It’s probably not worth responding to Tony Abbott’s ‘column’ in yesterday’s Herald, except to critcise the newspaper itself. Plenty of people have commented on how completely inappropriate it is to publish these thoroughly partisan polemics as opinion. It’s one thing to reproduce a speech, or parts of it, on the news pages, but opinion columns should be reserved for commentators who assert at least some claim to independence against which they can be held accountable. This sort of thing amounts to free advertising.
Abbott ridicules Rudd for getting his analysis of recent productivity developments wrong. The scorn is probably deserved, though I concede I haven’t had time to look into everything Rudd said. The main point is that Rudd hasn’t done himself any service by trying to capitalise on short term productivity changes: it’s as stupid as pointing to one summer’s heat wave as evidence for global warming. Rudd’s basic story, put forward in his budget reply, is sound: long run productivity growth has fallen this century, and more investment in education and training should raise it again. Short run changes in productivity — that is, from one year to the next — are mostly due to cyclical factors. Labour productivity (as distinguished from multi-factor productivity) is, after all, just the ratio of GDP to hours worked. If those two things don’t grow at exactly the same rate, productivity’s growth rate goes up or down.
Rudd should never have talked about short term movements, first because they’re not significant, and second because they’re tricky to analyse, as you will see if you persevere. As he seems to be able to learn lessons, we can hope he’s learnt this one, although I’m afraid he won’t get rid of the team of advisers that keeps getting him into trouble with their bullying ways and cynical attitudes.
Notwithstanding that there’s some truth in Abbott’s accusations, it needs to be pointed out for the record that Abbott doesn’t know what he’s talking about either. To make sense of what follows, refer to the graph over the fold, which shows a stylised business cycle. Note that there is no upward trend in any of the variables, because we are interested purely in the cyclical factors. The basic assumption is that employers adjust their workforce to their production needs with a lag.
If changes in output and employment were always synchronized, there would be no cyclical changes in productivity. But employment changes (even when measured in hours) lag output changes. In the phase when output and employment are high (call this the boom), output at any moment is growing more slowly than employment (rising more slowly before the peak and falling faster afterwards); so productivity rises. The converse applies to the slump. Alternatively, note from the bottom diagram that productivity growth is negative in the boom (output growth exceeds employment growth), and positive in the slump.
If you happen to be interested in whether the growth rate (as opposed to the level) of productivity is rising or falling, note that it rises in the downswing and falls in the upswing.
Now, let’s see if Abbott makes any sense. His first relevant comment is:
Unfortunately for Rudd, his advisers’ analysis confirms that slower productivity growth is the result of fast employment growth, not government neglect, and that sustained productivity growth is now likely.
But for this story to make sense, it’s not enough that employment grows fast: it has to do so faster than output. And for this development be a specifically cyclical phenomenon, it needs to have happened because output growth itself is slowing down. Assuming Abbott understands this, does he think it’s because we’re heading into a slump, or because the economy is normalizing after some abnormally rapid growth burst? There’s no evidence that either of these explanations applies.
The rest of Abbott’s analysis is all in one paragraph:
Students of basic economics, which Rudd seems never to have taken, know that productivity growth is normally high when unemployment is high because fewer workers nearly always produce more per head.
This is exactly wrong. Refer to the diagram, and you see that in the boom (which is when unemployment would be low relative to the average) productivity growth is low. From the reasonning he supplies, I suspect he means productivity itself, not its growth. But if he’s going to attack Rudd for not understanding basic concepts, he needs to be held to the same account.
Conversely, productivity growth is normally low when unemployment is falling fast because it takes time for production to increase.
First note that that ‘unemployment falling fast’ is not the converse of ‘unemployment high’. Loosely speaking, in some contexts, you might get away with this, but not in a technical discussion like this. Unemployment is actually falling fastest right at the beginning of the boom, and at this point (cyclical) productivity growth is around zero but falling.
Temporarily lower productivity growth is a function of high employment growth, not a reason to support the collective bargaining arrangements that kept unemployment up.
It’s not clear what he means here by lower productivity growth. Lower than what? But if he is talking about the nadir of productivity growth, that corresponds to the peak of the boom. This is not in fact the point of highest employment growth at all, but the point at which employment growth ceases.
Of course, Abbott does not imagine for a moment that any reader will attempt to make sense of these purely mathematical points, let alone examine the data to see whether they actually apply to present circumstances. The sole purpose is to give the impression that he knows what he’s talking about (or at least that his staffers do).
He finishes with a quote from Rudd which he correctly characterises as Beasley-like blather. But his own ‘analysis’ above shows that blather doesn’t always sound like blather: it can be dressed up to look quite coherent and precise.