Why the Australian government’s latest report on automation ignored the rise of self-driving cars

The government has made a glaring omission in the latest research on the future of Australian industry.

Can you spot what’s wrong with this graph?

It’s perhaps the biggest and most notable tech trend of 2014, yet it’s missing from the government’s latest research on the future of Australian industry.

According to the government’s latest report, truckies, bus drivers and taxi drivers are among the least likely to have their job replaced by automation within the near future. But anyone tied to the technology sector would know that all three of these jobs are in the firing line for disruption by self-driving cars.

The wheels of disruption are already in motion. Regulation is changing ahead of the self-driving car revolution. For example, the UK passed laws earlier this year to allow self-driving cars on the road, well before their anticipated mainstream release. The state of California followed suit just a couple of months ago.

Closer to home, Business Spectator has already documented how BHP Billiton is using automated trucks to cut costs and drive productivity.

Mainstream adoption of the technology could occur sooner than we think. According to Accenture’s Australia and New Zealand managing director David Maunsell, so far analyst reports and estimates have consistently overshot the pace of the trend. He believes the forecast made by former General Motors corporate vice-president Larry Burns -- that self-driving cars will be mainstream by 2018 -- is too conservative.

This all begs the question: why wasn’t this mentioned or accounted for in the report? The answer is rational, albeit convoluted. 

According to Mark Cully, the report’s main author and chief economist at the Department of Industry, the report attempted to replicate and localise a prior MIT study into job automation. It suggested that role automation depended on the nature of the role: routine tasks could be automated, whereas manual and abstract tasks could not.

Here’s the problem: for control, this study negated any influence of current or future technology. It simply considered the nature of the roles in isolation.

It basically said that routine jobs, like the ones below, could be automated. 

The report also suggested that roles which include manual labour or abstract work would be more difficult to replace with machines. But this isn’t exactly the case. In fact, another study from Oxford University considers taxi and truck drivers roles to be significantly more endangered by automation.

Regardless of the report, Cully believes automation will, in fact, disrupt the transport sector.

“As well as trucking, it [automation] may also affect mining, insurance, and public and private transport, including taxi services,” he told The Ticker.

“Equally, as the report makes clear, new technologies create all sorts of new job opportunities, as well as the benefits they offer consumers in lower prices and higher incomes,” Cully added.

This situation is an important lesson in interpreting research on automation and even wider technological disruption. Regardless of the figures or the credibility of the research, sometimes the most logical answer is indeed the most accurate one.