After working for the RBA for 30 years, here’s how I’d make it more accountable
- Written by John Simon, Adjunct Fellow in Economics, Macquarie University, Macquarie University
How can we tell if the Reserve Bank of Australia – which sets the interest rates that shape our national and household budgets – is doing a good job?
That’s partly what the Albanese government set out to test in setting up a review of the bank in 2022.
When that review reported back last year, many people were surprised it devoted so much of its attention to governance and culture, rather than things such as its inflation target, which helps the bank set interest rates.
The review recommended a more open decision-making process and an interest rate board made up of experts, something the parliament is yet to agree on.
However, for people such as myself who have worked at the bank (I was there for more than 30 years), that focus was exactly the right one.
To boost Australians’ trust in an institution that’s central to all our lives it is important to pass the legislation and implement the recommendations of the review.
In the meantime, there is more we can do: we can expect the bank be more transparent with all of us about what its plan is. This is not just the outcome it is aiming for, but how its decisions will get us there.
Bad luck vs bad forecasts
So how can we test the Reserve Bank’s performance? One way would be to see whether inflation is in its target band.
But there are all sorts of reasons why inflation might be out of the band that have nothing to do with whether the bank is doing its job. It might be out of the target band because of bad luck or because of things outside of the bank’s control.
Another method might be to look at the reasons inflation is not in the target band.
But this isn’t as useful as it sounds. If, for instance, inflation was outside of the target band because government spending was higher than the bank anticipated in the lead-up to an election, it could be argued it was the bank’s job to be better at anticipating and to set monetary policy accordingly.
It would certainly help for the bank to set out clearly the assumptions behind its forecasts and the reasons it made those assumptions at the time it makes its forecasts. But even then it would be hard to tell bad luck from bad forecasts.
As it happens, the bank regularly publishes an account of the reasons why its forecasts have missed their marks. But the accounts generally tell us more about how the forecasts were wrong than why.
The account might, for instance, tell us services inflation was higher than forecast because productivity was lower than forecast and population growth was higher than forecast. But it won’t tell us much about whether the forecasts were reasonable in the first place.
Early last month, the bank’s deputy governor Andrew Hauser went further and seemed to argue the economy is subject to so much radical uncertainty that forecasting is itself a fool’s errand.
In such an environment, looking at what happened in the past won’t provide much of a guide to what will happen in the future.
Now I happen to think the argument that the economy is as hard to predict as what Elon Musk will do next – as Hauser argued in a speech this month – is wrong.
I think you can learn about the future by studying the past. Hauser himself devoted much of his speech to learning from the past.
Spell out responses in advance
But Hauser and the bank make an important point: forecasting the economy is difficult, and it is unfair to hold forecasters to outcomes.
There is an alternative: hold the bank to inputs.
Bruce Preston at The University of NSW has proposed requiring the bank to state ahead of time, with sufficient specificity to be verifiable, what it will do in certain circumstances.
We could then discuss whether its plan is reasonable ahead of time, and verify whether it did what it said it would do when those circumstances came about.
This would not require a slavish adherence to a rigid path for interest rates. But it would require a clearly articulated strategy, with responses to various eventualities set out ahead of time.
As an example, the bank might say ahead of time that
if inflation is no longer expected to be below 3% by mid-2025, we will raise rates.
Or it might say
if unemployment rises above a certain point, we will lower rates.
Unfortunately, the bank has recently been reluctant to give such “forward guidance”, other than saying its decisions will be “data dependent”.
Partly this has been because of a concern that any such guidance would be taken as a promise about what it will do to interest rates.
But there’s a difference between articulating a clear strategy and saying what will happen to rates. The bank might be wanting to give itself wriggle room.
In any event, if the bank did articulate clear strategies, it wouldn’t completely solve the problem. There would still be lots of ambiguity about what the circumstances were and how to apply the clearly set out strategies.
So what should be done?
Processes are what matters
Standard procedure in the world of auditing is to focus on the processes used to arrive at decisions, and it happens to be the focus of the Reserve Bank review.
In essence, the review found the processes the bank followed were insufficient to guarantee it was doing a good job.
It recommended
setting up an expert monetary policy board and appointing members through a transparent selection process
making the culture of the bank less insular, more welcoming of challenges and more engaged with research
establishing a specialist governance board with an external chair to drive higher standards of performance across the organisation
That’s why it is vitally important for the legislation putting those findings into effect to make it into law.
Treasurer Jim Chalmers intended the two boards to start work in July. The treasury called for expressions of interest in March.
Chalmers says agreement with his opposite number Angus Taylor is close.
While it is true that we will never be able to tell whether the bank is doing a good job by looking only at outcomes, we can make its processes more trustworthy and verifiable. The legislation before the parliament is an essential step.
Authors: John Simon, Adjunct Fellow in Economics, Macquarie University, Macquarie University
