Development impact of the new technology revolution

The geopolitical implications of advancing automation are quite inimical to workers in developing economies like Sri Lanka, whose comparative advantage lies essentially in “embodied labour”, which comprises cheap goods made by low-wage workers, or providing domestic and construction work overseas, Sri Lanka’s highest foreign exchange earner, now for a long time greater than tea exports

By C.R. de Silva

Development issues raised by automation and AI

The digital revolution which was begun in the West, and is now progressing in Japan, China, Singapore and even India, is basically the irreversible and inexorable march of history, initiated during the Industrial Revolution and progressing through globalisation and the spread of information and communication technology (ICT), to the adoption of wide-ranging innovations. These innovations include sensors, robotics, automation and machine learning, resulting in labour saving and job reduction, which will eventually lead to their selective elimination. 

The most comprehensive treatment of this transformation and its implications are presented in the recent publication ‘Mastering Digital Transformation’ by Dr Nagy Hanna, former World Bank ICT expert. The amazing developments he describes have occurred throughout modern history, resulting in today’s stagnant wages, shrinking jobs and rising income inequality, lowering the living standards of affected cohorts. Effects will be felt more dramatically in developing economies, where the share of routine employment has already fallen by 8% (1996-2012), while highly skilled non-routine jobs have increased.

Brookings Institution research in Washington D.C. has confirmed that while containerisation of long-distance transport and the application of computer technology has enabled the expansion of global supply chains, facilitating location of production of various consumer goods, components and parts in cheap labour locations throughout the world, China may be the last country to use low- and middle-income labour, to ride the wave of globalised industrialisation to relative prosperity, due to offshoring of high-wage, manufacturing jobs from the developed West. 

But the process is now reversing with the re-shoring of jobs to their origins, due to robotics, machine learning and 3D printing, resulting in premature, global, labour-centric de-industrialisation, posing real challenges to developing economies. Therefore, low labour costs in developing economies like Sri Lanka will not insulate them from the effects of automation, as evidenced by rising robot sales in China, which recently overtook the US as the largest market for industrial automation, as well as in Singapore, Thailand, even India (as see later). 

Due to the potentially transformative impact of these ongoing trends in the technology revolution on labour markets in developing countries like Sri Lanka, decisive policy responses are needed to change both industrial policy and export strategy, to ensure that corresponding gains are shared with industrialised economies; and that safety nets are improved and extended to losers in shrinking labour markets. 

These actions will include changing the appropriateness, focus and quality of current education, including university, polytechnical and vocational training, installation of fast broadband networks, fiscal incentives to protect employment and a reasonable level of protection and subsidies not only for nascent local industries, but also through an effective safety net to protect adversely affected and vulnerable workers from labour market distortions as well as disruptions.

In their recent book ‘The Second Machine Age,’ Brynjolfsson and McAfee of the Massachusetts Institute of Technology (MIT), suggested several initiatives which will help expand employment opportunities and promote economic growth, e.g. : continue mass education with vocational training and life-long learning; extend the classroom to cover study at home; instil a culture of continuous learning and improvement; incorporate digitisation and analytics into teaching and learning models; attract and retain well qualified teachers; introduce longer school hours and terms; have rigorous student as well as teacher testing, as in Singapore and South Korea; emphasise recognition of hard-to-measure skills, like creativity and problem solving, which should be fostered; introduce strong incentives for achieving hard-to-measure goals; and encourage highly motivated self-starters to take advantage of available online education resources, about which more later.

These authors add that entrepreneurship is an innovation engine and needs encouragement, as well as financial assistance. Following the precedent set by Japan, in Korea the Government financed and supported the Korea Technology Development Corporation (KTDC) in the 1970s and 1980s, which provided support and seed capital for start-ups, pioneering innovation in manufacture and technology; and Korea’s example was later followed by other successful economies like Taiwan and Indonesia. It was found that net job creation is much higher at start-up companies. In fact, Apple and Microsoft had started with one employee – their own founders! 

Excessive regulation and taxation obstructs entrepreneurship and reduces start-up activity. Ways should be found to facilitate matching the right people with appropriate jobs. Local and national data bases of job opportunities and candidates can have huge pay-offs, and should be incentivised. Government in collaboration with the private sector should upgrade allied infrastructure, like broadband width, which like education and research, is subject to positive externalities.  

These are difficult issues to address adequately due to low government interest and effectiveness, and less resilient political and economic systems in developing countries, like Sri Lanka, which require consensus decision-making – essentially to overcome availability of only limited budgets, widely under-performing governance indicators, doubtful political stability, deficient regulatory regimes, and imperfections in the rule of law, control of corruption, etc., and already high, widespread and rising levels of income inequality. 

When Stanford University made its ‘Introduction to Artificial Intelligence’ online course open and free of cost about four+ years ago, 160,000 people in some 190 countries signed up for it; and 100,000 enrolled in a similar online ‘Machine Learning’ university course. Both ran for 10 weeks. In the ICT industry, these courses are called ‘massive open online courses’(MOOCs). 

Harvard University and MIT also entered the MOOC field as early as 2012, demonstrating the huge potential for easily restructuring education and training systems, and demonstrating that AI pioneers are discharging their ethical responsibility to help address structural problems ICT advances have caused, mostly to development strategy and trade policies, as well as in labour markets. 

In developing countries like Brazil, an adaptive-learning startup is guiding high school students in thousands of schools; and MOOCs have proliferated now on numerous for-profit, online websites, and are re-educating thousands worldwide. 

The era of continuous, life-time education and training, combined with work, in order to stay current in the fast-moving world of the new technological revolution has indeed arrived, if both economies and human beings take the challenge that thinking machines have mounted on our unsuspecting universe.

Automation’s impact on development

The geopolitical implications of advancing automation are quite inimical to workers in developing economies like Sri Lanka, whose comparative advantage lies essentially in “embodied labour”, which comprises cheap goods made by low-wage workers, or providing domestic and construction work overseas, Sri Lanka’s highest foreign exchange earner, now for a long time greater than tea exports. 

The emerging risk now is lack of skilled local worker support for implementing development projects. When automation makes rich countries more self-sufficient in these respects, the demand for products and services driving growth in the developing world will erode, and with it their comparative advantage. 

Premature de-industrialisation is already on the horizon, with manufacturing employment said to peak soon in Brazil, China and even India. Sri Lanka among other emerging economies will need to find new growth models, as conventional development strategies falter, labour markets fracture and export markets thin out due to greater sophistication in consumer demand. 

The primary concern is how to create meaningful and rewarding jobs in a context of inclusive growth in the face of widening income and wealth disparities, which could disrupt social cohesion, even create unrest and complicate the capacity of policymakers to address the challenges of employment creation and economic growth. On the other hand, the same advances in technology carry unprecedented potential in their wake to boost productivity, increase incomes and broadly improve living standards.

(The writer was a member of the former Ceylon Civil Service, and was later a senior professional at World Bank Headquarters for over 20 years).