In October 2022, Canada’s Parliamentary Budget Officer (PBO) generated a report that assessed the life cycle cost estimate (LCCE) at $306 billion Cdn for 15 River Class destroyers being acquired under the Canadian Surface Combatant (CSC) project. One media report dubbed it “more torpedoes” of the project and quoted a former procurement chief of the Department of National Defence as saying the project was “out of control”.
But does an LCCE justify such claims?
The PBO’s LCCE appropriately includes four phases of the ship’s life cycle: development (options analysis, studies, R&D), acquisition (purchase, construction, integration into service), operations and sustainment (ships’ deployments, support & maintenance, technology insertion, mid-life upgrade) and disposal (demilitarization, dismantling or destruction of systems, disposal of hazardous materials, sale of parts and metal scrap).
Canada has directives that require a cost estimate “when decision-makers consider whether to undertake an investment” and when making procurement decisions that are based in part on “full life-cycle costs whenever possible”. Missing is a definition of when an LCCE is not possible.
Given the Parliamentary Budget Office’s (PBO) report was released in 2022, one wonders whether the LCCE and the potentially growing project acquisition costs could be the reasons that Canada’s Arctic Foreign Policy (December 2024) indicated that the CSC project would deliver “up to 15” River Class destroyers. Until now, 15 was the definitive number and the minimum number of surface combatants defined as required by the RCN.
There is also a wide body of knowledge on the Internet – particularly in terms of accounting institutions – suggesting that LCCEs should be employed to select between two or more options based on the lowest life cycle cost. I personally have heard a foreign naval costing official state that the only credible employment of warship LCCEs is to assess the anticipated life cycle annual costs of various procurement options around the fifth-to-tenth year point, to inform bid evaluations.
One of the most comprehensive documents on the subject is the U.S. Cost Estimating and Assessment Guide – Best Practices for Developing and Managing Program Costs produced by the General Accounting Office in March 2020. It identifies what constitutes a proper LCCE. Aside from the two obvious attributes of comprehensiveness and accuracy, they also call for the LCCE to be well-documented (e.g. traceable to the original sources of evidence, the rationale and the assumptions employed for each cost element) and credible (e.g. identifies the level of uncertainty of the analysis in terms of the risk of both source data and assumptions, as well as the LCCE’s sensitivity to source data risks and major cost-driving assumptions).
The literature indicates that various LCCE estimating methods are in common use, both individually and in combination. Generally, the methods are based on the historical data of comparable assets (e.g. the Halifax Class frigates or United States Navy equivalent surface warships). These are then analyzed employing parametric models, regression analysis and/or cost estimate relationships. Variations on these approaches were employed in three separate estimates in the PBO’s River Class destroyer LCCE report, with the three results averaged to generate the single point estimate of $306 billion.
Neither the concept of a LCCE to inform decisions nor the approaches used to produce such estimates are in question. Rather, the issue is understanding how confident we can be in these estimates, given that their reliability depends on a variety of factors, some of which are identified below:
• the accuracy of the historical information for one or more comparable exemplars;
• the comparability of the exemplars’ operational profiles and major equipment systems’ maintenance requirements to the new platform being acquired;
• where knowledge is lacking in areas such as new technologies, inflation indexes, foreign exchange impacts (e.g. for systems being repaired-by-replacement by offshore original equipment manufacturers) and economic crises (recessions, pandemics) over the 25–30-year life of a platform, the large numbers of assumptions involved and the variances in levels of agreement by experts;
• the potential for significant variations in the type, number and impact of disruptors between the historic comparator(s) and what may occur going forward in the weapons platform to be acquired in terms of operational profiles (e.g. time on contingency operations in distant theatres of conflict), ‘infant mortality’ and through-life major system failures (e.g. replacement of major electrical generators driving electric propulsion), and sensors and weapons that are rendered ineffective by new technologies (e.g. drones) as well as the impacts of quantum leaps in general technologies (e.g. AI driving many more combat system software upgrades); and
• the method of converging on a LCCE when multiple approaches are employed, which provide differing estimates, whether by simple averaging or by weighting the approaches based on the levels of uncertainty of each.
Also important are the available analyses supporting estimate credibility mentioned above and missing from the PBO report in question – confidence ranges based on sensitivity analysis for the major cost driving parameters, and the aggregate level of the estimate’s uncertainty based on the probability and costs of known risks materializing (e.g. a major fire while operating the platform) and the uncertainty of assigned costs where they are based on judgment and assumptions (e.g. the consumer price index out 30 years).
One other factor stood out after literature survey of papers available on the Internet. There appeared to be no verification case studies indicating the accuracy of an initial LCCE when compared to actual costs over the life of the subject artifact. A naysayer would liken this to throwing a dart but never watching to see where it lands on the dartboard.
One therefore is left with many questions regarding the LCCE in the PBO’s report on the River Class destroyers. Not only is unclear how the generated single-point estimate will be employed, but the lack of detailed documentation in supporting annexes prevents one from understanding the estimate’s accuracy and credibility. I am sure that the PBO has this information, and I may have missed some appendices beyond Appendix A; perhaps the PBO could enlighten the public.
In the meantime, one should therefore be careful about reaching conclusions based on such life cycle cost estimates. One should not conclude that they are another ‘torpedo’ of the intended acquisition project, proof that the project is ‘out of control’, or that such estimates warrant a reduction in the number of platforms to be acquired.
