Difficulties of economic projection
Updated: Jan 25
Economists often present estimates (guesses) of future economic events. Smart economists present their guesses in general statements or, at least, like weather forecasters, with different outcomes identified by different probabilities.
At times of dramatic happenings point estimates are likely to be dramatically wrong. Terry McCrann in today’s Australian reminds us of recent embarrassing forecasts of the projectors of the Reserve Bank and other official entities.
‘A year ago we didn’t have a clue about what would prove to be in store for us through the course of 2020. Do we really have even the slightest bit more accurate idea about 2021?’
As Australia approached Australia Day a year ago, a month earlier Josh Frydenberg had ‘Proudly, confidently predicted the first budget surplus since Peter Costello’s last in 2007.’ The outcome was a deficit of around $85 billion.
‘That mid-year budget update was also predicting growth in the economy of 2.25 per cent over 2019-20. The actual figure would prove to be a thumping 6.3 per cent decline’. Treasury, the RBA and the Treasurer will all say ‘facts changed dramatically’ but the changes are so dramatically wrong that one must ask is there a better way?
Share prices also surprised. Not this writer, who sold many shares three years ago and has been waiting for a large contraction. (This point is a confession, not a gloat.) In March 2020 US and Australian shares both fell by 37 % but still I failed to re-enter the market. (Still waiting for a serious fall in share prices.)
Property prices in Australia ware also predicted to fall, but after a brief hiccup the boom continued. But no increase in cash interest rates that some theories once suggested, specifically those of an eminent American economist and his student Dr Philip Lowe, who has turned out to be a rate cutter rather than a rate hiker.
To return to guru McCrann: ‘Apart from the overall message – at this point, statement of the bleeding obvious – that 2020 was unpredictable in anticipation, it largely proved unpredictable even through the year when we supposedly knew what we were dealing with and how we were responding’.
‘Simply, obviously, the virus. And then, the massive fiscal stimulus – here and everywhere – and the QE/MMT money printing and zero interest rates, again everywhere and also here’.
Here is the killer point. ‘My core suggestion is that nothing has changed to render 2021 predictable in the way 2020 clearly was not. In 2020 it was the virus. In 2021 it is the vaccine’.
‘Just as a year ago there was a sort of casual, all persuasive, blind spot to the developing reality that the virus would totally define and determine 2020, there is exactly similar, but actually opposite all-persuasive casual assumption about the vaccine’.
Terry McCrann only risks two broad scenarios for 2021.
* ‘The first, less likely, is that [the vaccine] does work in the crucial economic sense relatively quickly. That would combine with all that stimulus to generate an economic boom way stronger than expectations, while yet further inflating - (already over inflated , bubble territory?) - asset values: shares and property.
* The more likely scenario is that we don’t get that. This would give us a 2021 that was a, hopefully milder, replay of 2020. In short a combination of anti-virus lockdowns/travel bans and fiscal and monetary responses – starting obviously, from zero rates, existing massive QE and massive deficits’. (Caveat – perhaps this scenario would also lead to the same (over?) inflated asset values.)
Terry McCrann ends with a (to me) rather enigmatic comment on China’s predicted economic growth, with 83rd ranking on the list of virus control.
Anyone wishing to be an economic forecaster, or anyone currently practising the arcane field, should consider Terry McCrann’s advice to consider scenarios. And my advice to try to add probabilities to scenarios like humble weather forecasters.
Read the full discussion here.
Fiona Prior sees ‘DORR-E DARI: A POETIC CRASH COURSE IN THE LANGUAGE OF LOVE’. More here.