As part of its National Shipbuilding Strategy (NSS), the Government of Canada has outlined a long-term project to renew Canada’s federal fleet of combat and non-combat vessels. One group of ships within this strategy is the Canadian Surface Combatant (CSC). This program consists of building up to 15 ships to replace Canada’s 12 Halifax-class frigates and three Iroquois-class destroyers.

The original budget to build these 15 ships was set in 2008 at a total of $26.2 billion. Given the factors of inflation, rising cost of material, labour and other expenses, the original budget is not enough to construct the planned number of ships by the anticipated start year of 2020.

Rod Story

With a mandate of providing independent analysis to Parliament on the state of the nation’s finances and to estimate the financial cost of any proposal for matters over which Parliament has jurisdiction, the Office of the Parliamentary Budget Officer (PBO) prepared a report: The Cost of Canada’s Surface Combatants. This report which was released on June 1, 2017, was spearheaded by Rod Story, Financial Advisor – Analyst on the Expenditure and Revenue Analysis team at PBO.

Marcello Sukhdeo, editor of Vanguard spoke with Rod Story recently about this report and the methodologies used in arriving at an estimated cost for CSC.

Q: The Office of the Parliamentary Budget Officer issued a report on the cost of Canada’s Surface Combatants recently. It was noted in that report that you played an important part in preparing it. What methodology did you use in arriving at the cost estimation for CSC?

We used three different cost estimation methodologies: a primary one that is parametric-based, as well as two secondary methodologies that are heuristic-based to confirm the results of the primary methodology. For the primary one we used TruePlanning 16.0 by PRICE LLC, a software which can perform costing of all sorts of different things. But in this case, I had to tailor it by first calibrating to a known program of a similar type, for example, using an earlier surface combatant, and then changing the parameters to reflect the new program. The theory is that past performance will inform future performance.

The other two methodologies were simpler heuristic-based ones. They were derived from two RAND reports that had investigated the cost of naval shipbuilding in the United States and Australia.

Q: Can you share a little more about the process you used with these methodologies to arrive at a costing for CSC?

In estimating the cost of the CSC program, TruePlanning was first calibrated to the CPF program using a CPF Cost Performance Report from March 31, 1994. This report was from near the end of the CPF construction program; as such, the cost estimates it contained were unlikely to change significantly. The parameters that were changed from the CPF model to create the CSC estimation were: development dates, production dates, salaries, weights of the seven work breakdown structures, inflation rates for the various work breakdown structures, updated ammunition costs, and spare parts costs.

On the heuristic-based methodology we used U.S. data from the RAND paper (Why Has the Cost of Navy Ships Risen?) for the factors that increase surface combatant costs and applied them to the CPF to estimate the CSC cost. Some of the heuristics we used were the cost of the ninth ship, 2 per cent inflation a year to account for non-obvious capability, increasing cost linearly with the increase in LSW and power density, inflating at 1.2 per cent above GDP inflation when in program plus 0.4 percentage points between programs, and taking into account tax rate differences between the two programs.

The other heuristic-based methodology is based on an idea taken from the RAND paper Australia’s Naval Shipbuilding Enterprise. We used the idea of the difference between U.S. and Canadian labour rates to convert the cost of building a ship in the U.S. to building it in Canada. This was done by comparing the cost of the ninth ship, using GDP inflation plus 1.2 per cent, using the current U.S./Canadian exchange rate, adjusting the cost for the difference in LSW, and adjusting for the difference in labour costs between the U.S. and Canada.

Q: The original budget for the CSC program was $26.2 billion, which based on today’s dollar value is not enough. What’s the estimation of the total cost to build 15 ships for this program at this current time?

It’s important to point out that when we state these numbers, these are nominal dollars, or as-spent nominal dollars. The as-spent nominal dollars that we project the cost to be now is $61.82 billion dollars not $26.2 billion. With the $26.2 billion budget, the cost per ship was $1.7 billion in then-year dollars. Today, we estimated that it will cost $4.1 billion in then-year dollars per ship, which is a cost of roughly 2.4 times more than originally budgeted.

Q: So, you are saying $61.82 billion dollars to build 15 ships. If the government should stay within the original budget of $26.2 billion dollars, how many ships could be built?

We estimated that they can only build six ships if they were to stay within that budget.

Q: Inflation is a critical factor in costing; how did you account for inflation in your report? How was this calculated?

There are different inflation factors depending on what you’re looking at. For inflation on the government side of things, like government project management, government facilities, and things of that nature, I used the standard GDP inflation, which has been around 2 per cent since the early 1990s. But when it comes down to the ships, we used the inflation rates of the Congressional Budget Office, which is GDP inflation plus 1.2 per cent. Due to the basket of goods used in naval shipbuilding, which are more expensive than the general economy, inflation for those items are at 3.2 per cent. But the most expensive item for inflation is the combat system, which has an inflation rate of 6.5 per cent because of its increasing size and complexity and ability.

Q: The PBO report states that Canada would save 25 per cent, or $10.22 billion, if the ships are built at a foreign shipyard using an original ship design rather than in Canada. Can you elaborate further on that?

Just to be clear, the numbers we spoke about before are as-spent or nominal value. This one is based on 2017 dollars. If you spent all the money exactly today, the total budget for CSC would be $39.94 billion, not $61.82 billion. So, this $10.22 billion is in real numbers, or you can say 2017 dollars.

So two things are driving that cost difference. One, it’s built in a foreign shipyard that has already built at least nine of the ships. So they are no longer needing to go through a learning curve. In addition to that, the assumption is that there are no design changes; that is, we take the ship as it’s already scoped and designed and in operation. So, Canada is not going to go in and do a large number of changes. Basically, they’ve built nine ships, and we’re taking the 10th through the 24th ship.

You have two things driving the cost saving: one is the learning curve. When Irving starts to build in the Halifax shipyard, clearly they’re going to go on a learning curve. And they have two things affecting that learning curve. One is, any time any shipyard builds a new design, there’s a lot of churn in the first eight to nine ships. During that time, they learn how to build most efficiently.

The other aspect that’s driving this is: Irving has not built a surface combatant. Surface combatants are vastly different than what they’re building now in the case of the Arctic Offshore Patrol ships. They’re much denser, much more complicated, and again that also affects the learning curve. So, your first eight ships will be that much more expensive purely because you’ll have that much more to learn.

Q: So, it takes about nine ships to really get it right?

Well, to reach your maximum efficiency. Basically, theoretical analysis has shown that by the ninth ship, you’ve now reached that point. This is analysis done in the original 2006 RAND report. So yes, it takes nine ships before you get to that maximum efficiency.

Q: Recently PwC issued a report on costing for CSC. It says that it’s cheaper to build in Canada than overseas. Based on what you’ve just explained, I can understand why your conclusion is different. Can you shed some more light, for the benefit of our readers, on the difference in the two conclusions?

We don’t comment on other organizations’ reports. But what I can say is that we did not consider economic multipliers and the effects in the overall economy, while they looked at that.

Q: What are the challenges of building the ships here in Canada?

Basically, learning curve is the primary challenge. We have not built surface combatants since the finishing of the Halifax frigates in 1996. All that knowledge has been lost; it has to be relearned.

The other challenge is the amount of changes that DND will want to make on the design. If they are to take a design from another country, how much are they going to change that design? It’s not like you change 5 per cent and expect the cost to increase by the same percentage. In fact, it multiplies. Once you’ve changed about 20 or 30 per cent, you may as well have redone the design from scratch. It’s one of those things – it’s very multiplicative. These ships are so dense in the sense of so many things are packed in so tightly and so dependent on each other. You make one change, it ends up propagating throughout the whole ship quite often. They have to be extremely careful. Managing those changes will be quite a challenge.

Q: If there is any delay in the production start date of 2021, what would the consequences be? What would this cost Canada in dollars?

There are two things that one needs to consider with the delay of the program. One is just the absolute inflation cost of the delay. As I mentioned earlier, the inflation rate for surface combatants is higher than just general GDP or consumer price index inflation. The cost of moving the ship out by each year will invariably increase the cost by $1.5 billion dollars.

The other thing to consider when things are being delayed is that one cannot lose sight that Irving Shipyard is building the AOPS. The assumption is you need effectively about a 4- to 5-year overlap; I wrote down 5-year overlap in the report, minimally, so you do not lose your employees. Then what happens is, you go on an even steeper learning curve, which increases your cost.

Q: Well to close off, congratulations on your report, and thanks for sharing the insight into the PBO’s costing for CSC with Vanguard magazine. We truly appreciate it.

You’re welcome and thanks for the opportunity.

For the PBO’s report: The Cost of Canada’s Surface Combatants, go here

NOTE: Through the new Defence Policy, DND has issued a new estimate for the CSC project.

“Through Strong, Secure, Engaged the Government will deliver the capabilities the RCN needs to meet future defence and security challenges, both at home and abroad, and to carry out the tasks required of a modern navy,” according to National Defence. “We will replace the surface fleet through investments in 15 Canadian Surface Combatants. Defence conducted a year-long re-costing of the Surface Combatants. This involved private sector firms as well as international experts, such as the U.S. Navy. Based on this review, Defence estimates the cost of 15 ships at between $56 – $60 billion. The policy sets aside funding to deliver the full complement of ships the Navy needs to provide capability across the full range of operations. They will replace both the Iroquois-class destroyers and the Halifax-class frigates with a single class of ship capable of meeting multiple threats on both the open ocean and the highly complex coastal (littoral) environment.”