Misplaced interest

Govt equity in chip design firms is impractical

The government’s efforts to stimulate a domestic semiconductor industry are well-meaning but could fritter away taxpayer money. It has launched a two-pronged effort in this regard. First, it has offered to underwrite 30-50 per cent of the expenditure on foundries, depending on the nature of the semiconductors being manufactured. Second, it is offering some Rs. 55,392 crore production-linked incentives for electronic hardware and semiconductors. The third element, announced in 2022, is the design-linked incentive, or DLI, which offers financial incentives and design infrastructure support to domestic companies, start-ups, and small and medium enterprises for up to five years. The aim is to nurture at least 20 domestic companies involved in semiconductor design and enable them to achieve a turnover of more than Rs. 5,000 crore in the next five years. In the second stage of this scheme, the government is considering picking up equity in domestic chip design companies. The idea is to forestall domestic design companies from selling stakes to global players.

The desire to retain intellectual property rights for chip manufacture within the country is sound in theory. The absence of a domestic ecosystem for semiconductors has resulted in a brain drain of sorts, with leading Silicon Valley chip majors setting up design centres in India to deploy India’s vast talent base. The core of the value addition in the semiconductor business lies in design rather than manufacture. This is why Silicon Valley majors such as Intel, Qualcomm, Nvidia, and later AMD and others chose not to invest in expensive manufacturing processes. Instead, they focused on exponentially refining and upgrading technology and farming out manufacture to cheaper locations in Asia such as Taiwan, Singapore, South Korea, and Japan in the 1970s to create a mammoth global supply chain. By focusing on intellectual property, Silicon Valley wields enormous power over the industry without dominating the manufacturing process. The proposal for the Indian government to take equity stakes in domestic chip design companies can be seen in this context.

The problem with this scheme is that it ignores the characteristics of the industry. The semiconductor design industry thrives on a hyper-competitive innovative environment that relies on speed and, importantly, the ability to absorb serial failures. Fostering this chaotically creative ecosystem will require the government to miraculously transform itself overnight. Clunky, time-consuming approval systems that require designers to submit proof of concept to qualify for investment won’t help. Second, the Indian government has little or no track record of tying up successfully with the private sector. The experience of companies, such as Balco and Hindustan Zinc, in which the government retained a stake after disinvestment, does not engender confidence in its ability to encourage the requisite competitive dynamism to make India a centre of domestic chip design. A more constructive approach would be to create the conditions to incentivise domestic chip designers by helping them create the markets for products in India and abroad. This will partly require reducing frictions in ease of doing business. Silicon Valley thrived by focusing on the fast-growing market for consumer goods. Selecting industry champions in isolation from the broader economy can yield only minimal results.