Japan and the Rise of the Machines

Japan's investment in robotics is starting to prosper, as they become leaders in the innovation space.


This article first appeared on ANZ’s BlueNotes. The original article can be found here.

Japan faces some deep-seated structural issues, such as population decline. The demographic trajectory looks terminal.

But much is far from obvious in Japan. The negative story often obscures the real story.

It is not just that these factors are overplayed – for example, the annual labour force decline on a percentage basis in Japan is in fact lower than Singapore, China, Korea and many other countries – but it also ignores the opportunities sometimes created by the way the country responds to these challenges.

“Much is far from obvious in Japan. The negative story often obscures the real story.”

When we look at Japan in 2017, we can see one of the tightest labour markets in the world. Just 2.8% unemployment, compared with over 9% in Europe and just under 5% in the US. The jobs-to-applicants ratio is at its highest level since the early 1970s.

The under-supply in certain parts of the Japan labour market is now at levels which should induce two main market reactions:

The most ‘natural’ reaction of course is wage growth –something Japan has seen suppressed for the past two decades. However there are some signs wages are finally starting to simmer, beginning with pockets where there are no immediate solutions – such as the recent rise in wages for delivery workers for example.

The other reaction is already well in train, the move towards robotics and automation, replacing many roles traditionally undertaken by people. A movement which began in manufacturing – and accounts for Japan’s G7-leading levels of factory productivity – has now moved to the service sector.

Importantly, it is not just market forces driving change. The government has become a strong proponent and polices relating to technology substitution of labour – for example Prime Minister Shinzo Abe’s proclamation Japan will quadruple the robotics sector in a “robot revolution” to overcome the declining labour force.

However, these technology goals do create tension with the pursuit of rising wages – witness for example the reaction of leading delivery company Yamato Transport to its recently increased wage costs.

The company is piloting autonomous delivery vans which will park in neighbourhoods for people to retrieve their deliveries.

Opposing forces

It is unclear how these competing forces will play out. In the short term it appears likely there will be some demand-pull inflation on wages, as we have already seen in the non-permanent worker group over the last two years, where wages have risen between 2 and 3% per annum.

To the extent this can feed through to the permanent work force, it should at least stabilise inflation and potentially also give the government an opportunity to pursue the long-delayed increase in consumption tax.

With corporate earnings at historic levels – in 2016, the bottom line of the non-financial sector grew on average around 21% – one could argue companies can certainly afford the increased wages.

It’s not simple however. There are two issues in Japan which have impacted the ability for labour market tightness to translate into overall wage growth and hence consumption.

The first is the two-tiered structure of the labour market. In Japan, there is a stark contrast between the permanent workforce and the temporary or ‘non-regular’ workforce.

The concept of equal work for equal pay does not exist. In fact, a temporary worker earns around 53% less than a permanent worker on a comparable monthly basis.

There are many reasons for this – the ingrained view permanent work is more valuable and less aggressive union representation are just some. This disparity has, perhaps not surprisingly, resulted in non-regular workers comprising an increasing percentage of the workforce over the last 20 years.

The good news is this is now changing, as recent labour data shows. The demand-supply imbalance is now so high with non-regular workers it is having the twin impacts of wages rising for those workers as well as employers turning back towards more permanent staff.

The other reason employers have continued to prefer non-regular workers, at least in the service sector, has been the fact labour productivity has continued to decline.

Japan is certainly not unique in that regard – the decline is pervasive across the developed world. However in Japan it stands in stark contrast to the automated factory sector which has made employers reluctant to raise wages.

Ready for the Revolution

The real conundrum then remains the intersection of the strategic push for automation versus the immediate and clear tightness in the labour market.

No-one knows exactly how this will play out. But we can say for certain Japan will advance more quickly than most towards the automated society.

What is perhaps an interesting discussion point – and I’ll freely admit this is nothing more than a personal thesis – is Japan is almost uniquely well-placed to deal with the transition from human labour to automated machine labour.

Consider in Tokyo for example, there are currently more than two job vacancies for every job-seeker. If one of those two vacancies happened to become automated, then it is not replacing available human labour.

It is new supply filling a demand gap. In other countries this gap may normally be filled by immigration. But in Japan, substitution by machine is a culturally and politically more palatable solution.

Contrast with say Europe, where automation would theoretically be replacing under-employed youth. Or, worse still, consider the situation in developing markets such as Bangladesh or Cambodia.

In such countries, almost all labour is employed in unskilled or semi-skilled activity – such as clothing or agriculture – which is liable to automation. Assuming automation increases faster than a country can upskill, then you create a huge cadre of jobless, with little prospect of change.

These technological gains could therefore lead to social dislocation and potentially be a drag on growth despite the gains in “productivity”.

Japan, due to its demographic profile and the increasing excess demand for labour, will not face this issue – at least not in the same way. Arguably it may actually lead to greater social cohesion rather than less.

Fault Lines

The flaw in this thesis – or more accurately the inherent conflict – is the latent availability of women workers.

Whilst female participation in the Japan labour force overall is not as low as one might think at just under 70%, there is a huge gulf of untapped talent due to ‘participation’ statistics disguising the fact many of these women are working only a small number of hours – and generally not by choice.

This can be driven by the huge backlog in child day-care facilities, or by the structural disincentives posed by Japan’s spousal tax deduction, amongst other things.

Ideally, Japan will resolve as much of the labour shortage as possible through the increased participation of women in the labour force.

However because of the deeply embedded opposition to immigration, it is highly likely whatever demand gap remains will be filled by the very urgent pursuit of technology-based solutions.

Given the demographic context, Japan is likely to pursue this automation agenda with more vigour than other nations.

Exportable Innovation

This agenda may create exportable solutions for other countries – though of course as mentioned, the social outcomes of those solutions may be quite different to Japan. However some countries may offer some real potential for productivity gains in key sectors.

The main message perhaps is the path for Japan and its labour market will be anything but linear. As always, there will be contradictions, and competing forces. But these forces are also likely to create opportunities both for Japan and for other countries.

A declining labour force does not necessarily lead to an Armageddon outcome – as they say in Japan, “if you believe everything you read, better not read”!


I witnessed a good example of the cross-border potential of Japanese automation recently in relation to the agricultural sector in Japan, which has potential application to ANZ’s home markets of Australia and New Zealand.

A number of household names in Japanese technology – names such as Hitachi and Fujitsu – have been working on both software and hardware tools for automating activity in agriculture.

A key driver of this is the fact the agricultural sector in Japan is rapidly aging, with the average farmer aged nearly 75 years. Labour is very hard to find in rural Japan.

Hence automation is a necessary way out to ensure the sustainability of the sector (which despite being a tiny part of Japanese GDP, still has enormous potential given the available arable land mass).

Amongst other things, these Japanese companies are working on a suite of solutions which allow a farm to be run dashboard-style from highly sophisticated applications combining geo-location with sensor technology, drones and driver-less machinery. The potential application to Australia and NZ agriculture – both of which are target markets – is clear.

Both markets share an issue of high labour costs combined with difficulties in sourcing labour and in Australia in particular the massive physical scale of some farms means they traditionally require a high labour input.

The potential is there for Japanese technology – originally built to solve Japanese problems – to have a positive impact on agricultural productivity in in Australia and NZ.