So why is it happening? That’s anybody’s guess. A host of possible explanations is being advanced and debated. It could be another decade before a new conventional wisdom emerges.

I’ve written before about the thesis that the digital revolution won’t boost productivity the way earlier waves of general-purpose technologies did, about the thesis of “secular stagnation” and yet another idea that the main trouble is decades of weak business investment.

But last week Dr Luci Ellis, a Reserve Bank assistant governor, offered her own thoughts on yet another possible piece in the jigsaw puzzle. Productivity is generated by firms, but Ellis notes that, both in Australia and abroad, the evidence suggests that levels of productivity vary widely between firms, even within the same narrowly defined industry.

“Firms that are highly productive – so-called superstar firms – tend to grow faster, grow employment faster, and pay better than firms that are a long way from the frontier of productivity”, she says.

But there’s a problem. Because these superstar firms are more productive than average, they gain market share at the expense of less-productive competitors.

The leading firms could start moving further and further ahead of the pack.

Those that lag behind would then find it harder and harder to catch up. The result could be that markets become more concentrated.

The market leader begins to reap monopoly profits, which isn’t good for consumers and might not be good for long-run innovation and [society’s] welfare.

Luci Ellis

“The market leader begins to reap monopoly profits, which isn’t good for consumers and might not be good for long-run innovation and [society’s] welfare”, she says.

But must the laggard firms never catch up? That may depend on why so many firms are lagging. If it’s because they lack managerial ability, it ought to be possible for them to copy the leaders’ superior approach or even poach their rival’s managers. If so, this would lift the whole industry’s – and the nation’s – productivity.

Loading

But what if the laggards have lower productivity because they aren’t adopting the latest technology the way the superstars are? There’s evidence this is the case in other advanced economies, but Ellis says we don’t yet know if it’s true in Australia.

If this superstar pattern has arisen only recently, it could be something to do with the nature of developments in digital technology and their ease of adoption.

Previous waves of general-purpose technologies, such as electricity or the earlier round of computerisation, had the benefit of reducing the level of skill needed to operate them, whereas innovations such as machine learning and artificial intelligence seem to have a very different character, she says.

“Using machine learning and other emerging techniques to automate routine business processes seems to involve specialist skills and, often, PhD-level training in statistics or computer science. These skills are much rarer and take longer to develop than those required for the jobs that are thereby replaced.

“That doesn’t mean it’s impossible, but it could take a long time,” she says.

And get this: if leading-edge technologies are (at present, anyway) unusually costly or difficult to adopt, they become a kind of barrier to entry protecting the firms that are already using those technologies.

That would be a worry if lagging firms never caught up. And if incumbents never face rivals, they’re more likely to become complacent. “Innovation could slow down, and growth in living standards with it”, she concludes.

So, is this the big reason productivity improvement has slowed throughout the advanced economies? Far too soon to say.

But it makes an important point: the problem, and the solution, lie in the hands of our big companies.

Governments may have a role in spending more – and more wisely – on education and training, but giving up a lot of revenue to cut the rate of company tax isn’t likely to make much difference.

Ross Gittins is the Herald’s economics editor