We've long been hearing about the threat of "robot redundancy" – and just whose jobs are at risk of automation.

Fewer than 5 per cent of jobs can be done entirely through automation today - and researchers have thrown out wildly varying forecasts how this might soon change.

One headline-grabbing study from Oxford University, suggesting that 47 per cent of all employment in the US was at risk of being replaced by computers and algorithms in the next 20 years, appeared overblown when set against findings of a forensic OECD investigation.

Of the 21 advanced nations that report looked at, only 9 per cent of jobs were "potentially automatable".


And, going by surveys here at home, Kiwis don't appear fazed.

A recent Massey University study found 87.5 per cent of respondents either disagreed or strongly disagreed with the statement "smart technology, artificial intelligence, robotics or algorithms could take my job".

But that said, perhaps Kiwis should be concerned.

Parts of our jobs indeed could be automated - especially those highly repetitive and uncomplicated tasks, like data entry.

"When parts of the job are automated, you may see a net loss of employees needed overall, or, that nature of the job simply changes to include more social and creative aspects," Massey's Dr David Brougham recently told the Herald.

"Employees should constantly assess technology that could do parts of their job, and consider what this will mean for their profession. It would be wise to plan for a possible shift in the labour market as a result of automation and other technological disruptions."

Further, he warned, technology and market pressure could continue to push away traditional nine-to-five work, as the casualisation of labour may continue to make work more temporary and unstable.

Now, an issues paper just put out by the Productivity Commission has explored the future of work in four potential scenarios – ranging from more tech and fewer jobs, to much the same as what we've seen over the past two decades.


The Government had tasked the commission to produce a wider report into the possible impacts of new technology on the labour market by March next year.

Its initial paper, released last week, found tech-driven change could have a "hugely positive" impact on Kiwis' wellbeing – yet the benefits might not necessarily be shared evenly throughout society.

For instance, many Kiwis, particularly in low to medium skill occupations, involuntarily had lost their jobs during the past few decades in part because of advances in technology.

"Workers who get made redundant may struggle to cope mentally and financially and face diminished future job prospects and earnings," said the report.

"But technological change can also affect workers and work in other ways.

"Examples include people who take time out of the workforce but struggle to re-gain employment as the types of skills in demand have changed, or those who invest in learning the skills for a profession but find it difficult to gain a job because the demand for that profession has diminished."

More tech and more jobs

It's true that technological change has already created many new jobs for Kiwis.

For example, since 1999, the number of jobs classified as "computer systems design and related services" increased from 8700 to 32,600.

And because of the new tasks created by tech, the net effect of new technologies could be to increase the total number of jobs.

One third of new jobs created in the US in the past 25 years, the paper noted, were types that did not exist - or barely existed - in areas including IT development, hardware manufacturing, app creation, and IT systems management.

At the same time, this change could free people from onerous or physically-demanding work - especially in agriculture.

Across the developed world, technological improvements had led to dramatic falls in the share of the workforce employed in the primary sector, while production grew.

This share of the workforce in New Zealand fell from 36 per cent in 1901 to 7 per cent in 2013.

Over time, workers simply moved to new and expanding opportunities in other industries.

Tech could also slash the cost of producing an existing good or service, which in turn may raise demand and grow jobs, the report added.

Where competition was strong, adopting a cost-reducing technology could allow a firm to increase its market share by lowering its prices.

Competitors then faced pressure to lower their prices to keep pace – and lower prices typically led consumers to purchase more of that good or service, spurring growth in employment to meet the increased demand.

This was once described as the "Uber effect", after ridesharing services such as Uber and Lyft provided a cheaper alternative to taxis.

What's called the
What's called the "Uber effect" means technological change can create many new jobs, rather than just replace current ones. Photo / File

While the number of traditional taxi trips in New York city fell between 2015 and 2018, this reduction was more than offset by a dramatic increase in trips using ridesharing apps.

The first and most optimistic scenario painted by the paper would see adoption of tech keep accelerating, and, like Uber, create more jobs than tech replaced.

Both capital and labour productivity climbed - perhaps substantially – while the quality-adjusted price of technology fell, partly because firms only adopted what they expected would give them a boost.

Although such an economy would lean toward high levels of employment, that didn't mean there wouldn't be concerning disparities in wage incomes.

Low-skilled labour might attract low or stagnant wages while the wages of those with skills in demand might soar.

"On the plus side, in such an economy there are more resources, in total, to enable redistribution of the benefits," the report noted.

"Higher levels of job churn are likely in such a labour market."

That might lead to increased demand for mid-career retraining – and jobs likely to be lost were those that were routine, as Brougham warned, and thus more amenable to automation.

Job gains would be those that are more complex, deal with unusual or unpredictable situations, or require essentially human elements such as judgement and compassion.

There would likely be greater demand for education and training, including from displaced workers and those in employment, and a greater need for flexible delivery options for training.

More frequent and widespread displacement might create financial pressure for affected households, increasing calls for wider income support measures - including for those undertaking retraining.

The report inevitably touched on the idea of a universal basic income – or UBI – which was designed to provide unconditional payments made to people without means tests or obligations to seek work.

This concept had some basic benefits – one them notably being that it eliminated poverty traps – but also had some obvious drawbacks.

A comprehensive UBI would have huge fiscal implications - a UBI of $200 a week for each Kiwi aged 18 to 65 would cost roughly $30 billion per year – and its impacts on employment weren't clear.

More tech and fewer jobs

This was the potential flipside to tech-driven prosperity.

It was well known that some types of roles could vanish when tech improvements dramatically improved the capability of capital equipment - prompting firms to invest more in equipment rather than jobs.

In the early 1960s, for example, nearly 27,000 people were employed as "waterside workers and related freight handlers" in New Zealand.

The widespread adoption of containerisation and other technology that has enabled the automation of ports has seen the number employed in "water transport support services" reduce to fewer than 6000, while the volume of freight handled has increased dramatically.

This graph shows the share of jobs at risk of automation or
This graph shows the share of jobs at risk of automation or "significant change" in OECD countries, as at 2016. Source / Nedelkoska & Quintini / Productivity Commission

In some cases, investments in capital could completely replace some jobs.

One well documented example was the near-disappearance of typists as an occupation following the widespread adoption of personal computers.

Technologically driven improvements in the quality of capital equipment could further lead a firm to invest more and also hire more staff with the skills required to operate the new assets.

These types of improvements are known as "skill-biased" technological changes – which had been cited in some countries as an explanation for the growing divergence between wages for higher-skilled workers.

In a scenario where technology ultimately turned out to be labour-replacing, capital productivity rose substantially, as firms increasingly adopted productivity enhancing technologies.

Labour productivity could also rise, as lower-skilled roles are increasingly automated.

An expected consequence of this combination of drivers was widespread unemployment.

While workers losing jobs would contribute to job churn, overall job scarcity might encourage those with jobs to hold onto them for longer.

Average wages could fall over time as labour supply would exceed available jobs – and even those who hung on to their gigs could face wage inequality.

"However, income differences between the employed and unemployed could drive an increase in overall rates of inequality," the report noted.

"With higher rates of unemployment, a much larger share of households will face financial pressures, and there may be calls for more generous income support."

There may also be calls for direct taxation of technology - sometimes called for robot taxes - to fund income support programmes and other interventions.

Elsewhere, people might demand greater sharing of existing jobs and increased employment protection for those still in work.


Another possibility was a slowing of tech adoption, perhaps due to falling innovation as technological bottlenecks became harder than expected to overcome.

Alternatively, slower change could occur as technology adoption by firms slowed – perhaps because newer technologies are less productivity enhancing for firms than those of the past.

In this scenario, the report found, there would be less change in the volume, churn and nature of work, and income and productivity growth will slow.

There could be reduced opportunities for people to find jobs that are well matched with their skills, preferences and circumstances.

Slower technological change could prompt calls for more government intervention to encourage firms to undertake research and development and innovate.

Steady as she goes

It's worth noting that, for all the fear about tech-driven disruption, New Zealand's relatively low unemployment rate, frequent reports of labour shortages and growing labour market participation rates suggested that the overall availability of work wasn't a serious issue today.

Source / Productivity Commission
Source / Productivity Commission

That wasn't to say the status quo would carry on – but a fourth scenario nonetheless assumed this possibility.

In this case, tech drivers of labour market change over the next one to two decades would stay within the bounds over what had been seen since the turn of the century.

While change would happen here, it would be at the rate at which we were familiar with: the advent of smartphones, eCommerce and social media.

This scenario assumed the continuation of slow productivity growth and generally slow technology adoption by New Zealand firms – and was compatible with full employment and stable levels of income inequality.

"Actual outcomes will vary," the report noted, "as other drivers of labour market change – such as variation in the business cycle or demographic change – are likely to predominate".

Submissions on the discussion paper close on June 5 and the report is due out on March 31 next year.