No one seems to know quite how it started, but ‘two quarters of negative growth’ has become the widely-accepted definition of a recession in Australia. Using that definition, it’s impossible to diagnose a recession in a state (unless it lasts for a year or more), because the quarterly National Accounts don’t tell us how much each state’s economy has grown or shrunk.

The ABS does give us a quarterly update on ‘state final demand’, which isn’t quite the same thing. It doesn’t include the goods and services that a state has produced for export, either to other states or overseas. For some states (like WA), omitting exports can matter quite a bit. After last quarter’s National Accounts, I explained why I think that using state final demand figures to diagnose a recession is misguided; Ross Gittins has made the same point.

Despite this, Tim Colebatch declared last week on the basis of the final demand figures that “Western Australia has joined the other AFL states in recession.” The Treasury Secretary referred to this and other such reports as “hysteria,” with David Gruen adding that:

The idea that, in the face of the largest investment boom we have ever seen, you ignore exports and focus on that piece of the economy that is demand by Western Australia and claim that that is a recession belongs in the comic books, not in serious newspapers. 

I agree that final demand is far from sufficient to diagnose a recession. But the output figures – Gross State Product – are only published once a year, with a lag, so they aren’t much help to someone wanting to get a read on how a state’s economy is performing.

For that reason, I like to look at the labour force figures, as Mark The Graph did here. Okun’s law (really more of a rule of thumb than a law) suggests that there’s a reasonably stable relationship between output growth and the change in the unemployment rate. Here’s what that relationship looks like for WA:

wa okun

Source: Calculations based on ABS 5220 and ABS 6202. The unemployment rate is the annual average in original terms. Each dot represents a financial year – ’91’ is 1990-91.

This graph tells us that over the past twenty years, WA’s unemployment rate has tended to fall when its economy has grown at 4.1% or higher. When growth is less than 4.1%, unemployment has tended to rise. The state’s economy hasn’t experienced an annual growth rate of less than 1.7% in this period (even in the 90s recession, which surprised me), but this chart suggests that if the WA economy didn’t grow, we’d see the unemployment rate rise by around 1.5 percentage points. You can also see why I described this as a ‘rule of thumb’ rather than a ‘law’ – variations in the state’s growth rate explain less than a third of the change in the unemployment rate (unlike at the national level, where R2= 0.6).

So is WA in recession? Well, the unemployment rate (in trend terms) has risen by 1.1 percentage points over the past year. That would suggest that output has grown by about 1.1% over the same period – that’s sluggish, but not a recession. But the pace of decline in the WA jobs market has accelerated, with the unemployment rate rising by 0.7 points in the six months to April. If we annualise that, it’s a 1.4 point rise in unemployment, which is very close to what you’d expect from the WA economy if it isn’t growing at all.

On this basis, my answer to the question of whether WA is in a recession would be “no, but it’s getting too close for comfort”.