Parliament should fund the Reserve Bank annually

Readers may well be getting a little tired of the run of posts this year on issues relating to the review of the Reserve Bank Act.  In truth, so am I.    However, The Treasury is inviting comments, with a deadline of later this month.

There are two stages to the review.  The first, led by Treasury, is around implementing reforms Labour campaigned on (a statutory monetary policy committee, and adding some sort of employment goal).  The second –  as yet ill-defined – is to be led jointly by the Reserve Bank and Treasury and is to look at any other issues that the Minister agrees warrant review.  For now, The Treasury is inviting views on what issues should be looked at in Stage 2.

My broad response to them on that question has been “anything and everything”.  The Reserve Bank Act –  and the other Acts the Bank operates under –  has grown like topsy over 30 years.  Roles and functions have changed.  Expectations around transparency have changed. And models that might have seemed sensible in 1989 – eg the role of the Board –  no longer do so today.   Probably the Act should be rewritten from scratch, but even if that doesn’t happen, no part of the current Act should be outside the scope of the review.   My post the other day proposed structural separation –  spinning out of the Bank a new Prudential Regulatory Agency.

Today I want to focus on funding the Reserve Bank’s operations, an aspect of the Act that could be easily overlooked as part of the review, but where current legislation falls far short of best practice.

Historically, there were typically no checks on the spending of a central bank.  Central banks “printed money” –  physically or electronically – and didn’t need to rely on tax revenue appropriated by Parliament.   They typically earned a lot of income by issuing (zero interest) bank notes, and investing the proceeds (often in government bonds).  I regaled my kids recently with the tale of the night –  the RBNZ 50th anniversary –  when the Bank took over the Michael Fowler Centre and hosted a banquet and dance (free) for staff and spouses, includng a McPhail and Gadsby floor show.  The kids asked who paid for it, and struggled to get their minds around the notion that “we just printed the money”, but that was the way things were.  These days –  rightly –  the Taxpayers’ Union would be all over such extravagance –  or the chauffeur for the Governor (and, if I recall rightly, even the Deputy Governor).  There was an Annual Report, of course, but little detail and no effective spending control or accountability.  It wasn’t that the Bank was always extravagant –  it wasn’t – but there were few or no external constraints.

The reformers of the late 1980s recognised the need for things to change.   But expenditure control took a distinctly secondary place to promoting and preserving the operational independence of the Bank in the conduct of monetary policy (recall that at the time the Bank was conceived of largely as a monetary policy agency).    There was much less concern about controlling spending than about closing off backdoor ways for a Minister of Finance to exert pressure on the Bank’s monetary policy.  The ability to threaten an annual parliamentary appropriation –  “go easy on monetary policy or I’ll cut your funding”  –  was seen as an important risk

And thus we ended up with the Funding Agreement model.  Under this model, the Governor and the Minister of Finance may reach a five-yearly agreement, which has to be ratified by Parliament,  on how much of the Bank’s income can be used for operating expenses in each of the subsequent five years.   I wrote about the model when the current funding agreement was ratified in 2015.

There had even been talk –  probably not welcomed at The Treasury –  of using the Permanent Legislative Authority route (a model used for judges’ salaries and Crown debt servicing), a model under which no parliamentary authority would ever have been required for the Bank’s spending.  Even at the time, that seemed a little self-important.

The Funding Agreement model certainly means that for most of any five year period, the Reserve Bank does not need to worry about ministerial pressure exerted via its budget.  And, as it happens, there has always been a Funding Agreement in place, and those funding agreements did help to lower substantially the level of spending at the Bank.  In that sense, they were a step forward relative to what had gone before.   But they are very far from best practice.

Treasury has a nice document on its website outlining the principle and process of parliamentary authorisation of public spending.  It begins this way

A long-standing principle under the Westminster style of government is that no expenditure of public money can take place without the prior approval of Parliament. In New Zealand, both the Constitution Act 1986 and the Public Finance Act 1989 continue this historical requirement.

Appropriation ensures that Parliament, on behalf of the taxpayer, has adequate scrutiny of how public resources are to be used and that the Government is held accountable for how it has used the public resources entrusted. The Estimates specify for each appropriation:
• a maximum amount of resources that can be consumed
• the purpose for which the appropriation can be used.

As they go on to note, most appropriations are for only a year at a time.

The Reserve Bank model falls short in a whole variety of ways:

  • there is no legislative requirement for there to be a funding agreement at all, and no (formal) consequences if the Governor and Minister do not (or cannot) agree on one,
  • there is no breakdown of the agreed level of spending between the Bank’s (quite different) stautory functions, and thus no restriction on the ability of the Bank to spend the money  in one area rather than another as it chooses,
  • there are no penalties or formal adverse consequences if the Bank spends beyond the limits set out in the funding agreement,
  • while the Minister and the Governor can agree to a variation in the level of the funding agreement during its term there is no ability for the Minister (say, a new government) to override the agreement, even temporarily (as, say, there is in respect of monetary policy itself),
  • in (almost?) all areas of public spending, the Minister asks Parliament for approval, and Parliament approves (or rejects) the proposed level of spending.  Officials make bids and try to persuade ministers of a need for more money for their department or agency, but they have no ability to block or formally constrain choices of the Minister and Parliament.  But the Reserve Bank Governor –  himself, in effect, appointed by unelected people (the Board) –  has that degree of control (no agreement needed, any agreement needs the Governor’s assent, and no consequences from failure to agree).

And it isn’t even that workable a model.  No corporate sets operating expenditure budgets five years in advance –  circumstances change, the price level changes, and so on.

And this is not some trivial government entity doing some unimportant function.  It isn’t primarily a trading entity spending money to make money.  The Reserve Bank is the key entity responsible for short to medium-term macroeconomic management, and has a huge range of discretionary power in areas of financial sector regulation.   And it has several billion dollars of Crown capital. The financial expenditure provisions are a glaring anomaly, out of step with one of the fundamental principle of our system of government.

The system hasn’t been grossly abused: mostly the Reserve Bank seems to spend fairly responsibly, and the Minister is advised by Treasury at the point of renegotiation.   But, equally, it is not as if the Minister and the Bank have gone above and beyond the formal statutory provisions: they could, for example, lay out detailed annual estimates, by functions, at the time the Funding Agreement was released, and report actuals against those estimates over the following five years.  But they don’t even do that: it is the sort of lack of transparency that led me to suggest earlier that there is no more transparency around the Bank’s spending plans –  one line item per year –  than there is around that of, say, the SIS.  And unlike the Bank, the SIS needs an annual parliamentary appropriation or it can’t spend anything at all.

A common argument –  at least among central bankers –  is that somehow central banks are different.  There is only one important respect in which they are: they earn far more than they spend.  But even that isn’t very important here.  Central banks make money largely through the statutory monopoly on currency issue, which is just (in effect) another form of taxation.  And spending and revenue are two quite different bits of government finance: IRD might collect lots of money, but it can only spend what Parliament appropriates.

And what of those arguments about avoiding back-door pressure?  Even they don’t mark out central banks.  After all, we don’t want ministers interfering in Police decisions either (a rather more important issue than a central bank), but Police are funded by parliamentary appropriation.  So is the Independent Police Complaints Authority.   There are plenty of regulatory agencies where policy might be set by politicians, but the implementation of that policy is set by an independent Board, and where backdoor pressure could –  in principle be applied.  Other bodies publish awkward reports that make life difficult for politicians.  But those bodies too are typically funded each year by parliamentary appropriation.  It is just how our system of government works.

When I wrote about this issue in 2015 (having only recently emerged from the Bank), I was hesitant about calling for radical change.   The funding agreement system itself could be tightened up in various ways, which might represent an improvement on what we have now.   But there isn’t any very obvious reason not to start with a clean sheet of paper, and build a new system –  aligned with how we manage public spending in the rest of government –  starting from the principle of annual appropriations, with a clear delineation by functions (monetary policy, financial system regulation, physical currency etc), and standard restrictions on the ability of ministers to reallocate funds across votes).

I’m not aware of any country that funds it central bank by annual appropriation.  But historically, central bank spending all round the world was subject to weak parliamentary control.  This is one of those areas where the international models aren’t attractive, and the standard should instead be the way in which we authorise spending across the rest of government.   This is a policy and regulatory agency –  not, say, primarily a trading entity (like, eg, the New Zealand Superannuation Fund) –  and should be funded, and held to account, accordingly.