Beyond subsidies: what’s really driving China’s industrial climb

*Originally published in the South China Morning Post.
A short version of this article was published as a letter in the Financial Times.
By Zongshuai Fan, Cambridge Industrial Innovation Policy

Engineers, supply chain depth and policy execution matter as much as spending for countries chasing advanced manufacturing

Driven by intensifying competition in advanced manufacturing, the world is waking up to “China shock 2.0”. The first “shock”, associated with China’s accession to the World Trade Organization in 2001, focused on low-tech manufacturing.

Of course, this latest “shock” isn’t framed as such by Chinese policymakers. They present it as an upgrade in export strength, often described as a shift from the “old three” of textiles, furniture and home appliances to the “new three” of electric vehicles, batteries and solar panels.

What many abroad call “China shock 2.0” is a marker of industrial upgrading and a reflection of long-standing national strategies.

Beijing is also not waiting for the world to fully digest this latest wave. As of late 2025, official messaging has begun promoting artificial intelligence (AI), industrial robots and innovative medicines as the “newer three”, pointing to their rising presence in global markets. This is not just domestic cheerleading. These sectors are also drawing growing attention internationally.

It may be too early to place the “newer three” on the same footing as the previous trios, given their relatively combined smaller export base and the likelihood that the composition will change. Still, the trajectory is striking. It took roughly two decades for China to climb from the “old three” to the “new three”. Just years after the “new three” entered the spotlight, Beijing is already signalling what’s next.

Why subsidies are only part of the story

Industrial subsidies are the explanation many reach for. Public funding is indeed part of the story, especially in frontier sectors such as AI. Compared with the United States, China has relied more heavily on state-backed channels than private capital. Government funds invested an estimated US$184 billion in AI firms between 2000 and 2023.

But attributing China’s industrial achievements to subsidies alone is too easy and ultimately less useful. It blocks a more practical discussion of what other countries could learn from China’s experience, especially those that lack deep pockets.

Beyond subsidies, a large engineering workforce and a comprehensive industrial system played an important role in China’s ability to move up the ladder.

Official data shows that China produced more than 5.2 million STEM bachelor’s degree graduates between 2020 and 2022, roughly the size of Ireland’s entire population. However, scale is not the only edge. In 2022, about two in five bachelor’s degree graduates were STEM-trained, higher than other manufacturing powers. The pipeline is still widening, with STEM graduate numbers up about 13 per cent between 2020 and 2022.

Beijing is also trying to steer STEM talent towards the real economy, not finance. Regulators have tightened oversight of parts of the financial sector and imposed salary reforms at some state-linked institutions. As economist Yao Yang has argued, one aim is to shift incentives from finance to manufacturing.

The labour market signals are starting to shift. At Tsinghua University, information technology has reportedly become a top choice for graduates, ahead of finance. Some finance professionals with engineering backgrounds have also pivoted into frontier industries. Entrepreneur Liang Wenfeng, who trained in information engineering, later shifted from hedge fund work towards AI and founded DeepSeek in 2023, as regulatory pressure on the sector intensified.

China’s edge also lies in the industrial system it has built over decades. If we apply NVIDIA CEO Jensen Huang’s “five-layer cake” view of the AI industry, China’s advantage in the infrastructure layer is tied to manufacturing capacity. A recent policy document set a goal of building 100 high-standard industrial data sets by 2027, drawing on China’s industrial breadth to train large language models for a wide range of applications.

Unitree illustrates the advantage of a dense home-grown industrial robotics supplier base. Most of its key suppliers are Chinese, with foreign-sourced components accounting for about 20 per cent. This contrasts with Tesla’s humanoid robot Optimus, where Chinese suppliers are widely seen as central to the supply chain. Unitree’s edge comes from rapid product iteration, reliable shipping and cost controls to its tightly integrated supplier network.

Together, the “old three”, “new three” and “newer three” map neatly onto Beijing’s industrial playbook. The emphasis is to upgrade traditional industries, scale strategic emerging sectors and cultivate future industries in tandem. That approach also captures China’s broader ambition to reinforce industrial capability at every rung of the value chain, rather than betting on a single wave of winners.

Lessons for industrial strategy

As governments pursue industrial strategies to boost international competitiveness, especially in advanced manufacturing, China’s experience offers a practical reminder. Investment matters, but it is often amplified by complementary strengths. When these elements are in place, they turn frontier technologies into capability at scale.