Donald Trump’s visit to China comes as the country’s new Five-Year Plan places technological self-reliance and frontier scientific innovation at the centre of its strategy for long-term economic and geopolitical competition.
The strategic centrepiece of China’s 15th Five-Year Plan (2026–30), formalised at the March 2026 National People’s Congress, is a commitment to high-level scientific and technological self-reliance. But a tension exists. No amount of industrial scaling can substitute for ‘original innovation’, yet the system charged with delivering these breakthroughs has historically rewarded speed and volume over the patient, risk-tolerant inquiry that frontier science requires.
The plan’s architecture represents a strategic pivot to address this gap. While the 14th Five-Year Plan treated innovation as an enabling tool for high-quality development, the 15th portrays foundational science as central to China’s competitive survival. The cultivation of ‘new quality productive forces’ – rapid technological and scientific innovation and industrial upgrading – signals a shift from scale-driven growth towards total factor productivity gains from frontier research. Without mastering source elements like algorithmic principles, physical models and advanced materials, Beijing argues, the industrial system risks ‘building on sand’ and remaining structurally vulnerable to external disruption.
The plan’s response takes the form of a double helix model, comprising two intertwining strands – long-cycle ‘original innovation’ and industrial applications that convert discovery into productive output. While the 14th Plan delivered impressive industrial outputs, basic research remained underemphasised.
The 15th Five-Year Plan proposes that the strands must be entangled through ‘whole-chain’ promotion. Its instruments include proof-of-concept centres and pilot platforms, a ‘new national system’ for Beijing to deploy lump-sum research funding, and enterprise-led innovation consortia to bridge the gap between laboratory and market.
China’s capacity to execute this agenda is stronger than critical accounts often concede. CATL and BYD together hold over 55 per cent of the global EV battery market. DeepSeek’s January 2025 release of R1 – a reasoning model competitive with frontier systems at a fraction of the cost – challenged the idea that massive compute spending gives Silicon Valley an unassailable AI advantage. And SMIC’s 7-nanometre chip for Huawei’s 2023 Mate60 Pro smartphone demonstrated that export controls are binding but not hermetic.
As Chinese firms continue to gain market share in pharmaceuticals, solar and high-speed rail, the system is proving capable of producing frontier capacity even under pressure, though the strongest cases remain in industrial applications rather than frontier science.
Yet China also faces real structural friction. Projected growth remains sluggish. Persistent deflationary pressures constrain the patient capital commitments that long-cycle science requires. Duplicative construction and capital misallocation continue to absorb resources. These difficulties are compounded by an ageing workforce, strained local government finances and tightening scientific talent exchanges and academic collaborations. Basic research accounted for less than 7 per cent of China’s total research and development (R&D) expenditure in 2024 against 15–25 per cent in leading OECD economies. Despite a decade of policy attention, Beijing has yet to resolve the structural impediment of an evaluation system that rewards quantitative outputs over open-ended discovery.
Perhaps the most consequential factor determining China’s frontier research future is one which the plan’s national-level framing cannot fully address – subnational variation within the country’s innovation geography.
Research on high-tech parks in three of China’s most prominent innovation hubs – Beijing, Shanghai and Shenzhen – reveals distinct differences in focus and configuration. Beijing concentrates basic research infrastructure, anchored by state laboratories, elite universities and central funding channels – a configuration suited to the long-cycle foundational science system that the 15th Five-Year Plan elevates.
Shanghai enjoys a strong municipal developmental apparatus and access to the global innovation ecosystem, generating dense linkages between the state, firms, universities and global markets within strategic sectors including integrated circuit manufacturing and biopharmaceuticals.
Shenzhen is a market-driven manufacturing and hardware ecosystem where private firms, returnee entrepreneurs and contract manufacturers drive rapid iteration with less state choreography – a configuration that produced Tencent, BYD and the drone manufacturer DJI.
The relative weight of these competing innovation models could determine what the 15th Plan ultimately produces. Whether unified national frameworks can align these divergent local models, or whether top-down discipline will compress the diverse local dynamism driving Chinese technological catch-up and innovation, is where the plan’s progress will be most clearly observed.
These subnational variations sit within a broader political economy debate that the 15th Five-Year Plan inherits but does not resolve. The classic developmental state discussion – extended in recent scholarship on how startup-centred policies in East Asia have evolved into hybrid forms blending state direction with market agility – captures the historical mechanisms through which Chinese innovation gained ground. But security-first governance compresses specific conditions – tolerance for failure, open-ended inquiry and decentralised initiative – that long-horizon frontier R&D requires. The plan will test whether Beijing can preserve these conditions while enforcing political discipline.
External pressure cuts in both directions. ‘Small yard, high fence’ technology restrictions, US semiconductor export controls and allied measures have tightened bottlenecks in leading technology sectors, notably advanced lithography systems and AI chips. China is positioning itself to occupy the vacuum left by Washington’s scaling back of research funding, but this remains contingent upon whether it can offset external constraints and resolve the contradictions within its own model.
The 15th Five-Year Plan wagers that a state-mobilised system can cut across subnational differences and cultivate the open-ended inquiry it has historically struggled to sustain. Confident readings in either direction – triumphalist or dismissive – reach beyond what the present evidence supports. How the ambition evolves will matter beyond China itself. What this system can and cannot achieve in advancing frontier science will shape the architecture of global technological competition in the decade ahead.
Republished from East Asia Forum
Steven Jiawei Hai
Dr Steven Jiawei Hai is Assistant Professor of Technology Entrepreneurship at Xi’an Jiaotong–Liverpool University and Affiliate Fellow at King’s Institute for Artificial Intelligence, King’s College London.