Worrying subsidy race in building high-tech industry

The consumption of high-tech products across the world has been increasing. Mobile phone was a luxury not long ago, even in advanced economies. On the other hand, conventional products are getting high-tech. For example, automobiles are no longer mere mechanical machines. An average modern car has as high as 1500 microchips. The role of microchips and software has been increasing for creating sensing, perceiving, communicating, and decision-making capabilities in products, turning them into artificially intelligent (AI) machines. As a result, products are becoming capable of responding more effectively, safely, and efficiently. For example, intelligent farming machines adapt their actions to the varying soil and crop conditions-resulting in higher yield, better quality, and less wastage. Hence, due to the growing role of microchips in machines in making them intelligent, both the size and importance of the microchip-making semiconductor industry have been growing. Therefore, every major economy has been after establishing and expanding its footprint in the high-tech sector. Hence, there has been a subsidy race for showing quick result, instead of creating snowball effect out of specialisation. But does it create a worrying situation?

The backbone of the high-tech industry has been semiconductors. In addition to its growing volume reaching $1 trillion by 2030 from $600 billion in 2022, the supply of microchips has been crucial for producing all kinds of products. For example, during COVID, microchip shortages halted production in many industries. For instance, in 2021, the global automotive industry suffered a significant loss; manufacturing plants remained idle, resulting in eleven million fewer vehicles and loss of $200 billion in revenue. Hence, a growing number of countries have been linking the semiconductor industry to national security issues. Besides, in addition to volume and its increasing importance, semiconductor, and other high-tech industries are sources of high per capita income, primarily due to value addition scope out of knowledge and ideas. Therefore, building a high-tech industry has been a development agenda for increasing per capita income, reaching and sustaining high-income status, and ensuring supply.

Despite growing revenue in high-tech industries, a few countries have monopolistic positions. For example, more than ninety per cent of high-end microchips are produced in Taiwan. On the other hand, although India's consumption of semiconductors is expected to cross $80 billion by 2026 and $110 billion by 2030, no microchip is produced in India, except for a few for space and defence applications. Furthermore, due to the consolidation of microchip production in Taiwan and South Korea, global market shares of Japan, the USA, and Europe have substantially reduced. But the USA has close to 50 per cent market share in the idea-based innovation and design segment. On the other hand, in semiconductor equipment and chemical, USA, Japan, and Dutch-led Europe have a monopolistic situation. For example, Dutch ASML is the only producer of extreme ultraviolet lithography machines.

Over the decades, there has been a race to leverage economies of specialisation in making high-tech products better and cheaper. This race has been in producing ideas and integrating them into products and production processes, resulting in the growing role of human capital in idea creation and the diminishing role of labour in production. Besides, economies of scale, scope, and externality effects have been growing. Hence, conventional factors like land, utilities, physical infrastructure, and wage differentials are suffering from an eroding role in determining production cost. As a result, an effort to develop new nodes of high-tech production in the green field has been facing increasing barriers to economies of production. Hence, there has been a subsidy race to attract high-tech production. Despite this, major high-tech companies have been complaining about higher production costs in new territory. For example, despite receiving subsidies from USA's $52 billion Chips fund, TSMC claims that clients in the U.S. will have to pay up to 30 per cent more for chips made in America than for chips made in Taiwan.

On the other hand, India's ambitious $30 billion subsidy fund for building high-tech industry by incentivising foreign firms is yet to show results. In addition to offering as high as 75 per cent subsidy in capital expenditure, India has also been offering subsidies at the output level through its production-linked incentive scheme (PLI). It has been learnt that the $24 billion PLI programme provides a 4 per cent to 6 per cent subsidy to incremental production, whether for domestic consumption or export. For example, for assembling each iPhone, Foxconn has been receiving 6 per cent of the invoiced value, resulting in a $60 subsidy for each iPhone's production, with a $1,000 price tag. But the value addition from local labour in assembling each iPhone is as low as 2 per cent of the total cost. Furthermore, in a recently announced investment deal with Micron Technology, India has offered a $1.95 billion subsidy to attract an investment of $825 million in a semiconductor assembling and packaging plant, creating 5,000 low-end jobs.

Why is there a need to offer growing subsidies for incentivising foreign high-tech firms? Not long ago, the availability of land, basic infrastructure, favourable labour law, and liberal duty structure were good enough to attract high-tech foreign firms in less developed countries. For example, to attract Intel to seed the semiconductor industry's formation in the Penang paddy field in 1971, Malaysia had not to dish out any cash subsidy. In the 1960s, even there was resistance to the entry of foreign high-tech firms for sourcing low-cost labour in less developed countries. As explained, due to the growing economies of specialisation, the high-tech industry value chain comprises only specialised nodes. Worth noting that, more or less, all these hard-to-replicate nodes started in a humble form and grew through the cumulative effect of specialisation and cluster formation.

Initial analysis indicates that a massive subsidy-led approach in attracting high-tech investment will not likely generate a net gain, particularly in less developed countries. For example, no analysis suggests that wage earning by 5,000 low-end jobs in the forthcoming Micron plant in India will generate more gain than the given $1.95 billion subsidies. Furthermore, massive subsidies being offered by major economies have been creating insurmountable barriers to the aspiration of less developed countries like Bangladesh. Consequentially, investments made in building infrastructure of high-tech parks and expanding higher education in science, technology and engineering in Bangladesh and many other less developed countries run the risk of failing to attract foreign high-tech firms.

The focus should be on creating the snowball effect from specialisation and cluster formation. Instead of attempting to leapfrog by attracting multinationals through massive subsidies, the focus could be bottom-up. It may start through the local capacity for pursuing microchip-based incremental innovation of products being produced, whether for local consumption or export. Hence, instead of joining the race of massive subsidies and mega infrastructure development to attract high-tech firms for creating low-end jobs, the focus should be bottom-up through incremental innovations-creating a snowball effect of specialisation.

M. Rokonuzzaman, Ph.D is academic and researcher on technology, innovation and policy. Zaman.rokon.bd@gmail.com