China's Limited Role in Semiconductor Innovation
In recent years, fears have become widespread in the US that China's rise in industrial manufacturing and innovation will challenge America's leadership in advanced manufacturing.
An important concern is that China's indigenous innovation policy gives rise to a model of state capitalism that unfairly favors domestic producers at the expense of foreign firms, undermining innovative capacity and competitiveness and harming the prosperity of business and workers.
A different story emerges, however, when we examine the state of China's semiconductor industry. The global semiconductor industry critically depends on sales to China, where many electronic products are manufactured, yet China's own role in semiconductor production remains limited.
The gap between China's semiconductor consumption and production keeps growing, from US$5.7 billion in 1999 to a record US$100.5 billion in 2011. And China's semiconductor trade deficit more than doubled since 2005 to US$138 billion in 2011, due to growing imports of advanced microprocessors for the manufacture of wireless communications products, especially from US firms.
As for microprocessor innovation, China remains way behind the technology frontier in both fabrication and design, reflected in a weak portfolio of essential semiconductor patents. China still has a long way to go before it can shape, or even co-shape, the industry's technology trajectory.
This poses an interesting advanced manufacturing puzzle: Why is it that, despite massive government efforts to build indigenous innovation and production capabilities, China still plays a limited role in semiconductor design and production?
One popular explanation provided by trade economists points to state direction of industry as the main culprit, arguing that "technology nationalism" and "infant industry protectionism" prevent China from reaping the "gains from trade" that would facilitate learning and productivity improvements.
That explanation fails to convince, however. Joseph Stiglitz and Bruce Greenwald offer a sophisticated theoretical critique arguing that, for developing countries, the benefits of broad trade restrictions may outweigh their costs. And former World Bank Chief Economist Justin Yifu Lin, one of China's most respected economists, highlights fundamental flaws in economic theories that neglect to analyze how institutions and industry structures in developing countries differ from those that have facilitated economic growth and prosperity in developed nations. For Lin, government policies can make companies more competitive in an open market, provided they reflect the economic structure and institutions of a particular country and industry.
A defining characteristic of China's semiconductor industry is its deep integration into the global semiconductor value chain through markets, investment and technology, but China's leadership views such deep global integration primarily as a threat to its domestic innovation capacity, rather than an opportunity. An important concern is that foreign firms with manufacturing plants in China dominate the country's semiconductor consumption, and that these foreign firms retain control over key technologies, core components and capital equipment, as well as chip design and system integration capabilities and process know-how.
The semiconductor industry is one of the priority targets of its indigenous innovation policy, including a plan to create a group of glob¬ally competitive semiconductor firms that will develop into global leaders in market share, manufacturing excellence and innovation capacity. Indigenous innovation in semiconductors is based on the idea that access to foreign core technologies may be too costly or restricted, and hence China needs to shift the balance from technology imports to domestic R&D and innovation in advanced manufacturing. The overriding concern is to replicate as much as possible of the semiconductor value chain in China and to use foreign technology only to create new technologies based on intellectual property rights owned by Chinese companies.
Much of the official indigenous innovation policy doctrine remains married to the idea that China needs to develop leading-edge process technology at home, but this approach neglects the critically important demand side of innovation and, most importantly, China's dearth of complementary "soft" capabilities related to patenting, licensing and the capacity to coordinate multi-layered international innovation networks.
China's indigenous innovation policy thus tends to neglect that the integration of Chinese firms into the global semiconductor value chain enables them to source foreign technology and complementary capabilities that would be too costly and time-consuming to develop at home. As the CEO of Spreadtrum, one of the few successful Chinese chip design companies, told me recently: "The availability of circuit design tools, semiconductor fabrication services and open-source smartphone software allows Chinese firms to circumvent their weak spots and develop their strengths in hardware, design, and integration."
In fact, Chinese semiconductor firms are exposed to intense competition that is driving down profit margins, reducing funds available for capability development, R&D and innovation. These firms are eager to reap the benefits of global production and innovation networks. Global technology sourcing through returnees from the US or Chinese expatriates who are US citizens often plays an important role in accelerating the speed of learning and the exposure to international management practices.
My argument, in a nutshell, is that in China's semiconductor industry two innovation strategies co-exist, but with only limited interaction: the government's emphasis on indigenous innovation and firm-level efforts to enhance global technology sourcing. China's fundamental advanced manufacturing challenge is to find ways to combine the benefits of both innovation strategies.
There is no doubt that for a latecomer to advanced manufacturing like China, strengthening domestic innovative capacity is the key to a sustainable economic transformation beyond the export-oriented "global factory" model. Without smart innovation policy in support of domestic capabilities, China will find it difficult to catch up with the productivity and income levels of the United States, the European Union, and Japan.
In addition, China's leadership feels that intellectual property and technical agreements that are part of its World Trade Organization membership constrain options for national industrial and innovation policies that were available earlier to Japan, Korea and Taiwan as they developed their economies. Attempts to wriggle out of some of those compliance constraints may lead to a sharpening of China's indigenous innovation policy.
Furthermore, Beijing fears that America's push for an advanced manufacturing renaissance, as laid out in the administration's Advanced Manufacturing Partnership strategy, will erode China's competitiveness in manufacturing, implying that China needs more than ever to strengthen its domestic innovation base in order to upgrade its manufacturing ¬– which it has been doing in specific target industries such as lasers, nanotechnology, advanced computing and new energy technology.
Yet, while strengthening domestic R&D and innovation is necessary, it is not likely to be sufficient for bootstrapping China's semiconductor industry into the primary league. China has reached a level of development in this industry where indigenous innovation needs to be combined with sophisticated global technology sourcing. Now is the time for China to identify and counter unintended negative consequences of an indigenous innovation policy that places too much emphasis on technology autonomy.
As long as indigenous innovation policy and the real-world practice of global technology sourcing co-exist in isolation, the result will be a fragmented innovation system that is ill-equipped to address both the opportunities and the challenges that result from China's deep integration into the global semiconductor value chain. In addition, this fragmentation adds further to China's deeply entrenched innovation barriers that reflect its status as an industrial latecomer and the legacy of the planned economy.
It will take time to overcome those barriers, which include severe quality problems in education; plagiarism and derivative research in science; and a fragmented innovation system that reflects intense rivalries between the central and regional governments, as well as between different government agencies. In addition, government R&D support and procurement aids state-owned enterprises and neglects small, innovative private firms, while a still weak intellectual property rights regime stifles entrepreneurship, innovation and private R&D investment. Most importantly, China lacks capabilities that are necessary to transform new ideas, discoveries and inventions into commercially successful innovations, such as sophisticated approaches to patenting, licensing, and standardization.
China's persistent innovation gap in core sectors of the information and communications technology industry implies that Chinese firms need access to American technology, whether in terms of equipment, core components, software or system integration. Thus, China‘s upgrade of its semiconductor industry could create significant new markets for American firms, provided they stay ahead of the innovation curve. Not only is such US-China cooperation in the semiconductor industry possible, it could become a powerful "win-win" proposition. The Chinese would get more of the technologies they want and in some instances still desperately need, and the US would get new quality jobs resulting from exports of US technology to China.
Such cooperation, however, needs to be on an equal footing, with reciprocity of rights and obligations on contentious issues, such as the right balance between the protection of intellectual property rights and China's interest in technology diffusion. China needs to put something more substantial on the table, providing safeguards against forced technology transfer though policies such as compulsory licensing, information security standards and certification, and restrictive government procurement policies. The US, in turn, needs to acknowledge that Chinese firms feel disadvantaged by restrictions on Chinese foreign direct investment and on the export of so-called dual-purpose technologies to China.
Establishing such reciprocity between countries at such different stages of development will not be easy, but progress toward adjusted rules of reciprocity should be possible if the United States and China accept that while their economic institutions and innovation systems are different, they are deeply interdependent.
(Dieter Ernst is a Senior Fellow in economics at the East-West Center in Honolulu, Hawaii, focusing on global production networks and R&D internationalization in high-tech industries. This was first published by the EastWest Center.)