China Turns 'New R&D Institutions' Into Core Engines for 2035 Tech Goals

China Elevates Flexible Research Bodies in Its 2035 Innovation Push

China is formalizing a distinct class of research organization — the 'new-type R&D institution' — as the connective layer between state labs, universities, regional innovation zones and industrial commercialization. These entities are defined by three features: no administrative rank, no fixed staffing establishment and no guaranteed fiscal budget. In return, they receive autonomy over research direction, spending and personnel, and are meant to operate under market mechanisms.

The policy push now has legal and top-level backing. A 2019 Ministry of Science and Technology guideline first defined the model at national level; the 2021 revision of the Science and Technology Progress Law codified state support; and the 2024 Central Committee reform decision wrote their encouragement and regulation into the country's core reform agenda. The result is a system in which new-type R&D institutions sit alongside national laboratories, universities and enterprises as an explicit instrument of high-level sci-tech self-reliance.

The scale of the surrounding investment is substantial. According to the article, China's total R&D spending reached 3.93 trillion yuan in 2025, about 2.80 percent of GDP — the first time intensity exceeded the OECD average of 2.70 percent. Basic research spending was 277.8 billion yuan, more than 7 percent of the total. Invention patents in force reached 6.318 million, and PCT filings stayed first globally for a seventh straight year.

Regional clusters are the other half of the design. The government is building three international innovation centers — Beijing-Tianjin-Hebei, Shanghai-Yangtze River Delta, and the Guangdong-Hong Kong-Macao Greater Bay Area — and has called for differentiated local innovation hubs rather than homogeneous competition. In the latest WIPO ranking, China had 24 of the world's top 100 innovation clusters, with the Shenzhen-Hong Kong-Guangzhou cluster overtaking Tokyo-Yokohama for first place.

Where New R&D Institutions Change Innovation Economics

Why the 'New R&D' Label Changes Institutional Behavior

The core logic is that traditional institutes face administrative and evaluation constraints, while new-type R&D institutions are designed to bypass them. The article describes a 'three noes and three yeses' structure: no administrative grade, no fixed establishment, no fixed fiscal allocation, but full research, funding and personnel autonomy. That is a meaningful incentive shift: research teams can spend and hire according to project needs, and funding is tied to performance and market conversion rather than bureaucratic allocation. The risk is the mirror image — institutions must continuously prove value to retain funding and talent.

How Beijing, Shanghai and the Greater Bay Area Are Betting Differently

The regional data point to specialization rather than replication. Beijing is using institutions such as the Brain Science and Brain-Inspired Research Center and Quantum Information Science Research Institute to strengthen basic and frontier capabilities; the article cites a superconducting qubit decoherence time of 503 microseconds as one output. Shanghai is organizing Pujiang Lab and Qizhi Institute under a 'three noes and one comprehensive' budget model and developed the 'Shusheng' large-model family. In the Greater Bay Area, Pengcheng Lab and the Shenzhen Institute of Advanced Technology are explicitly tied to the Guangzhou-Shenzhen-Hong Kong corridor and Hong Kong-Macau connection. Jiangsu's Industrial Technology Research Institute has built a commercialization model around funding-plus-equity, team-controlled entities and contract research, with a provincial regulation taking effect in January 2025.

The Commercial Pipeline Is the Real Measure, Not the Headline Projects

Many of the article's indicators center on links to industry. National technology contract turnover reached 7.57 trillion yuan in 2025, up from 0.64 trillion yuan in 2012, and the patent industrialization rate for enterprise inventions was 53.3 percent. The supporting infrastructure is also dense: over 2,300 university technology-transfer offices, over 200 proof-of-concept centers, roughly 2,400 pilot-test platforms, and 16,000 incubators covering 95 percent of county-level areas. A concrete commercial marker is Hengrui Medicine's 43.7 billion yuan GLP-1 out-licensing deal, cited as a record for domestic innovative drug licensing. The policy interpretation is that Beijing wants new R&D institutions to be the middle layer that converts laboratory breakthroughs into these deals and scale-ups.

The Open Question: Efficiency and Evaluation Reform

Officials acknowledge that R&D investment still faces waste and inefficient duplication. The proposed responses include more performance evaluation, central-local cost sharing, and steering resources toward strategic frontiers. For new R&D institutions, the claimed advantage is that flexible evaluation can avoid an overreliance on papers, titles, degrees and awards. But that also makes their performance harder to compare. Whether the flexible model genuinely improves productivity or simply adds another category of institutes will depend on the monitoring and exit mechanisms that follow.

What Policymakers and R&D Leaders Can Actually Act On

For teams and institutions in or near the named innovation corridors, the actionable next steps are specific:

• For new R&D institution leaders: Adopt or benchmark the operating templates already codified — Beijing's 'five new' mechanism, Shanghai's 'three noes plus comprehensive budget,' or Jiangsu's January 2025 industrial research institute regulation — rather than recreating governance from scratch.
• For regional planners: Use the central directive against homogeneous competition to specialize on existing strengths: quantum and brain science in Beijing, photonics and AI in Shanghai, network communications and cross-border Hong Kong-Macau collaboration in the Greater Bay Area, and conversion-heavy models in Jiangsu.
• For corporate R&D and commercialization teams: Find partners among the named transfer infrastructure — 2,300-plus university technology-transfer offices, 200-plus proof-of-concept centers and 2,400-plus pilot-test platforms — or through cluster-specific institutes such as Pengcheng Lab and Shenzhen Institute of Advanced Technology, where incubation pipelines are already producing enterprises.
• For investors and licensing executives: The policy direction supports early and hard-tech investment, but the model carries no permanent fiscal guarantee; due diligence should focus on an entity's contract research pipeline and conversion record, not only its scientific reputation. The Hengrui GLP-1 out-license is one benchmark, but one deal is not a systematic market.