Something interesting happened at the Department of Defence’s Senate estimates hearing last week.

Buried among the largely predictable and interminable questions about how much Defence pays eminent persons in advisory and advocacy positions were a couple of pieces of information about the cost of SEA 1000, the future submarine project.

The first was either a very telling word or a misstatement—but considering it was made by Rear Admiral Greg Sammut, the head of the future submarine program, who is famous for both his absolute precision in language and his almost preternatural calmness ^[1], it would be odd if it were a misstatement.

When asked about the acquisition cost estimate for the future submarines, Sammut answered $50 billion constant (see page 32 of the Hansard record ^[2] or video footage ^[3]). As I explained in chapter 7 of ASPI’s recent The Cost of Defence ^[4] budget brief, constant cost estimates don’t take inflation into account. If you buy something for $10 this year, and buy it again next year and the year after, in constant prices it’s still $10 every time. It’s essentially a real value expressed in today’s dollars. But when Defence presents cost estimates, it usually does it in outturned dollars, which take inflation into account. The standard outturning rate that Defence uses is 2.5%, so what costs $10 this year will cost $10.25 next year and $10.50 the year after.

The ‘investment values’ provided in Defence’s public 2016 Integrated Investment Program ^[4] (IIP) are given in outturned dollars. So it would be reasonable to assume that the IIP’s figure for the future submarine of greater than $50 billion is an outturned number. But Sammut’s statement was that the cost is $50 billion constant, which is an entirely different kettle of dollars. And when outturned, it becomes a very different number.

To convert a constant number to an outturned number, we need both a spend spread (how many dollars per year and over how many years) and an outturning rate. As an exercise, we set out an indicative $50 billion constant spend spread based on a ramp-up in spending to achieve commissioning of the first submarine in 2032–33, followed by subsequent submarines on a two-year drumbeat. When outturned at 2.5%, the total becomes $79 billion. Of course, if you apply higher indices to account for the fact that the cost of military equipment increases at a faster rate than general inflation, you quickly get much bigger numbers.

In The Cost of Defence, I stated that the figures in the public IIP are very woolly and really just present a broad band. We now may be starting to see the figures for the future submarine become a little more defined and refined. So when Defence states that the acquisition cost of the future submarine is greater than $50 billion, it may be that the cost is actually almost $30 billion greater than $50 billion.

The other interesting piece of information that Sammut revealed was his estimate for the sustainment cost of the submarines—another $50 billion, again in constant dollars. What does that look like in outturned dollars? I set up an indicative spread assuming commissioning of the first submarine in around 2032–33, with subsequent deliveries on a two-year drumbeat and a 24-year service life. That outturns at 2.5% to about $124 billion.

Incidentally, if you’re thinking that the $50 billion sustainment figure has to be outturned, not constant, that actually results in a 28% lower annual sustainment cost per boat than the current Collins submarines. Again, as I argue in chapters 6 and 7 of The Cost of Defence, it runs counter to the laws of physics and the even more iron-clad laws of military cost escalation to think that a platform that is 50% larger and a leap of a least a generation in capability could cost less to operate than its predecessor. So it makes eminent sense that Sammut’s $50 billion is a constant figure.

So a lot hangs on a word. If Sammut did actually mean that the $50 billion acquisition and $50 billion sustainment are constant dollar estimates—and there’s no reason to think that he didn’t—then we’re actually talking a total of around $200 billion in outturned dollars.

But as I argued in chapter 6 of The Cost of Defence, total numbers are largely irrelevant because after the $200 billion, Defence will just keep spending billions on continuous submarine building off into the distance. The real question is how much of Defence’s annual cash flow is now tied up in the future submarines (and shipbuilding and the future fleet in general). The Cost of Defence estimated that the continuous naval shipbuilding enterprise would require an annual cash flow (in 2020–21 dollars) of $3.5–4 billion, or around 30% of Defence’s capital investment budget.

And the future submarines alone would require around $2 billion per year in acquisition funding and another $2 billion in operating costs (between sustainment and crewing) for a total of around $4 billion per year. Or put another way, around 10% of Defence’s total budget will be locked into one capability. Nothing in last week’s estimates hearing should lead one to revise that assessment.