How unusual is Australia’s record without “recessions”?

Over the weekend a reader posted a link to a blog post by a couple of researchers at the St Louis Fed.  People often point out –  sometimes to boast (if you are an Australian politician), sometimes just out of curiosity –  that Australia has gone 28 years with “a recession”, where “recession” here is defined by that convenient (not overly helpful) benchmark of two successive quarterly falls in the level of real GDP,

As a definition of recession it seems to serve mostly because there is no better definition in general use. In the US, there is the NBER business cycle dating committee, which draws on a range of different indicators to reach a judgemental determination of the dates of recession, but those dates are typically only available with a lag, and there isn’t anything similar in most other countries (including New Zealand).

The rather obvious, but all too frequently forgotten, point that the St Louis Fed observers are making is that any sensible interpretation of GDP growth has to take account of the rate of population growth.  There are countries in eastern Europe where the fall in population has been large enough that one could see GDP falling every quarter and yet the average citizen would still their per capita GDP rising.  Even setting aside that extreme, there is a big difference between, say, Germany (with almost no population growth this century, Japan which has had a slight fall in its population in the last five years, and New Zealand with population growth in excess of 10 per cent over the same period.     Australia is another advanced country with rapid average population growth.

As the St Louis Fed researchers point out, once you focus on per capita real GDP the chart for Australia looks a bit different

st louis fed 1

On the “two consecutive negative quarter” rule, Australia would have had three (fairly mild) ‘recessions’ in the period since the end of the early 1990s recession.

In case you are wondering, New Zealand (using the average of our two GDP measures) would also have had three “recessions” (rather more severe) on this per capita metric: 1997/98, 2008/09, and 2010.

It is still fair to note that even on a per capita basis, Australia did not have a recession in real GDP in 2008/09.  As far as I can see that makes it the only OECD country to have avoided two consecutive quarterly falls in real per capita GDP over that period.

But (as the NBER business cycle dating methodology recognises) simply looking at GDP isn’t a particular sensible basis for assessing whether (more broadly) things are going backwards.  Personally, I find looking at the unemployment rate quite useful.   All else equal, if the unemployment rate rises the economy is generating fewer new jobs than the increase in the number of people ready and willing to work.  As with any measure there is some month to month and quarter to quarter “noise” (often just measurement challenges), but if the unemployment has been rising for a couple of quarters (especially in a climate where the long-term trend is downwards) that too is probably a suggestion of an economy going backwards where it counts.

Here is the quarterly series for Australia’s unemployment rate back to the early 1990s.

Aus U to sept 19

Over that period there were seven episodes where the unemployment rate rose for two or more consecutive quarters.  Most of the increases were small, but there were three episodes where the increase was 1 percentage point or more.

Over the same period, New Zealand had four episodes where the unemployment rate rose for two or more consecutive quarters, two of them by a percentage point or more.  The UK and the USA had three and two episodes respectively –  including the very sharp increase in the US unemployment rate at the time of the 2008/09 recession.

It isn’t even as if the extent of the rise in the unemployment rate in Australia over 2008/09/10 was unusually small.  Australia’s unemployment rose less than in the other Anglo countries, but on OECD numbers there were about 10 European countries where the unemployment rate rose by a similar amount (mostly) or even less than in Australia.

Another variable I find worth looking at for Australia is a measure of real income that takes account of terms of trade fluctuations (for commodity exporters much of any adverse shocks show up in price rather than volume, whereas for manufactures and services exporters the balance is the other way round).

This is a chart of real net national disposable income per capita series, produced by the ABS.

RNNDI to 19

It really quite a startling record of steady growth in this series from about 1993 to about 2008.  But the period since then has been quite different (a point that various of the more downbeat domestic Australian commentators often point out – real Australian incomes have not been doing well).  There have been seven periods since 2008 when RNNDI per capita has fallen for two or more consecutive quarters.    Somewhat to my surprise, when I looked at the closest New Zealand series there had only been two such falls since the early 1990s.   The fall in Australia’s per capita RNNDI over 2008/09 was a actually larger than the fall in New Zealand’s per capita real GDI measure.

The St Louis Fed researchers ended their post this way

So should we use Australia as a benchmark when thinking about possible duration of expansions? If so, we have to take it with a grain of salt because looking at just GDP growth doesn’t paint the whole picture. It is important to look at per capita GDP growth to have a broader view.

Their goal wasn’t to bag Australia but to put its experience in some perspective, specifically the importance of taking account of population growth trends when looking at GDP numbers and headline about presence or absence of GDP “recessions”.

This post is in much the same spirit.   I really like Australia and can’t stand the “chip on our shoulder” too many New Zealanders seem to have about the place.  It is different from most other OECD countries –  heavy resource dependence does that, as does rapid population growth – and that needs to be taken into account in comparing cyclical economic performance.      Through some mix of good luck and good management –  mostly the latter in my view (including choices around a floating exchange rate, low and stable public debt, and keeping the state out of the housing finance market) Australia has avoided anything like the worst downturns (whether per capita GDP or unemployment) seen in some other OECD countries (eg the US and –  even more savagely – a number of euro-area economies).  But it is a normal economy, it has upswings and downswings: if no one else the unemployed (and the underemployed) know it (relative to, say, the end of 2007, the median OECD economy now has an unemployment little changed from then, while Australia’s unemployment rate is almost 1 percentage point higher than it was then).

Incidentally, and while on the topic of per capita GDP growth, I had a look at the latest annual growth rates for New Zealand and the 30+ countries for which the OECD has data.  Despite all the talk  –  including from the PM –  about New Zealand doing better than its peers, our real per capita GDP growth in the year to the June quarter (or four quarters on four quarters) was just lower than the median country in the sample (and quite a bit below the unweighted averages of those countries).  Better than Australia over that year (see the first chart) but then –  as this post and the St Louis Fed one help illustrate –  Australia is no stellar performer.

 

6 thoughts on “How unusual is Australia’s record without “recessions”?

  1. I do have issues with the unemployment rate or at least the variables that determine an unemployed. Historical comparisons are of no use because we went from counting real people to an estimate. We have 3 people doing what one full timer with overtime would do. We have employees gaming their hours under the working for families which creates more short shift hour positions, etc. It would all be fine if it didn’t create more transaction costs and lower productivity. We can then get on to under utilisation of skills, participation rates and poverty pockets and ghettos where stats don’t penetrate. I could go on but to think the unemployment rate is a good indicator is a stretch.

    Like

    • Thanks. I didn’t because my focus here wasn’t on predicting recessions (which is what Sahm is trying to do). The sort of lift in unemployment rate she uses in that “rule” is probably the sort of thing one would expect to see in a genuine downturn, but probably not as a leading indicator (historically the unemployment rate has lagged output/income a bit).

      Like

  2. Michael,

    Yes, she would like to use these numbers in a predictive sense to trigger some direct stimulus payments to individuals. But it can also be used in a historical sense: “did we have a recession in a particular period for example. In fact, her rule captures the NBER dates very well, despite my initial reaction that unemployment rate is a lagging indicator. Numbers suggest that Australia had a mild downturn in 2001-2, and a significant one in late 08-09.

    Like

    • Yes, fair enough. And would no doubt show up the NZ downturns in 97/98 and 08/09 (but not the often forgotten 2010 double dip). For history, something akin to an NBER cycle dating committee in each country would be great.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s