Yup. And DCostS has the multi-year inflation model which rarely survives contact with reality, but they don’t certify the cost ING’s unless you use their model, then everyone gets up to go to the bathroom when the Auditer General comes by to ask you why your cost was off be so much… sigh
Funny you mention the car…I walked a few guys through a $265K Toyota Camry…once I told them that’s how the GoC costs MCPs, the lights went on. (And why you can’t say price per helicopter is $340,000,000….)
It also seems that no one uses the same models so TBS, PBO, Dept of Finance (FIN?), all do some different things with nuances to assumptions, inflation and phasing that result in different costings.
The other variable is the actual contract milestones, which can make huge differences in cost. If a major payment milestone is delayed a few years, your costs might go up a lot just from billable hours shifting to higher rates. So everytime they pause major decisions for elections, budget reviews etc it adds on hundreds of millions to the project in real terms over the life.
Inflation is another fun one; StatsCan has dozens (hundreds?) of different indices, and they each apply differently to different sectors. So you may get a general 3% inflation, but a 10% on things like steel and copper for supply chain reasons, 5-7% on labour etc etc. They also don't really do +/- to show the uncertainty on budget demands, so if you work through something, and end up with a most likely cost of around $10, but may be as low as $7 best case, or $20 worst case, you basically have to ask for $20, then get kicked in the balls when you return unspent money that was essentially contingency.
That happens a lot on the in service side as well, where we have been forced to go to task based annual contracts for in service work, which requires us to have all the money in the project in March to get the DND 626 task authorization form actually issued. So we do things like tie up $1M in April, track monthly burn rate against it, and then generally either have to return money too late for anyone else to use, or scramble to get more money. Before the updates to the Log SOW for R&O we just got monthly invoices, and would ask for budget top ups as the year went on, so was much less PM work on the financial side.
The amount of time and effort that the CAF and the public service writ large puts into all of this is pretty appalling, and probably doesn't get much better real results compared to the initial back of the napkin cost estimates somebody did a SWAG on during the initial bit. But no one wants to ever hear about the real SWE costs of all this internal processes, or the lost opportunity cost of those same SMEs working on basically process busy work that doesn't really add value but is required to get things approved.
Travel is probably a great example, where we have people doing thousands of dollars in staff time for the bureaucracy to travel, with delays making things like airline tickets cost a lot more, while also not working on things that are their core job.
