Technological Change and the Future of Work

Technological Change and the Future of Work

Domestic Policy
Thursday, 07 December 2017

Dr David Gruen, Deputy Secretary, Economic and G20 Sherpa


Thank you for inviting me today.

Just last week, the Presidency of the G20 passed from Germany to Argentina. Argentina has made clear that the future of work will be a priority for its Presidency.

So the topic of the future of work is a topical one. I suspect I will be spending many hours over the course of the Argentinian Presidency discussing this topic with other G20 Sherpas, including my friend, Indonesian Sherpa, Pak Rizal.

Many young Australians visit Indonesia every year. In my generation, upon arrival, with no phone and travelling only with a ‘Lonely Planet’ guide, they would negotiate a good rate with an ojek driver and head to the closest warung.

Those young people never envisaged a world where tourists could use smartphones to request a ride, order food or have their hair styled.

But, of course, now Indonesia has exactly that with GoJek.

As the physicist Niels Bohr is reported to have said ‘Prediction is very difficult, especially about the future’.

Nobody can predict exactly what technology will mean for the future of work.

But in many ways our time is a most exciting one.

Today I want to talk about what opportunities our time may bring for the millions of working men and women in our region ― including, of course, those in both of our countries, Indonesia and Australia.

Displacement of jobs by technology is an old fear

Many people around the globe approach technological change with caution, even dread, rather than excitement.

They fear new technologies may take their jobs or their children’s.

This is an old fear. Indeed, the common term used for those who fear technological change harks back to events that occurred two centuries ago. Early in the 19th century, English textile workers who destroyed new weaving machines they blamed for taking their jobs were called ‘Luddites’, after their supposed leader, Ned Ludd. Here we see Ned Ludd and followers destroying the machines that they feared were driving their families into poverty.

There is no doubt that, over the past two centuries, waves of technological change have eliminated jobs, and rendered some occupations obsolete. But they have also facilitated the creation of new jobs to take their place ― either directly, or indirectly as a result of rising standards of living generating new demands. 

Technological change leads to change in demand for labour between sectors

Changing technology and consumer demand have always changed the nature of jobs.

The charts below tell the story for Australia and Indonesia over the past 40 years.

In Australia, 40 per cent of the workforce was employed in agriculture or industry in 1976, compared to less than a quarter today.

The Indonesian story shows some clear similarities. In 1976, two thirds of the workforce was employed in agriculture and this share has fallen to around a third today, replaced by growing industrial and services employment.

Source: International Labor Organization, ILOSTAT.

Why are we so focused on technological change now?

Some people think the pace of technological change has been accelerating, or is about to accelerate. Or alternatively, that the coming wave of technological change (associated with artificial intelligence and machine learning) will be particularly disruptive, making it harder for workers to adjust. But there are also prominent analysts, such as Robert Gordon, who argue that the rate of technological change has slowed.

Yet another claim is that the type of technological change may now be affecting more influential groups, such as the middle class, or whole geographic regions, such as manufacturing in the mid-west of the United States.

One thing is clear. Technological disruption is much harder to manage in the absence of a stable macroeconomy. Some of the fear in the developed world about technological change may reflect the lasting scars on the labour market from the global financial crisis in many advanced economies.

The chart below shows recent employment growth in Indonesia, Australia and the United States. It serves as a reminder of just how severe the employment shock in the US was as a result of the global financial crisis ― in stark contrast to the outcome for our two countries.

Source: OECD Short-term Labour Market Statistics; Indonesia Badan Pusat Statistik

Job polarisation in developed nations and distributional consequences of automation

It is important to recognise that even if macroeconomic policy is well managed and employment holds up, technological change can have an impact on the distribution of income.

Many developed countries, including Australia, have seen ‘job polarisation’ ― with stronger growth in both high and low skilled jobs, at the expense of mid-skill jobs.

The chart below tells the story for the US. During the 2008-09 recession, employment losses occurred throughout the economy, but were concentrated in mid-wage occupations. By contrast, during the recovery in the US labour market, employment gains have been concentrated in lower-wage occupations.

One of the implications of this job polarisation is arguably the rise in disaffected populations and populism in many advanced economies, with people seeking representation from those who sympathise with their lived realities ― which are quite distinct from the realities of the urban elites.

Source:  NELP analysis of the Current Population Survey (all estimates measured at Quarter 1).

But emerging economies have been winners in this new world

Taking a longer-term view, it is worth reflecting on how extraordinary the recent era has been in terms of income growth for a significant majority of the people living on the planet.

Larry Summers recently noted that, over the 2300 years between the time of early Athens and London in the early 1800s, standards of living rose by an estimated total of about 75 per cent. During the Industrial Revolution, growth accelerated so that (real) incomes grew by up to 50 per cent over a human lifetime. Yet in the recent era, 50 per cent income growth has occurred over periods as short as six years at various times in China and other countries in our region.

In the recent era, while incomes within many advanced countries have become more unequally distributed, incomes across the globe have become more equal.

You can see this in Branko Milanovic’s famous chart. Over the 23 years to 2011, emerging economies, including Indonesia and other countries in Asia have seen average real incomes more than double. By contrast, the average real incomes of ‘Mature Economies’ have grown only sluggishly.

*1988 positions are PM&C approximations using data from Lakner and Milanovic (2015).
Source: Milanovic B. (2016), Global Inequality: A New Approach for the Age of Globalization, Belknap Press, p31.

The Indonesian and Australian stories

It is also of interest to look specifically at real income growth in Australia and Indonesia over the same 23 year time period.

Australia has performed much better than the average of other mature economies shown in the previous chart, with strong income growth right across the income distribution.

Indonesia has also out-performed other countries in ‘its cohort’, the Asia (ex-India, ex-China) group. It has seen roughly a doubling in real incomes for poorer income deciles, and significantly stronger income growth for richer income deciles. It has also seen stronger income growth than Australia across all income deciles ― consistent with the significant opportunity for strong catch-up growth for emerging economies that are open to new ideas, new technology, investment and trade.

Source: Global Consumption and Income Project.

Technology and jobs

Turning now to the link between technology and jobs, history is littered with the dire warnings of prominent analysts about the widespread job losses that would result from new technology. To give but one example, the great economist and Nobel Laureate, Wassily Leontief, wrote in 1953 that ‘labour will become less and less important … More and more workers will be replaced by machines’.

The experience of the subsequent 64 years suggests, at the very least, that that prediction was premature.

An insightful way to understand the overall effect of new technology on the number of jobs in the economy is as a race between two dynamic processes. On the one hand, automation tends to take jobs away while, on the other, the invention of new complex tasks creates new jobs.

In the history of technological advance, at least in the two centuries since Ned Ludd, the second process has broadly kept pace with the first. That is, there have been sufficient useful new complex tasks invented for people to do that there has been no secular rise in technological unemployment over that time. And importantly, with rising levels of education, people have risen to the challenge of mastering these new complex tasks.

The historical experience has been that technological advance has not adversely affected the overall level of employment, but rather has raised the overall level of income. Technology makes us richer.

The key policy issue for the impact of technology on employment is facilitating the transition of workers, either to adopting new technologies in their current jobs, or into new jobs that become available.

Workers need the skills to take up new technology or to find new jobs. Because people will have more money, some of these will be in traditional industries that technology finds hard to do ― such as hairdressing. Many others will be in areas that only exist because of new technology.

This of course is the political irony with technology ― it threatens jobs we can see, but creates jobs that we can’t yet imagine. Last decade few workers would have anticipated jobs today as ‘big data’ scientists, cybersecurity experts or online-reputation managers for pop stars like Afgan and Rossa.

While the GoJek app has undoubtedly displaced some taxi drivers, other jobs have been created in service industries dependent on accurate and low cost transport.

Technology sometimes eliminates whole jobs. But usually it replaces the most boring, dangerous or repetitive tasks associated with a job. To give but one example, the spreadsheet has removed a lot of the ledger balancing done by accountants, leaving them free to do more of the creative tasks in corporate accounting.

Indonesia is remarkably well connected to the internet, particularly in this city. I am told by a colleague now living in Jakarta that he can access better internet speeds here than he could at home in Canberra.  Almost everyone has a smart phone, and the penetrations of Twitter and Instagram are some of the highest in the world. I mention this because these platforms represent some of the leading edges of technology, with many Indonesian businesses advertising their products and services on these sites. 

Some jobs are more vulnerable than others

While I am confident there will be jobs, I am less confident about what jobs they will be.

Some jobs are clearly more vulnerable to automation than others. The chart below shows a representation of which jobs are more or less exposed to automation.

Machines are good at routine, repetitive, precise and predictable tasks. They are less good at non-routine jobs ― those involving abstract or creative effort, involving judgement and/or inter-personal skills. Interestingly, these attributes that robots find difficult are the qualities that most clearly distinguish humans from other species.

The future of work for us is doing tasks that are ‘more human’.

Take health care ― a nurse I know of tells the story that she and her colleagues have always had to adapt to the latest technology to help with patients. Doctors, though, are increasingly turning back towards dealing with patients, relying more on technology to do clinical work. Perhaps the most at risk are specialists who rarely if ever deal with patients, such as anesthetists.

The gender story in all of this is interesting. The International Labor Organization suggests that, in Indonesia, women are about 50 per cent more likely than men to occupy a job at high risk of being automated.[1]

But this is only part of the story. Women also may be in a strong position to take up the economic opportunities in the growth industries of the future.

Source: PM&C infographic. 

The decline of routine jobs

The next chart presents an estimate of the proportion of jobs by type in Australia since the mid-1980s. Routine jobs, both manual and cognitive, have been in decline, while non-routine jobs have become increasingly prevalent.

*Non-seasonally adjusted.
Source: Reserve Bank of Australia, Alexandra Heath (2016).

The path to economic development is less clear

Technology may also be having an impact on countries’ optimal development path.

Traditionally, as they have developed, countries have tended to transition from agriculture to simpler manufactured industrials, before moving to higher technology manufacturing and service industries. This pattern was followed in Europe in the 17th and 18th century, and subsequently in the US. Japan was the first Asian economy to industrialise in the late 19th century, South Korea and Taiwan in the 1960s, China from the mid-1970s and Indonesia and India more recently.

But it may be that this development path is now less obvious than it once was.

At any given level of income, countries are now less reliant on manufacturing (or industry) than they were in the past, as a proportion of employment and output. Rodrik (2015) suggests that this trend towards earlier de-industrialisation is due both to technology making manufacturing more productive, and to globalisation allowing countries to specialise.

This de-industrialisation trend applies both to developed and developing countries ― nearly everywhere but Asia. In Asia, notwithstanding strong productivity growth, manufacturing has remained robust because of Asia’s comparative advantages, including low costs of transportation.

The chart here highlights the peak of the industry sector in several Asian economies. As it shows, developing economies are reaching this peak at progressively lower shares of employment. Japan peaked in the 1970s and Singapore in the 1980s at around 35 per cent. China peaked at around 30 per cent earlier this decade. Indonesia’s industry share is still growing but, if this history is a guide, it may also soon peak. 

* ‘Industry’ consists of mining, manufacturing, construction and utilities.
Source: World Bank, World Development Indicators.

The emerging future of work in our region

Thus, technology may be changing the nature of the development process. Technology has made it easier to ship not just goods, but services and ideas across borders. Measured in value-added terms, the share of goods trade fell from 71 per cent of world exports in 1980 to 57 per cent in 2008. Much of the value of an iPhone, for example, derives from the original design and embedded intellectual property of the product rather than from its physical components and assembly.

So long as workers in emerging countries have the requisite skills, we may see more of them staying at home and providing services internationally ― as we already see in India’s IT industry.

Technology is also breaking down barriers between traded and non-traded services.

Recent technologies may be making geographic location less important. Rather than factories situated together, exported services can now be delivered from people’s homes almost anywhere. Technology also offers the potential to allow people to work around their own family and cultural values, while also being more inclusive for people with disabilities.

I will end by offering some reflections on what policies are most likely to best support the future of work.

First, staying open to the world ― to ideas, trade and investment.

No country by itself can come close to developing and harnessing the ideas that the over 7 billion people in the rest of the world can collectively offer.[2]

Second, and the corollary to openness, is to have in place policies that help communities adapt to technological change, especially when jobs have been lost and alternative employment possibilities require people to move.

Third and finally, domestic policies that promote a more productive economy. Picking winners for industries or geographic areas is probably harder to do than it was in the past. But improving infrastructure and education are critical ingredients to give workforces the best chance for whatever the future will bring.

Thank you.

[1] Chang and Huynh, ASEAN in Transformation: The Future of Jobs at Risk of Automation, International Labour Organization (2016) and Bureau for Employer’s Activities, Working Paper No. 9, page 20.

[2] UN World Population Prospects: The 2017 Revision (current world population of 7.6 billion).