not sure why other cities continue to build with LRT. Calgary – $170 million/km; Ottawa – $170 million/km; Edmonton – $228 million/km. These are at grade systems, with some tunnel sections that are no more technically difficult to build than Evergreen Line at $127 million/km.
As far as performance goes, because it’s automated, Skytrain has much better frequency than LRT, and it’s average speed is much better than LRT. Calgary’s C-Train’s average speed is 35 km/h vs Skytrain at 45 km/h. Both systems have 1.5 km station spacing.
When the Evergreen Line is complete, Vancouver will have the longest rapid transit system in Canada. We must be getting pretty good bang for our buck when you consider Toronto is over twice the size of Vancouver.
13 Dan Cooper // Mar 18, 2014 at 3:10 am
Speaking to the question of accessibility: I have never seen as badly-designed stations for a person who cannot use stairs as those at 41st and 49th, at least not in a system built after the 1950s (or most built after the 1920s). Three separate elevators to go from the platform to ground level, or the other direction? A sick joke.
14 Randy Chatterjee // Mar 18, 2014 at 9:06 am
Dan@13: Thank you for pointing out one of the many glaring examples of poor system design for accessibility, a critical measure of success since transit is the only choice for so many differently-abled.
Vonny@11, see Jan@12. Difference in part is the cars. To your second point, due to the ridiculous and confusing overlap between Millennium and Expo, separating their data is complex. However a look at per station boardings and exits shows Expo to be by far the more heavily-used route choice.
Finally, more than 85% of Canada Line ridership is cannibalized and goosed from cancelled and redirected routes, especially the urban and suburban B-lines, hardly a ringing endorsement of its ability to attract new ridership. Many former B-line users have stopped using transit to get downtown due to RAV Canada Line’s inconvenience, inaccessibility, and crowding. Overall the route is sound and speed good, but the financing, durability, accessibility, crowding, and system changes to “adapt” to it are a huge net drag to the overall transit network.
15 Voony // Mar 18, 2014 at 10:14 am
Notice:
Circa 2004 (when ridership study was done),
transit ridership in the corridor was ~40,000 (~15,000 for the 98B + all suburban routes).
This number was a rational used mainly by Canada line opponent, to consider this project as way over-sized.
today, there is close to ~120,000 daily ride on the Canada line…
I notice that the exactly same people whose yesterday were claiming it was over sized, can be seen today explaining the line is undersized.
Again, the net positive effect of the Canada line is difficult to miss on the first graph http://voony.wordpress.com/2011/11/28/some-translink-statistic/ (even thought it is unlinked trips)
16 Rico // Mar 18, 2014 at 11:34 am
Randy Chaterjay,
‘Finally, more than 85% of Canada Line ridership is cannibalized and goosed from cancelled and redirected routes, especially the urban and suburban B-lines, hardly a ringing endorsement of its ability to attract new ridership. Many former B-line users have stopped using transit to get downtown due to RAV Canada Line’s inconvenience, inaccessibility, and crowding.’
And yet ridership is way higher on the corridor than pre Canada Line (as Voony notes probably between 2 and 3 times higher) so obviously a lot more people find it better than the bus services it replaced than those that find it less convenient. Most changes to a network will benefit some users more than others, a change may even be to the detriment of some. It is clear from the usage that the Canada Line positively impacted more users than negatively impacted them (as witnessed by the massive increase in ridership on the corridor).
I also agree with Dan about the accessibility of stations, it is frustrating to take a stroller on a bunch of different elevators to get to the platform. Thankfully most of the stations are fine.
Also of course the Millenium Line is not as well used as the Expo Line. When it was built it went from nowhere to nowhere. With the Evergreen line that will start to change (although it already has with new development at the existing stations). If we finally build Broadway it will be as useful as the Expo line and will get ridership to match.
17 Randy Chatterjee // Mar 18, 2014 at 1:08 pm
Within a year of starting operation, the 98B alone was at 18,000 boardings. (http://www.tc.gc.ca/media/documents/programs/98b_eval_final.pdf)
Now add the previous ridership of the Cambie 15, and part of the Oak 17 and Main 3, plus all of the following cancelled or redirected routes: #488 Garden City, #490 Steveston, #491 One Road, #492 Two Road, #496 Railway, #311 Scottsdale, #351 Crescent Beach, #352 White Rock, #354 White Rock South, #601 South Delta, #602 Tsawwassen, #603 Beach Grove, #604 English Bluff, and #620 Tsawwassen Ferry.
Finally, add all the other later Vancouver bus route redirects to Marine Drive station.
These total far more than 90,000 boardings per day, and many of these are now double-counted due to the forced transfer at the RAV Canada Line.
And if you have any more doubts, look at the mode-shift that 60,000+ new transit riders would have created. It’s simply not there. And what is there are Cambie Street traffic volumes now higher than in 2009. Oak as well.
Again, the RAV Canada Line had a decent alignment, but politics and graft made an unnecessary mess of it. Translink and its ridership are still paying for it through the nose and in lost time and convenience. It is worth noting that average suburban transit times into Vancouver are now higher due to the loss of nearly a dozen B-line and express bus routes.
18 Randy Chatterjee // Mar 18, 2014 at 1:19 pm
And why still is NO ONE talking about the actual final cost of the RAV Canada Line, the subject of this thread: $2.1 billion to start and $120 plus million per year.
And what do we get for this nearly $6 billion project (in nominal dollars)? Water continues to pour into the tunnel guideway at every seam, breaching the seals, eroding the concrete, and rusting the steel. Leaky condo, leaky subway. How much more can BC take from our construction “industry?”
19 Rico // Mar 18, 2014 at 2:40 pm
Randy Chaterjay,
‘Within a year of starting operation, the 98B alone was at 18,000 boardings. (http://www.tc.gc.ca/media/documents/programs/98b_eval_final.pdf)
Now add the previous ridership of the Cambie 15, and part of the Oak 17 and Main 3, plus all of the following cancelled or redirected routes: #488 Garden City, #490 Steveston, #491 One Road, #492 Two Road, #496 Railway, #311 Scottsdale, #351 Crescent Beach, #352 White Rock, #354 White Rock South, #601 South Delta, #602 Tsawwassen, #603 Beach Grove, #604 English Bluff, and #620 Tsawwassen Ferry.
Finally, add all the other later Vancouver bus route redirects to Marine Drive station.
These total far more than 90,000 boardings per day, and many of these are now double-counted due to the forced transfer at the RAV Canada Line.’ I guess all those people who were swearing that the Canada Line would never reach to current 40,000 boardings per day missed those routes….oh wait, I am familiar with most of those routes and WHEN the Canada line opened they had pretty minimal ridership (can’t say without looking up the numbers for the 15 Cambie or 17 Oak or a few others, but still…..). Where do you make up your numbers?
20 Rico // Mar 18, 2014 at 2:45 pm
Randy Chaterjay,
‘And why still is NO ONE talking about the actual final cost of the RAV Canada Line, the subject of this thread: $2.1 billion to start and $120 plus million per year.
And what do we get for this nearly $6 billion project (in nominal dollars)? ‘
I also would like more understanding about the ‘performance’ payments. That said my understanding is they are mainly accelerated debt repayments because financing was foolishly part of the P3 and Translink is trying to pay off the private financing part as fast as possible. Your 6 billion dollar figure makes you sound like Rail for the Valley….
21 Randy Chatterjee // Mar 18, 2014 at 10:03 pm
While all the mudslinging and straw men? No one here argued the Canada Line would not fill up, and I for one took note of the short platforms in 2006, very strange in comparison with Millennium and Expo. This style of argument in no way to adds to public knowledge. Start by reading Steven Rees’s 4yo blog here: http://stephenrees.wordpress.com/2009/08/16/canada-line-subsidy-will-be-felt-for-years-to-come/
Bus ridership figures are hellish to get, or trust, as Rees–who worked at Translink–notes there. However we do know that the 98B alone in 2002–the year after the strike and its first full year in operation–counted for almost half of the total N-S ridership Vonny claims existed in 2009. One bus route out of 14. Patently ludicrous.
On the question of system cost, all in for 35 years, the construction contract was $2.1 billion. That is not at issue, and well-documented. Then, FOIs from Translink over the past three years have shown escalating total annual payments to InTransitBC LLP of $120 million up in 2012 to $127 million. We do not know how these figures are calculated, but they clearly imply $3.7 to $4.9 billion for such payments over the life of the P3 contract, less the real value (undisclosed) of the actual private contribution to the P3 project (<= $600 million), and plus the $2.1 billion. The math is easy, and discounting could be applied if you'd like to suggest a rate. A cost nearing $6 billion is a no-brainer, and I have 5 years as a international development finance manager–including on rail transit projects–to back me up.
Facts, please.
22 Voony // Mar 18, 2014 at 10:37 pm
Ridership
Translink bus ridership are well documented on a route basis
bus 311, 351, 352, 353, 354, 601, 602, 603, 604 and 620, altogether carry no more than 10,000 rider a day.
Notice that bus 620, was not direct to DT even before the Canada line, so shouldn’t count)
I don’t have ridership for defunct, bus routes, 488, 490, 491, 492 and 496, bus I have the 2008-2009 schedule: altogether, it was ~125 day runs of those 5 routes…we can deduce a rider ship of ~5,000 for all those routes…
Inescapable conclusion:
the 98B alone was effectively carrying more people than those 14 buses route altogether.
the bus 15 was carrying 6,000 people before the advent of the Canada line (probably most of them connect to the C line today).
Adding the 98B ridership (which has significantly grew over the route lifetime) you end-up to the ~40,000 daily rider in the corridor number circa 2004,
It was 20,000 people using the bus 17 before the C-line. As a unforeseen, half of them have disappeared (probably using route 16, 25, 41, 49 to C-line), and among the rest some could effectively connect to the C-line.
ridership on the 3 bus has increased, and it is hard to see a significant share of its 22,000 rider connecting directly to the C-line.
So all that doesn’t add up to 90,000 pre-existing rides., 60,0000 could be already a very optimistic projection.
Where come from the 60,000+ new riders?
Transit work like road:
The better the offer is, the more people will use it…and increase their overall number of trip (mobility)…
for that reason, transit never reduce significantly congestion per its own existence
23 Voony // Mar 18, 2014 at 10:59 pm
system cost
The cost has effectively been well documented, …and the crown awarded a $1.4B contract
InTransitBC is assumed to have added $700M to finance the total construction cost estimated $2.1B
The ~ $100M annual payments done by Translink to InTransitBC covers the reimbursement of those $700M financing by InTRansitBC + the operating cost of the line.
One could hope to know more how those thing break down, but the bottom line is that when you consider the total life cost of the project (capital+operating costs), the Canada line cost around ~$4/rider, that is in line with the Skytrain, and will be a better deal than the evergreen line ($6/rider), what could be already considered unbeatable in the LRT world.
24 Rico // Mar 19, 2014 at 6:05 am
Thank you for the facts Voony. From personal experience I knew Randy’s numbers were bizarre but it would have taken me a lot of work to find the correct info.
25 Andrew Browne // Mar 24, 2014 at 2:17 pm
My only contribution to this is:
While the operating agreement no doubt lasts X years (whatever the term is – 30 years?), it is not at all necessarily the case that the accelerated debt repayments also last 30 years. Sure, $100 million times 30 years is $3 billion. But it seems to me that what is happening is that Translink is paying off the private-debt faster than required so as to minimize interest expenses. So perhaps the payments will drop in year 10 from ‘operating + debt’ to just ‘operating’. In year 10 that might look like a drop from $150 million annual payment to $40 million annual payment (or something similar).
26 Andrew Browne // Mar 24, 2014 at 2:27 pm
And to be clear the ‘debt’ I am referring to is the $700 million contributed by the P3 partner, inTransit BC (SNC-Lavalin et al).
Incidentally I believe the P3 money actually came from two public pension funds. BCIMC if I’m not mistaken? So we’re really just paying ourselves anyway. It’s a P3 in that project construction risk has been transferred to the private partner… but the private partner was financed by public money. Meaning the risk is still ultimately public!
Very odd knots we twist ourselves in to essentially ‘jog in place’, so to speak.