At first blush, the central narrative of the Defence Department’s 2018–19 budget, as presented in the Portfolio Additional Estimates Statements (PAES), appears to be about the falling Australian dollar.

Defence’s estimated appropriation for 2018–19 has increased by $449.6 million since the Portfolio Budget Statements (PBS) were published in May 2018 (table 1, serial 4; tables and pages in this post refer to the PAES unless otherwise noted). This is largely due to a foreign exchange adjustment of $476.6 million, as shown in table 6.

Defence is supplemented for exchange rate fluctuations on a no-win, no-loss basis. If the Australian dollar falls against the US dollar and reduces the department’s buying power, Defence doesn’t reduce the number of F-35 jets, say, that the government has agreed to buy. Instead, the government tops up Defence’s funding.

Such adjustments can be substantial. This year it’s $476.6 million, but by the back end of the forward estimates in 2021–22 foreign exchange adjustments are predicted to require an additional $941.5 million in funding (table 6).

It’s not surprising that the numbers are so big; with around 65% of Defence’s capital equipment acquisition budget currently going overseas, declines against the US dollar and euro have a big impact (ASPI’s The cost of Defence, table 5.1). For example, the F-35, Defence’s most expensive project for 2018–19, has gone from $1.821 billion to $1.933 billion, an increase of $112 million, with ‘the forecast variation primarily attributable to foreign exchange updates’ (table 63, page 81). A number of sustainment activities also attribute increases in spending to exchange rate variations (table 64).

But what’s surprising in light of those adjustments is that the department’s estimated overall capital spend on equipment, ICT and facilities has actually decreased by $434.9 million since the PBS (table 8, serial 6). At a rough estimate, around two-thirds of the increased cash requirement due to exchange rate variations can be attributed to the capital program. So the capital spend should have gone up, not down, because Defence needs to spend more Australian dollars to deliver the same capability. Yet once we add in, say, $300 million to account for exchange rate variations, the capital spend will miss its PBS prediction by more than $700 million.

In fact, compared to last year’s overall capital investment, we’ve actually gone down by $200 million this year, from $10.8 billion (PAES 2017–18, table 8) to $10.6 billion—even after including the extra money that’s required to compensate for the declining Australian dollar. (Defence doesn’t publish actual achieved capital expenditure for the previous year, so it’s only possible to compare with the previous year’s PAES prediction, not with actual achievement. So the 2017–18 figure may have been more or less than $10.8 billion.)

So how do we account for this? As always, it’s hard, if not impossible, on the basis of public information to determine what is at work. There are generally three suspects.

First, funds have been moved from acquisition to sustainment. And indeed, the sustainment budget has increased by $455.4 million (table 9). Some of this is due to exchange rate variations, but two big elements are an increase of $208.9 million in Chief Information Officer Group sustainment and $196.7 million in Joint Capabilities Group sustainment. Since these are the two groups essentially responsible for Defence’s ICT and ISR (intelligence, surveillance and reconnaissance) backbone, it would appear that keeping Defence’s core information systems functioning is costing more than expected.

Second, acquisition projects under-delivered against expectations. This can be unplanned—for example, in the case of LAND 121 Phase 4 (Hawkei protected mobility vehicle—light), which will spend only $203 million of a planned $396 million ‘primarily due to ongoing vehicle reliability issues, which is expected to delay the commencement of full rate production’ (table 63, page 84).

But it can also be planned, in order to manage cash flow pressures—for example, in the case of AIR 7000 Phase 2 (P-8A maritime patrol aircraft), which is reducing its spend by $183 million to ‘re-align payments with available budget’ (table 63, page 81). Another adjustment is a decrease of $150 million to the air warfare destroyer program due to ‘reprogramming’ of costs.

And the third possibility is that unapproved projects are being delayed to manage cash flow pressures. When the department is facing cash shortages, delaying approval of new projects is generally its first response. As we have argued previously, since Defence had to move $6.9 billion forward to pay for the ramp-up of the construction of the future submarines (page 49), it’s likely it has had to delay other projects to free up the cash.

What’s the evidence that this has happened? The bottom line in the PAES capital investment program table (table 8, which includes equipment, ICT and facilities) also includes unapproved projects (those that haven’t received second-pass approval to commence delivery), but they’re not broken out as a separate line so there’s no smoking gun that new project approvals are being delayed.

But if we examine the capital equipment program more closely, the evidence is a little more suggestive. Table 63 states that the cash flow required this year for the approved capital equipment program has increased by $107 million since the PBS (table 63, page 86). But table 8 states that the cash flow required for the total major capital equipment program (both approved and unapproved equipment projects) has gone down by $363.2 million since the PBS (table 8, serial 1).

So, if spending for approved equipment projects has gone up by $107 million while total spending on equipment projects has gone down by $363.2 million, it seems that delays to new equipment projects originally scheduled for approval this year are about the only thing that can account for the decline in capital equipment spending since the PBS.

Since there hasn’t been an update to the public Defence Integrated Investment Program in nearly three years, there’s no hard evidence to confirm that project approvals have been delayed. The government and Defence are claiming record numbers of project approvals, but there’s no comprehensive list of what the project approvals actually are, so it’s hard to know whether the IIP is being approved on schedule. But the PAES numbers give cause for concern.

Overall, the naval shipbuilding program is getting money out the door. Granted, SEA 1180 (offshore patrol vessels) will miss its $274 million prediction for 2018–19 by $53 million due to ‘delays in contract signature’, but at least it’s in contract and steel has been cut on the first ship (table 63, page 85). SEA 1000 (future submarines) will require $38 million more this year than its original PBS prediction of $418 million, suggesting spending is now ramping up apace (table 63, page 84). And with government approval for the design and production phase of SEA 5000 (future frigates) secured, its estimated spend for this year has gone to $222 million (table 63, page 85). The submarines and frigates are showing a healthy increase of 47% and 55%, respectively, in spending over last year.

Also on the subject of shipbuilding, there’s a tantalising line in table 6 (which lists variations to Defence funding since the PBS) called ‘Australian Naval Infrastructure Pty Ltd—additional equity’. Since ANI is the former arm of ASC that was hived off to build and operate the naval shipyards, this suggests that the construction of shipyards could be costing more than expected. But no numbers are provided ‘due to commercial sensitivities’.

It’s always hard to develop a reliable narrative based on a few data points, but the story seems to look like this: The government and Defence are continuing to pump what funds are necessary to ramp up shipbuilding. But that, combined with greater than expected costs to operate Defence’s dated and broken ICT backbone, is putting the squeeze on the rest of the capital program.