The Productivity Commission’s five-yearly report on our productivity performance seems to have sunk like a stone but, before it disappears without trace, it has one important thing to tell us: the obvious reason productivity improvement has slowed, and why, ceteris paribus, it will probably stay slow.
Economists like trying to impress people with Latin phrases. Many conclusions in economics depend on the assumption of ceteris paribus – all other things remaining unchanged. Economists are always holding all other variables constant while they see what effect a change in variable A has on variable B.
Trouble is, in the real world, all else almost never stays unchanged. In which case, the relationship between A and B that you thought you could rely on has been stuffed up by some other variable or variables between C and Z.
Back to the point. Everyone thinks they know what productivity means, but they often don’t. The commission’s report says productivity improvement is “the process by which people get more from less: more and better products to meet human needs, produced with fewer hours of work and fewer resources”.
“In many cases this growth occurs with lighter environmental impact” – a truth many scientists just can’t seem to get their heads around.
The report says that over the past 20 years, the rate of improvement in productivity has slowed in all the rich countries, but with Australia slowing more than most.
Why? Many reasons, no doubt, but one big one that ought to have been obvious, since the American economist William Baumol noticed it in the 1960s.
The fact is that most improvements in the productivity of labour come from advances in technology. You give workers better, “labour-saving” machines to work with, which allow them to produce more in a typical hour of work.
(The other big one is giving workers more education and training, which allows them to work more complicated machines – including computers and software – design more complicated machines and programs, and service complex machines.)
Trouble is, it’s easier to improve productivity in some industries than others. In particular, industries that produce goods – on farms, in mines and in factories – can, and have, hugely increase their productivity by mechanising and computerising. Same in utilities, transport and communications.
In the production of services, however, it’s much harder. Although some services can be delivered digitally – streaming video, say – with little involvement by workers, most services are delivered by people, from less-skilled services delivered by waiters, cleaners, bedmakers and shop assistants, to highly skilled teachers, nurses, doctors, lawyers and prime ministers.
You can give these workers a car or a mobile or a screen, or give a hairdresser a better pair of clippers, but there’s not a lot you can do to speed them up. As Baumol famously remarked, it takes an orchestra just as long to play a symphony today as it did in 1960 – or 1860.
After two centuries of playing this game, we’ve ended up with goods industries that are highly “capital-intensive” – lots of expensive equipment; not many workers – and service industries that are highly “labour-intensive”: many workers; not much equipment.
Which means the productivity of labour is sky-high in the goods sector, but not great in the services sector.
But here’s the trick. You might expect that wages will be much higher in the high-productivity goods sector and much lower in the low-productivity services sector. But no. Wage rates do vary according to the degree of skill a worker possesses, and on the demand for that particular skill.
But a cleaner in a factory gets paid pretty much the same as a cleaner in a lawyers’ office. And a doctor gets much the same working in a big factory’s clinic as in a hospital.
Why’s that? Because, if an economy is working properly (which ours isn’t at present), it’s the economy-wide improvement in productivity that tends to increase all real wages by about the same percentage.
This is brought about by market forces. Despite their low productivity, employers in the services sector have to pay higher wages to stop their workers moving to higher-paying jobs in the goods sector.
Remember too, that over time, mass production lowers the prices of manufactured goods. That’s particularly true if you judge it by how many hours of labour it costs to buy, say, a car or a restaurant meal.
What we’re saying is that, in rich, high-productivity economies such as ours, labour is the more expensive resource, and capital the less expensive resource.
It’s also true that there’s a limit to how much you can eat, how many cars you can drive and how many TV sets you can watch, but no yet-discovered limit to how many services you can pay other people to perform for you.
Put all that together and the goods sector’s share of the economy keeps getting smaller, while the services sector’s grows – to 80 per cent of the economy (gross domestic product) and 90 per cent of total employment.
But it also means that the sector which has little ability to improve the productivity of its labour also has to keep paying more for its labour as the goods sector increases the productivity of its labour.
Gosh, that’s not nice. No, which is why Baumol said that the services sector suffers from “cost disease”. And the services sector’s huge and growing share of the economy explains why productivity in the economy overall is improving more slowly than it used to.
But it could become even worse. If, as it seems, the goods sector has finally exploited almost all its potential to become more productive, and there’s not a lot of obvious scope to improve the services sector’s productiveness, it’s hard to see how we’ll get much more productivity-driven growth in the economy.
What a dismal prospect. Talk about the problems of affluence. You know, I don’t think the world’s poor have any idea how hard life will become for us.