[See update note at the very end of this post which means that some parts of this post – re 24/25 spending – need correcting and reframing]

In yesterday’s post I tried to present the Reserve Bank Funding Agreement for 2025-30, as approved by the Minister of Finance and the Bank’s Board, in the context of the previous agreement, and the variation to that agreement signed up to by Grant Robertson a few weeks before the last election (which hugely increased the amount available for operating expenditure). Against that benchmark, yesterday’s agreement didn’t seem to display much restraint at all, even on the headline figures.

The Minister of Finance had sought to make much of (a) how much lower the agreed numbers were than the first bid made by the Bank last September (a total of $1.03bn over five years for both capital and operating expenditure – capex being only $50m in total), and b) relative to a number we had never seen before, what she described as the Bank’s operating expenses budget for 2024/25 of $200m. This latter was the basis for the much-publicised 25% cut claim.

I was a bit dismissive of that presentation, and what seemed mostly to be spin (unlike the RB’s own press release).

So I’m now going to try to step through it all fairly painstakingly and more slowly/carefully than I did yesterday.

First, here are the operating expenditure numbers for things covered by the 2020 Funding Agreement, showing both the original agreement and the August 2023 variation.

There were really big increases granted in that 2023 variation (far more than, say, implied just by the inflation surprise). Note too that spending for 23/24 was allowed to be much higher than in 24/25. I’m pretty sure we never got a specific explanation for that, but it probably related to some major new initiative.

By law, the funding agreements are required to specify maximum expenditure for each single year (the Bank can’t just pick and choose, transferring money from one year to the text on a whim). However, in the 2023 variation Grant Robertson explicitly allowed the Bank to treat the last two years as a single aggregate pool (apparently he had done the same for the last two years of the previous agreement). In principle, I have no particular problem with that (although of course most government agencies would have to come to Parliament and get a fresh appropriation each year) but it opens up a number of risks.

And here are the old and new Funding Agreement opex numbers, expressed in constant 2024/25 dollars (for each year I’ve used the level of the CPI for the average of the relevant Dec and Mar quarters).

There are a number of ways to look at that series. One way is to compare spending in 29/30 with the average allowed (over the full five years) under the previous Funding Agreement. Those two numbers are almost identical.

Another might be to compare 29/30 with the approved Funding Agreement level for 24/25. That implies a cut of about 6 per cent in real terms.

If we wanted to be slightly partisan about it (Willis can only be held accountable for stuff that has happened on her watch), approved opex for 2029/30 will be a little over 10 per cent higher in real terms that what was approved for 22/23, the last full year of the previous government.

In each of those comparisons note that the fresh exclusions from what is covered by the Funding Agreement mean that 2025-30 numbers are understated relative to the numbers for earlier years. It is pretty unsatisfactory that neither the Bank nor the Minister (nor Treasury) has provided any sort of reconciliation table to provide a clearer sense of magnitudes.

But all of those comparisons are between Funding Agreement approved numbers for each year. What about comparisons with what the Bank has actually done, or planned to do, in the current (24/25 year).

In the Bank’s Statement of Performance Expectations, published in the middle of last year and fully adopted by the Board (signed by both Quigley and Orr), these were the Bank’s financial forecasts for the current year.

Total operating expenses were forecast then to be $231 million. That was an increase from $186 million the previous year, and $105 million as recently as the year to June 2021.

As it happens, and shortly after the Statement of Performance Expectations came out I wrote a post about some aspects of their numbers

It was pretty breathtaking stuff. What I don’t think I’d noticed by then was that the Minister of Finance had already, in April 2024, set out this is her annual Letter of Expectation to the Board.

That seemed pretty clear. $149.4 million was the baseline for what they were allowed for (Funding Agreement covered) opex in 2024/25, the Minister seemed to be pretty clearly encouraging them to focus on reprioritisation not bids for more and yet (a) the Board signed up to a massive increase in opex for 24/25 (well ahead of the Funding Agreement number for that specific year) and b) then in September bid for a big increase in real opex spending allowances for the coming five years (roughly a 25 per cent real increase).

But how does that $231 million number compare to the $149.4 the Bank was allowed under the then Funding Agreement for 24/25? There we get some hint from a helpful table in the Bank’s 2023/24 Annual Report

Now, this has to be read carefully too. You will notice that the Funding agreement: annual allocation number ($177m for 2023/24) isn’t the same as the Funding Agreement number in the first table above. That is because in the 2020-25 Funding Agreement there is a separate category for “direct net currency issue expenses” (currency is, after all, a profit centre for the Bank so there is some logic in treating it separately). In the year to June 2024, $13.5m was allowed for that set of expenses (subtract that from $177m and you get the $163.5 in the table above).

You will also notice that of the total operating expenses ($186m that year), $28m were for items not covered by the Funding Agreement at all (these exclusions matter, and as noted above they are growing).

So lets go back to that $231m budget for 2024/25 that the Board approved in mid 2024. Of that, $14.5 could have been for those direct currency issue expenses (that was what was allowed in the Funding Agreement), but another table in the Annual Report suggests that in 2022/23 and 2023/24 actual net currency issues expenses had only been $5-6 million per annum, and the Minister’s Cabinet paper says that the latest budget for currency issues expenses for 24/25 is $9m. In absence of further information, and with a little extrapolation, we’ll allow $30m as having been for things not covered by the Funding Agreement at all. Subtract those two items and you are left with $192 million, which is a lot larger number than the $149.4 million allowed for 24/25 in the Funding Agreement (2023 variation).

Now, you will also note in the table – penultimate line, although not clearly labelled – that in 2023/24 the Bank had underspent the Funding Agreement amount by $19m (some mix of currency and other opex), and so – per the Robertson agreement mentioned above – they could in 2024/25 spent up to $19m more than what had been specified for that year alone. That would take the permissible limit to $168.4. Even that number is still a long way short of what the Bank was actually budgeting for the year (see previous para), and it relied scope for one year’s spending that shouldn’t have been prudently used for anything other than one-offs, since a new Funding Agreement was just about to be negotiated, in a period of straitened fiscal circumstances, and when the Minister had already warned them about the future. And yet they were planning a 21 per cent increase in staff expenses in a single year – the final year of the old agreement.

As noted earlier, in the Minister’s Cabinet committee paper – and her press release – she referred several times to an (apparently revised) Reserve Bank operating expenses budget for 2024/25 of $200m. If we take off $30 million for spending not covered by the Funding Agreement and $9m for currency expenses that would get us back to $161m, still substantially more than the $149.4 allowed for that specific year, although lower than what would have been allowed in that year from carrying forward the previous year’s underspend.

You have to wonder at what point the published budget was revised so sharply (down 13 per cent for the full year). Perhaps Treasury had some input when the saw how far out of line the published budget was?

It wasn’t as if this big increase in opex for 2024/25 looks to have been all about one specific project. According to the table in the Cabinet paper, in 2024/25 there is a 40 per cent increase in spending on the core functions (monetary policy, markets, and financial stability), following a 48 per cent increase the previous year. And the support functions (now running at more than four times the level of spending in 2017/18) also see spending rise 17 per cent this (24/25) year.

The Minister’s Cabinet paper, released yesterday, told us that total staff numbers (FTE) had reached 660 by the end of January. Last year’s Annual Report tells us that as at 30 June last year – only 7 months previously – they’d had 601.3 FTEs. It is staggering increase in the last year of a five year Funding Agreement – relying for that year’s Budget on a big underspend the previous year, and despite the Minister’s warning of coming fiscal stringency.

It seems pretty clear that Orr, Quigley and the rest of the Board were engaged in a strategy designed, in effect, to try to bounce the Minister into agreeing to a new higher baseline spending number that (probably) would have been even higher in real terms for the next five years that what they were spending in 2024/25. If so, and it is difficult to read what happened any other way (given that we know what the Bank bid for last September) it is really pretty inexcusable conduct all round (and frankly pretty poor behaviour vis-a-vis staff, since it was a risky high-wire act and if it didn’t come off it was likely to be staff who paid the price.

All the exclusions (and changes in exclusions – for example, those net currency expenses now have just a five yearly total cap rather than annual provisions) mean it is difficult to know with any certainty how much the Bank is really being cut back relative to this year’s spending

Take that $200m budget for 24/25, subtract $30m for spending excluded even in the previous Funding Agreement and $10m for net currency expenses and you are back to $160m. The new Funding Agreement allows for spending in 2025/26 of $155m BUT there are new exclusions. There is this one, which is at least quantified

That might be $5m per annum. In addition (and as mentioned yesterday) spending around the CDBC project is also newly excluded and there are a number of other items. Subtract those of the $160m budgeted on Funding Agreement covered items and you get pretty close to that $155m. Even allowing for a couple of per cent of inflation, there isn’t much sign of real spending cuts – in an organisation that had massively increased spending as recently as, well, this current year.

For the period beyond that there looks to be somewhat more restraint imposed on the Bank but it is nothing very dramatic, in an organisation where spending and staff numbers have blown out in recent years, and kept doing so in 2024/25. (And as noted in yesterday’s post, the Minister rejected Treasury’s push for lower numbers.) Actual real spending later in the Funding Agreement period remains – per table near the start of the post – well above what was spent in the last full year of the previous government.

There are a lot of numbers in this post, and several at least are uncertain. But the bottom line seems to have been one where the Bank’s Board and Governor tried it on, with their preposterous bid for much higher Funding Agreement operational expenditure. Thankfully that made no headway with the Minister. No one seems to emerge with much credit, and if you were inclined to make an allowance for The Treasury there is the mystery of how that huge operating expenses budget ($231m) happened in the first place. Surely they were aware before the thing saw the light of public day?

As for the Minister of Finance, the attempt to claim a 25 per cent reduction relative to this year’s budget looks even more disreputable than I realised yesterday. At least 20 per cent of that budget seems to be on items that weren’t even covered by the (previous) Funding Agreement.

The Minister of Finance appeared on RNZ this morning and from the resulting story it appears that she wanted to emphasise a message that “New Zealanders are doing it tough…..We expect you to show some restraint. Focus on your core statutory requirements”. Which is good stuff as rhetoric, but the reality seems to be one where the funding the Minister has approved will only stop the Bank continuing to expand further (having already expanded for one more year than almost all other government agencies), and do little or enough to ensure they are focused on the core stuff. That’s a shame, and I’m sure most New Zealanders would prefer a few more (say) kidney transplants than cementing-in very high levels of spending – far far above pre-Orr levels in real terms, and above even levels late in Labour’s term – at the central bank.

Perhaps it is only a few geeks and nerds who really care about the law being followed in these obscure matters, but as a reminder

There is still no sign of a budget (let alone a nice to have like a proper reconciliation table).

UPDATE:

Here is another way to try to look at it

The blue bars are actual (real) operating spending by the RB on things covered by the Funding Agreement, with the 24/25 numbers being per the Bank’s budget disclosed in the Minister’s Cabinet paper. On the headline new Funding Agreement (and allowing for 2 per cent inflation) we get the red bar. Allowing $7 million for the new exclusions – $5m for the Deposit Takers Act implementation (see above) and (somewhat arbitrarily) just $2m for the rest – we get the yellow bar. If the yellow bar is roughly right, real spending (like for like) in 25/26 would be 4.4 per cent lower than in 24/25 and barely lower at all than the actual for 23/24 (more than half of which year was under the current government).

Because we do not know the precise value of new exclusions, the yellow bar is illustrative only, but the direction is pretty clear. All the other numbers are from RB documents or the Minister’s Cabinet committee paper.

UPDATE 19/4: There is a mistake in this post, in interpreting the 2024/25 budgeted spend. A full post, revising and amplifying the story, will follow on Tuesday.