Cotton is processed at the Janki Oil Industries ginning facility in Kadi, India, on Monday, Jan. 28, 2008. Debris and seeds are first removed from the cotton, which is then washed and packaged into 250 kilogram (550 pound) bales. Cottonseed oil is extracted from the seeds, and the organic byproduct of the process is used as cattle feed. Photographer: Keyur Khamar/Bloomberg News
© Bloomberg

Fresh analysis from the OECD has heightened fears that developing countries will be worst hit by an anticipated wave of “creative destruction”, as jobs are lost to the march of automation.

At least one strategist believes this is already happening, as technological progress swings away from favouring globalisation to incentivising the “localisation” of production back to developed economies.

“Technological progress, once a facilitator of globalised supply chains, now challenges emerging market manufacturers’ climb up the value-added ladder and productivity growth,” said Marvin Barth, head of FX and EM macro strategy at Barclays Investment Bank.

“It is apparent in the data for investment, foreign direct investment and relative developed market expenditure on manufactured goods already,” added Mr Barth, who feared this may lead to “a degradation of EM institutions that has the potential to create a vicious downward spiral in some EM economies”.

The OECD’s latest analysis pointed to the proportion of jobs it deemed to be at high or significant risk of automation. Although most of its members are developed countries, it also included a number of countries deemed to be “emerging” by various organisations, and they are predominantly the ones in the line of fire.

Of the 28 OECD countries analysed, the 11 adjudged to have the smallest proportion of jobs at “significant” risk are all developed. In contrast, all 11 emerging countries (at least by the definition of MSCI, the index provider) are among the 17 deemed most vulnerable to job losses, as the first chart shows.

A graphic with no description

In term of jobs deemed as being at “high” risk, six of the seven most exposed countries are emerging.

Overall, Slovakia comes out worst in this analysis, with 64.4 per cent of jobs seen as being at high (greater than 70 per cent) or significant (between 50 and 70 per cent) risk of being lost.

Several other central and eastern European countries, such as Lithuania, Turkey and Greece, are not far behind, while the likes of Scandinavia, New Zealand, the US and the UK are seen as being better placed to withstand an anticipated tide of automation.

Separate analysis by Standard Chartered bank added to the concerns. It suggested that the wave of technological disruption that has already hit developed countries, leading to a “hollowing out” of middle class, middle skilled jobs, has yet to reach the developing world, as shown in the second chart (although the data, despite being sourced from the IMF’s World Economic Outlook of April 2017 are somewhat dated).

Madhur Jha, head of thematic research at StanChart, argued this was because emerging countries have not yet adopted digital technologies to the same extent as in the developed world, where advances such as robotics have allowed “routine, repetitive jobs, both cognitive and manual, to be automated”.

A graphic with no description

This in turn could be because “many jobs in EMs are still low-skilled ones that are more manual, low-skill and non-routine jobs that do not lend themselves to automation easily” with, for instance relatively little employment in information technology-related occupations.

However, Ms Jha said there were two reasons why the “insulation” that developing countries have enjoyed thus far “is set to decline”.

Firstly, more labour is likely to move into middle-skilled jobs in the coming years as economies develop, while more employers are likely to adopt digital technology.

Secondly, many emerging market countries, such as India, Pakistan, Bangladesh and African states such as Ghana, Kenya and Nigeria have a high percentage of their labour force engaged in agricultural and manufacturing jobs, as the third chart shows.

While many of such jobs will fall into the category of low-skill, non-routine manual labour, Ms Jha said, “as economies progress and move up the value chain, many of these jobs in agriculture and manufacturing will ultimately lend themselves to routinisation and automation”.

A graphic with no description

This ties in with a warning from the World Bank in 2016 that two-thirds of jobs in emerging markets are “susceptible to automation”, but this process will be dependent on the pace of technological disruption.

“The share of occupations that could experience significant automation is actually higher in developing countries than in more advanced ones, where many of these jobs have already disappeared, but it will likely take longer in lower-income countries,” the bank said in its world development report, citing lower levels of IT take-up and lower wages, which means investments in technology may be less profitable than in the west.

Consequently, the World Bank saw workers in middle-income countries, such as Argentina, Croatia, Uruguay, Latvia and China, as being more exposed to automation than their counterparts in poorer countries, as the fourth chart indicates.

A graphic with no description

The analysis by Barclays’ Mr Barth added a separate concern to the mix. Whereas the likes of the World Bank have focused on economic activity continuing to taking place in the emerging world, but with fewer employees per unit of output, Mr Barth went further, arguing that technological progress will also mean some activity physically moving from EMs to developed markets.

His premise is that technology has been a major driving force of globalisation by enabling management of complex global supply chains and associated logistics.

However, he believed technological progress, in the form of robotics and 3D printing, now favoured “localisation” instead.

“With automation costs for most manufacturing processes falling below that of even the cheapest labour economies, there no longer is an incentive for DM firms to outsource production away from their main demand centres or home markets,” Mr Barth said.

Instead, cheaper automation allows for more specialised production tailored to local or even individual tastes and preferences, and increasingly requires collaboration that is better done locally, he argued.

Moreover, as the costs of local production fall, transportation costs and the risks from disruptions to complex, extended supply chains rise in relative terms.

As an example, Mr Barth cited Adidas, the German sportswear company, which is “seeing a return to local production in high labour-cost Germany due to the combination of technological feasibility and customer demand for immediate customisation of products,” meaning Adidas can create custom-fit shoes on the spot at some retail stores in major markets.

He believes these trends are already evident in macroeconomic data, although the picture is perhaps not entirely clear-cut.

For instance, Mr Barth said foreign direct investment (as a proportion of global GDP) rose “rapidly” in emerging markets in the 1990s and 2000s but has declined since 2006/07 in the EEMEA (eastern Europe, Middle East and Africa) region and since 2011 in China and middle-income EMs, as shown in the fifth chart.

A graphic with no description

It has only continued to trend higher in the “newly-industrialised countries” — Hong Kong, Singapore, South Korea and Taiwan, which he argued are more like DMs in their manufacturing capability — and the poorest EMs, where he said that a significant share of FDI is coming from China “and may be more politically than economically motivated”.

More broadly, Mr Barth said the investment share of GDP fell for roughly three decades prior to the global financial crisis in the developed world as manufacturing capacity shifted to the emerging world. However, it has risen steadily since then, as the sixth chart shows.

A graphic with no description

In contrast, his interpretation of the final chart is that investment shares in EMs generally trended higher from the 1980s onwards, despite the odd crash, such as in East Asia after the debt crisis of 1997, but that this trend has now reversed, with the possible exception of eastern Europe. Others may find the data here less convincing, but there is certainly no clear pick-up since the GFC, as seen in the developed world.

A graphic with no description

Mr Barth’s overall conclusion is a gloomy one. He envisions the potential for a “vicious cycle” in many EMs as decreasing FDI weakens institutions, such as the rule of law, while the broader trend leads EM assets, such as equities and currencies, to underperform those of the developed world.

StanChart’s Ms Jha is somewhat more upbeat. While she too envisaged some loss of demand and jobs from reshoring and automation in DMs, she believed some of this would be offset by the expansion of “south-south” supply chains, based on China, as well as rising domestic demand as EMs themselves become more affluent.

Get alerts on Emerging markets when a new story is published

Copyright The Financial Times Limited 2019. All rights reserved.
Reuse this content (opens in new window)

Follow the topics in this article