_{ARTICLE   I   THOUGHT LEADERSHIP}

Time to liberate the RBA from its ‘double Micawber’

Published 30 June 2025

It is almost 18 months since the RBA began introducing the recommendations of the RBA Review such as the RBA Governor holding press conferences after each monetary policy decision. With the re-election of the Albanese government, a new Statement on the Conduct of Monetary Policy between the RBA and the Treasurer provides an opportunity to re-evaluate what has worked well and what tweaks could be made. What changes should be made? There are 2 pieces of ‘low-hanging fruit’ that should be a priority.

First, the RBA should abandon the Review’s recommendation that the Bank should always aim for the mid-point of the 2-3% target. Second, the RBA’s econometric models should play a supporting role in its decision making rather than appear to be the headline act.

These changes, while modest, are nevertheless important as the focus on “2.500%” provides a false sense of precision about the ability of monetary policy to deliver a particular inflation rate two years ahead. In other words, it holds the RBA up to a standard that can’t be achieved. This is because inflation will be more affected by what happens with oil prices or Trump’s trade policies than if the cash rate is 4.1% or 3.85%.

Nor is this simply recognising the obvious reality of operating monetary policy in a complex and rapidly changing landscape; the singular focus on 2.50% has clearly caused confusion about what the RBA is trying to achieve and how their decisions are consistent with those aims.

This was exemplified when the RBA cut its policy rate by 25bps in February, an episode in which the unhelpful reliance on econometric models was also plain to see. With inflation and wages growth falling back to long-term averages, cutting rates was the correct thing to do. But because the RBA felt compelled to justify the decision based on its models, it became unwieldy and confusing.

So, while the RBA cut interest rates in February, it also revised up its inflation forecast to 2.7% in 2 years’ time as that was what its models indicated (given that the RBA models also assume that its policy rate will move in line with market expectations which is also clearly false). And that prompted some analysts to ask why the Bank cut rates if they expected inflation to be above 2.5%.

As a result, the RBA had to rush out the Deputy Governor to give an interview to Bloomberg in which he explained that if they hadn’t cut rates, their inflation forecast would have been fractionally below the mid-point of their target band. People often laugh about Charles Dicken’s character Mr Micawber who declared that the difference between ‘happiness’ and ‘misery’ was whether one’s annual spending was a sixpence more than one’s annual income. The RBA is now in a “double Micawber” – not only is an inflation forecast fractionally above 2.5% a cause for misery, but so too is one fractionally below 2.5%.

While this is a specific instance where the reliance on modelling begat more questions than answers, this point is generalisable. The British statistician, George Box, famously said that “all models are wrong, but some are useful”. Unfortunately, using models to explain a central bank’s policy actions to the general public, let along financial market participants, is the opposite of useful. This is because econometric models rely upon a whole host of assumptions about specific parameters which can’t be observed or verified ahead of time. These include the NAIRU, the neutral rate of interest and the output gap. Nor do they incorporate a role for President Trump’s tweets.

To understand the RBA’s view on these concepts we quote the Deputy Governor, again from his interview with Bloomberg. In that interview, he said: “…there’s a chart in the Statement on Monetary Policy that shows the range of estimates of the neutral rate, and it runs from something like 1 to 4. Now, 3 percentage points of difference in such an important macroeconomic variable is pretty useless, to be honest.”

His point was that people shouldn’t pay too much attention to it. But by the RBA mentioning it, it has the opposite effect; immediately analysts start engaging in arcane arguments about where the policy rate is compared to the neutral rate and where unemployment is compared to NAIRU. If the Deputy Governor of the RBA thinks this is “pretty useless” then I suspect the RBA Board members that actually decide where the policy rate is set are not kept awake at night wrestling with the question of whether r-star is 1% or 3%. But if it is a ‘pretty useless’ concept, it makes sense for the RBA to stop discussing it. If inflation has fallen from 7% to 2.5% surely that is sufficient evidence that interest rates have been high enough for long enough. In other words, it is time to liberate the RBA from the explanatory straight-jacket that the Review has forced upon it.

Download PDF