Some people remain mystified by inflation, even some of those who used to be a deputy governor of the Australian central bank, the RBA.

A week ago, Stephen Grenville, former RBA deputy governor, had a piece published in the Australian newspaper. The headline writer called it ‘The mystery of inflation’. It has taken me longer than normal to write this response, mainly because it took me time to recover from falling off my chair when I initially read it. Grenville’s premise is that central banks and economists failed to forecast the inflation to come out of the COVID era. True, but not all economists failed to forecast it – those that read this blog would know that. Grenville objects to the remedy proposed by many of the failed forecaster class, that of better modelling. You can’t model what you don’t understand, he says. True again. But then he goes off the rails.

Grenville believes that price setting behaviour has changed and that inflation forecasting cannot improve without first understanding modern price setting dynamics. For example, he says inflation expectations were different during the COVID era compared to the 1970s and that must have changed the process, he asserts. Well, expectations do change but expectations are irrelevant to actual prices, as I have explained before.

Secondly, he says there is more differentiation in goods, services and pricing methods (without specifying any evidence for this assertion) and that must, he asserts, change price setting behaviour. He even said consumers get more free goods, citing social media, government services and consumer credit as examples. That this statement can even be made by a former RBA senior executive beggars belief. Anyway, he said it.

Finally, he says technology-enabled buying options (think online shopping) has changed consumer options and therefore, he asserts, price setting behaviour.

Grenville then proposes his solution. The data rich environment must be harnessed by the Government institutions, the central bank, the statistics bureau and the anti-competitive practices watchdog. They must analyse the micro data to better understand price setting dynamics in the modern world. How depressing. These views expressed by Grenville are ignorant. And of course, what he left out of the piece is just as illuminating.

I will make three points in rebutting Grenville. Firstly, price setting dynamics have not changed. Nor will they. Price setting dynamics are driven by human nature, not by new online shopping options. Buyers want to improve their lot (their utility, satisfaction). So do sellers. They will trade if and only if two conditions are met: a) the buyer values the good/service more than the money that would have to be given up to make the purchase; b) the seller values the additional money to come from a sale more than holding on to the inventory, that is the unsold good/service. Both conditions must be satisfied. If the buyer is willing to buy at a lower price but the seller isn’t willing to sell, the transaction will not occur. If the price can be adjusted through bargaining to the point where both parties will have their value preferences met the transaction will occur.

Buying and selling is always and everywhere a transaction at the margin in the here and now. Web buying technology doesn’t change that. It might give the illusion of extra options to the consumer, but the underlying dynamics are the same. More free goods (I still can’t believe he said that) doesn’t change that. Consumer preferences are satisfied under the conditions that the consumer wants to satisfy their most urgent preferences first, and they want to do that sooner rather than later. The same goes for sellers.

Human nature explains the price setting process. Does technology change that by making price discovery easier? No. What Grenville demonstrates is the fallacy of aggregation thinking applied to individuals. While in an economy as a whole, price discovery is made very easy thanks to technology, buying decisions are made by individuals, not by the economy as a whole. Individuals buy relatively few items compared to the range of goods and services in the whole economy. Consumers are limited both by funds and time in their ability to consume. So weighing up the consequences of buying a holiday vs a car vs further education is not hard. No amount of additional data in a cloud is going to make that decision any different.

Secondly, Grenville’s assertion that modelling micro data will reveal wisdom is doomed for the same aggregation fallacy. Massive amounts of data and computing power cannot reveal why certain transactions happened. They can only show that the transaction did happen, but not why. Now, we know that each transaction happened because both parties to the transaction decided it was in their interests to proceed. However, two other individuals could have decided that the transaction was not in their interests. The transactions that did not happen are just as important as the transactions that did happen, yet they are not in the data. Grenville can model micro data all he likes but by ignoring equally important outcomes of other relative value assessments that go to the heart of price setting will mean his analysis will be worthless.

Thirdly, in his confusion about inflation, he failed to mention the role of Government monetary policy. Perhaps he thinks all inflation is driven by price setting between buyers and sellers. If that is the case, little wonder he is confused. Price setting through buyer and seller dynamics is not an inflationary process. It merely sets relative prices, not the overall price level.