A common theme emerging from the national accounts on Wednesday was a chorus of calls to improve Australia’s labour productivity performance.
That followed reports last week that Reserve Bank of Australia (RBA) governor Glenn Stevens, when appearing before a parliamentary committee, had called for an overhaul of industrial relations laws to restore growth in labour productivity.
In fact, he pointedly referred to multi-factor productivity – the output per unit of labour and capital (machinery, buildings and the like) – combined, and a long list of things that need attention, only lightly glancing on industrial relations.
But in the current political environment, it’s the raw productivity of labour – output per hour worked – that’s the focus of attention.
The standard measure of labour productivity, gross domestic product (GDP) per hour worked, is useful.
But so are guns, razor blades and petrol.
They have to be handled with care.
Which brings us back to the June quarter national accounts.
GDP per hour in the June quarter was right where it was a year before, after being virtually unchanged the year before that.
In other words, labour productivity has stopped rising.
Over the years, economists have come to expect something between one and 1.5 per cent annually.
Clearly, something is very wrong.
Or is it?
Industry-by-industry estimates of output per hour worked are hard to come by, but if you really want to, you can cobble them together using labour force data available from the ABS.
And the national accounts show gross value-added for major industrial sectors, and hours-worked figures – albeit not seasonally adjusted and for the mid-month of each quarter – can be found in a dark and distant corner of the bureau’s website.
And it’s when you look at productivity industry by industry that things begin to get interesting.
In the mining sector, the figures show total hours worked rose 17 per cent over the past year.
But output – gross value-added – fell nine per cent.
That translates into a fall in real output per hour in the mining sector of 22 per cent.
And it’s no aberration, either.
In the past decade, hours worked in mining have soared 167 per cent – nearly tripled – while the volume of output has grown by only 17 per cent.
In other words, labour productivity in mining has fallen 56 per cent – more than halved – in the past 10 years.
That’s to be expected when less productive mines suddenly become profitable and require more labour to dig up the lower-quality, deeper ore.
The mining firms are making bucketloads of money, so we don’t need to feel sorry for them.
It’s what this implies for the rest of the economy that’s more interesting.
Take mining out of the GDP estimate, then take hours worked in mining out of total hours worked, and the picture looks much brighter.
Rather than no growth in labour productivity in the latest year, once mining is excluded the numbers show growth of 1.9 per cent, well above the long-run average.
The same analysis of the non-mining economy does show some slowdown in growth in output per hour, to an annual average of 1.1 per cent over the five years to June compared with 1.6 per cent over the five years before that.
But faster GDP growth typically means faster productivity growth, most likely because existing workers are made busier and expansion quickens the rate of introduction of more efficient technology.
And with slower growth, such as we have seen recently, it works the other way.
So some slowdown in labour productivity growth would normally be expected, given the slowdown in non-mining GDP growth, to 2.6 per cent from 3.6 per cent between the same two half-decades.
None of this downplays the value in the ongoing effort to find new ways to improve productivity.
But the argument that economy-wide productivity growth has suddenly stopped in its tracks – whether due to rigid industrial relations laws, lazy management or anything else – is given no solid support by the hard data.
