Judith Sloan has written another denunciation of the Fair Work Act in which she makes a number of claims about the intent and effect of this apparently radical piece of legislation. Despite the fact that it’s the sort of topic that makes most eyes glaze over, I’d like to focus on her claim about productivity:
There is the rather awkward contrast between what Julia Gillard predicted – “the best thing about Labor’s industrial relations plan is that it will be good for productivity” – and the actual outcome on productivity, which is pretty much in free fall. Even the head of Treasury has been banging on about the importance of lifting productivity.
Bold claims from politicians about the effect of industrial relations changes on productivity growth are nothing new. When he was introducing the Work Choices bill into Parliament, Kevin Andrews said it was intended “to encourage the further spread of workplace agreements in order to lift productivity and hence the living standards of working Australians”. [fn1]
Was he right? Was Gillard right? It’s hard to know, and in some ways it’s too soon to tell.
It’s extremely difficult to measure the effect that a given policy change has on productivity. First, productivity growth is best measured over a number of years, as an average rate over the course of a cycle. Second, our productivity growth has been suppressed to some extent over the past decade or so by the big capital investments underway in the mining and energy sectors. Big projects that are under construction soak up a lot of inputs (capital and labour) without a lot of output to show for it in the short term; this reduces the rate of productivity growth relative to where it otherwise would have been.
The third problem with measuring the effect of a particular policy change on the rate of productivity growth is that our small, open economy is quite heavily affected by what happens around the world. When there are big technological developments, such as in IT, the productivity growth rate in developed countries tends to rise more or less in unison. When the rate of technological progress slows, that tends to be felt as a fall in the rate of productivity growth across the developed world. That is basically what has happened in recent decades; our productivity soared in the 90s, as did much of the rest of the rich world, but then grew more slowly in the 2000s, a slump that was common elsewhere. [fn2] Domestic policy does affect productivity, but not nearly as much as is sometimes claimed. Assessing precisely how much of our productivity performance can be ascribed to public policy, much less an individual piece of legislation, is an exceedingly difficult task.
Despite all this, we can get a sense of how productivity is changing from quarter to quarter via the National Accounts. The ABS reports the total output of the economy (GDP) and the total number of hours worked. Put these together and you get GDP per hour worked, a key measure of labour productivity.
A big problem with this headline measure is that it includes the public sector, where productivity is notoriously difficult to measure. Indeed, statistical agencies often just assume that the level of productivity remains unchanged over time in the public sector, so that a change in output equals the change in inputs.
For this reason, the ABS separates out GDP and hours worked in the “market sector”, where productivity can be properly measured. This is also the best place to look to see any effect of industrial relations changes on productivity; after all, the state public sectors (bar Victoria’s) are still in their own IR jurisdictions, unaffected by the Fair Work Act or its predecessors.
What do the data show? In the first 18 months of the Fair Work Act, productivity in the market sector grew by 1.9%. This was no different to the 18 months before that, when market sector productivity was also up by 1.9%. In the first 18 months of Work Choices, it grew by 1.4%. [fn3]
I wouldn’t make too much of the under-performance in that first 18 months of Work Choices. After all, as I said, productivity is best measured on average over the course of a cycle. Nevertheless, those figures don’t exactly suggest there has been any great productivity slump in the Fair Work era.
What, then, has prompted Sloan’s claim? It must be the fall in productivity in the seventh quarter of the Act’s operation, the March quarter of this year. Indeed, productivity did fall in that quarter, quite sharply. Market sector productivity fell by 2% in just one quarter, wiping out the gains of the previous 18 months.
This is clearly not a good thing. Productivity growth enables growth in real living standards; it’s in no one’s interest for productivity to fall. However, just looking at this one macroeconomic aggregate in isolation gives you an unnecessarily alarmist picture of what’s happening in the Australian economy.
Remember that labour productivity, the figure I’ve been using above and the figure to which Sloan is presumably alluding, is measured as GDP per hour worked. What happens, then, if GDP falls and employment doesn’t? You get a fall in productivity.
This is what happened in the first quarter of this year. Remember that GDP suffered its largest one-quarter fall since 1991. This was driven by theQueenslandfloods and other natural disasters, particularly their effect on coal exports. The collapse in exports shaved over two percentage points off our quarterly growth rate.
Despite this fall in GDP, employers chose not to shed their staff in a mass panic in March. The fall in output had its biggest effect on large exporters and disaster-affected regions. Much of the economy was still ticking along; consumption grew by a reasonably solid 0.8% in the quarter. Even where businesses were struggling, perhaps they could see that this was a temporary downturn driven by extreme weather, and that things would turn around in later months, so they didn’t want to sack their staff.
Whatever, the reason, the number of hours worked in the economy kept rising in line with its trend, even while output tanked.
This is what Sloan and others are railing against. If output falls, but hours don’t, productivity must fall as a simple matter of arithmetic.To castigate the Government and its industrial relations legislation for a single quarter’s productivity performance is extremely opportunistic and somewhat cynical. Would Sloan have preferred that employment fall by 1.2% in a quarter, just as output did, all in order to maintain the productivity level? Does Sloan suggest that the fall in GDP, driven as it was by a collapse in exports, was caused by the Fair Work Act? I can’t imagine that that’s what she is suggesting.
I’ll grant her the benefit of the doubt and assume, instead, that Sloan is referring to the long-run decline in the rate of productivity growth. This is certainly what Martin Parkinson was alluding to in the speech that Sloan mentions; his speech includes an analysis of productivity growth over various cycles, the most recent being 2003-04 to 2007-08.
The problem is that this story, of a long-run decline in productivity growth, doesn’t fit Sloan’s highly ideological characterisation of the Fair Work Act.
The decline in the rate of productivity growth is a long term story. Growth peaked in the mid-1990s and has been falling ever since. As I put it in response to another ideological tirade against the Fair Work Act,
The rate of productivity growth has been declining… through the era of rottweilers on the wharves, declining through the era of near-compulsory AWAs in higher education, the public service, and Government-funded construction projects; declining through the era of a building tribunal with coercive powers; declining through the era of AWAs that could undermine award conditions; declining through an era in which minimum wage setting was given to a new body entirely unsympathetic to the idea of minimum wages’ very existence.
Why blame an Act that has been in effect for only two years for a productivity slowdown that began in the decade before last?
[fn1] hat tip to Alex White
[fn2] Productivity growth has been stronger in the past year or two in other developed countries. This is another way of saying that they’ve experienced jobless recoveries from a deep recession.
[fn3] I’m using seasonally adjusted figures, from ABS 5206.0, table 1.