What is a “per capita recession”? Are we in one? And does it matter?

Posted on March 11, 2019 by Pete Faulkner - GDP, Population, QLD, State Accounts, Uncategorized

Last week we saw a weak set of National Accounts for the final quarter of 2018 which showed growth of just 0.2% q/q, 2.3% y/y and 2.8% 12 month average (commentary here). As this data was released we also started seeing a lot of talk (primarily on Twitter initially before it spilled over into the main-stream media) of a “per capita recession”, a term rarely, if ever used before. So what is a “per capita recession”?

A recession is generally defined as when GDP contracts for two consecutive quarters; a event that we haven’t seen in Australia since the second quarter of 1991. If we divide GDP by the resident population then we get a measure of GDP per capita. A “per capita recession” therefore, is when GDP per capita falls in two consecutive quarters. In the third and fourth quarters of 2018 GDP per capita did indeed fall (0.1% and 0.2% on the ABS seasonally adjusted basis) taking Australia into a “per capita recession”. The opposition, and many commentators (particularly, although not exclusively, those with links to the ALP), started making a good deal of noise about this. But how big a deal is it and does it actually tell us much?

Firstly, as Nick Behrens at QEAS has pointed out so clearly, economic growth is driven by three elements; population, participation and productivity. To somehow suggest that growth isn’t really growth just because it’s been driven by population increases makes little sense.

Secondly, does this measure really tell us anything useful? GDP per capita may be a useful measure of living standards, although it is by no means the only one. A more often quoted indicator of living standards might be the real net national disposable income per capita; this measure actually had two consecutive negative quarters in Q2 and Q3 2018 but was quite strongly positive (+0.8% q/q) in this latest data. I certainly don’t remember a lot of commentary about a “per capita recession” earlier last year.

To give some idea of the usefulness of the “per capita recession” concept Nick highlights in his post the reality that, based on annual ABS Gross State Product (GSP) data, Queensland was in just such a recession from 2009 through to 2011 and then again in 2015. I’ve taken that idea a step further by analysing the quarterly Gross State Product data from the Queensland Treasury. What that shows s that in the period since 1990 Queensland has experienced 10 such “per capita recessions” and yet just 4 actual recessions.

Looking at the chart below it’s clear that two quarters of negative GSP per capita is certainly not a guarantee of an actual GSP recession. A similar comparison for the national data (see chart above) shows “per capita recessions” in 2000 and 2006 (before the current example) neither of which coincided with, or preempted, an actual recession. Indeed, the weakest period of GDP growth immediately following the GFC wasn’t reflected in a “per capita recession” either (at least not on seasonally adjusted data. Trend data did show such an event).

What we can certainly say is that a “per capita recession” probably indicates a period of weak growth. The current example has lead to some focusing their attention on the fact that the weak growth we are currently seeing is indicative of weak productivity gains and that the country s in need of serious reform to help lift productivity improvements. If all the talk about a “per capita recession” does nothing other than focus more attention on that issue then it’s probably not been a bad thing.

It’s probably also worth pointing out that revisions to both GDP and population data are common and could change this picture. The ABS will be releasing their latest population estimates next week for Sept 2018. Previous GDP revisions can be quite dramatic. For example, the third quarter 2016 GDP was originally flagged as a 0.5% decline for the quarter; that has subsequently been revised to a 0.01% increase.