Editor’s Note — DCC.
This is the theory piece under the theory pieces. Published in 《政法论坛》in 2021 — before the Data 20 Articles — and cited 255 times since, Xu Ke’s “Data Rights: Paradigm Integration and Normative Differentiation” is the structural-theory scaffolding that the three-rights framework (hold / use / operate) and the data-source-rights debate both build on. For overseas counsel, its value is diagnostic: it explains why Chinese data rights are structured the way they are, and why the Western property / IP / license analogies keep failing to map. DCC summarizes the framework; the deep jurisprudential argument is in the original.
The two warring paradigms
Xu Ke frames the data-rights field as a contest between two paradigms:
- Formalism — the “empowerment” model (赋权模式). Treats data rights as a new property-style right to be defined and assigned: who owns the data, what the owner can exclude, how the right transfers. The instinct is to build a data-ownership right analogous to property or IP.
- Substantivism — the “conduct regulation” model (行为规制). Skeptical of defining a data-ownership right at all; instead regulates behavior — what processors may and may not do — through tort, competition, and data-protection rules, without ever vesting a property-style right.
Each has a fatal weakness. The formalist empowerment model struggles because data isn’t naturally exclusive or rivalrous — a single dataset can be held and used by many parties non-exclusively, which defeats the property analogy. The substantivist conduct-regulation model struggles because pure behavior-regulation can’t support the market the data-element economy needs — you can’t trade what you can’t define a right over.
Xu Ke’s move: the two paradigms are “different roads to the same destination” and must be integrated — through a third paradigm he calls reflexive law (反省法), which steps back from both to ask what structure data rights actually require.
Three possible rights structures
The integrating insight is structural. Xu Ke distinguishes three ways a “right” can be built:
- Rights-ball (权利球). A unitary, indivisible right — like classical ownership. One owner, one solid sphere of entitlement. Doesn’t fit data: data’s value comes precisely from being usable by many parties simultaneously.
- Rights-bundle (权利束). A bundle of separable sticks — like the Anglo-American “bundle of rights” property concept. Better, but Xu Ke argues it’s too loose: the sticks are enumerated but not structured, so the bundle doesn’t explain how the rights relate or cohere.
- Rights-block (权利块). Xu Ke’s proposal. A structured set of rights with a shared core and differentiated manifestations — capturing data’s “理一分殊” character (one underlying principle, many concrete manifestations). The block has an “overall design rule” (整体设计规则) that gives the rights coherence and “individual design rules” (个别设计规则) that differentiate them by scenario.
The “data rights block” (数据权利块), Xu Ke argues, both integrates and improves China’s existing “separation of powers/functions” (权能分离) theory — the theory that the Data 20 Articles operationalized into hold / use / operate. And because it’s a block (structured) rather than a bundle (enumerated), it connects coherently to the surrounding institutions: data security, data trading, statutory data use, data opening and sharing.
Why this matters for the three-rights framework
The Data 20 Articles’ three-rights structure (hold / use / operate) is, in Xu Ke’s terms, a rights-block: three rights sharing a common core (the underlying data), differentiated by function, structured rather than merely listed. This is why:
- The three rights are severable (you can hold one without the others) — that’s the “individual design rules.”
- The three rights are non-exclusive (multiple parties can hold the same right over the same data) — that’s data’s non-rivalrous nature, which the rights-block accommodates and the rights-ball cannot.
- The three rights cohere (they’re not a random bundle but a structured set) — that’s the “overall design rule.”
For overseas counsel who’ve read NDA’s interpretation of the three rights, Xu Ke’s piece is the theoretical answer to the question “why is it structured this way?” The structure isn’t arbitrary; it’s the rights-block design responding to data’s specific properties.
Why Western analogies fail
The piece’s most useful contribution for overseas counsel is implicit: it explains why the instinct to map Chinese data rights onto Western property, IP, or license concepts keeps failing.
- Property (ownership) analogy fails because it’s a rights-ball — unitary and exclusive. Data rights are a rights-block — structured and non-exclusive.
- IP analogy fails because IP protects single-author creation with exclusive rights; data is multi-party co-created and non-exclusive.
- License analogy partially works but misses the structure: a license is a grant from an owner; the rights-block has no single owner to grant from — the rights are structurally co-held.
The lesson: stop looking for the Western analog. The rights-block is a genuinely different structure, designed against data’s properties (non-rivalry, multi-party creation, low replication cost). Counsel who internalize the structure operate the regime; counsel who keep reaching for analogies stay confused.
What this tells overseas compliance teams
-
Use the rights-block as the mental model, not the property/IP/license analogy. When you encounter Chinese data-rights vocabulary (hold / use / operate, data-source rights, etc.), parse it as a structured rights-block — shared core, differentiated and severable manifestations, non-exclusive — rather than trying to find the Western equivalent. This is the single most useful conceptual correction for cross-border data lawyers.
-
The “non-exclusive” property is load-bearing in contracts. Because the rights-block accommodates multiple parties holding the same right over the same data, your Chinese data agreements can (and increasingly do) allocate non-exclusive use rights to multiple parties. Don’t assume exclusivity as a default; specify it where you need it.
-
The theory predicts the rulemaking. Xu Ke’s 2021 framework predicted the structure the Data 20 Articles (2022) adopted and the Data Property Rights Registration Guide draft (2025) is operationalizing. Tracking the leading academic structural theory is a forward indicator of where registration, trading, and rights-confirmation rules go next.
-
“理一分殊” is the design philosophy to internalize. One underlying principle (the data), many differentiated manifestations (the rights, by scenario). This is why Chinese data rules proliferate scenario-specific sub-rules (sector catalogues, FTZ negative lists, public-data vs enterprise-data vs personal-data regimes) rather than one unified rule. Expect differentiation by scenario as the structural norm, not the exception.
The structural takeaway: Chinese data-property law is built on a deliberate, theorized structure — the rights-block — not an improvised borrowing from Western property law. Xu Ke supplied that structure before the policy adopted it. For overseas counsel, the piece is the decoder ring: it tells you what kind of thing Chinese data rights are, which makes every downstream rule legible.
— 许可, 数据权利：范式统合与规范分殊 (Data Rights: Paradigm Integration and Normative Differentiation), 《政法论坛》Issue 4, 2021, pp. 86-96; reposted via 政法论坛 WeChat Official Account. Original article (Chinese).
Not legal advice. The above is DCC’s structured summary of Xu Ke’s analysis, with framing for overseas counsel; the paradigm-integration argument and the rights-ball / rights-bundle / rights-block taxonomy are Xu Ke’s.