When the Evidence Base on China Becomes a Policy Risk

Europe’s China policy now rests on a widening evidentiary contradiction. The EU wants to de-risk China-related supply chains, screen investments, enforce export controls, protect research cooperation, and manage exposure to economic coercion. Yet these tools depend on reliable evidence about China at the very moment when that evidence base is becoming more selective, unstable, costly, and legally uncertain. China is not becoming unknowable. The problem is specific: Europe has to make policy under conditions of curated access.

Information Has Become a Security Issue

Under Xi Jinping, Beijing has gradually reclassified information as a matter of national security, political sovereignty, economic control, and strategic competition. This reclassification is visible across several legal regimes, from data security and personal-information protection to counter-espionage and state-secrets regulation. What matters here is not the legal detail of each statute, but their shared implication: data generated in China may become sensitive not simply because of its content, but because of what happens to it – when it is collected, aggregated, compared, commercialized, or moved across borders. Ordinary empirical practices can therefore acquire political meaning in new contexts.

This ambiguity lies less in individual documents than in scale and context. A company record, a local procurement notice, a court entry, or a statistical series may look mundane in isolation. In combination, however, such sources may reveal ownership chains, industrial priorities, subsidy patterns, social stress, local debt, or gaps between central policy and local implementation. Categories such as “important data” and “work secrets” remain difficult to delimit. For foreign researchers, companies, journalists, and officials, the boundary between usable public information and security-relevant information has become less predictable.

Selective Access, Not a Blackout

The result is not an information blackout. China still publishes large volumes of policy documents, statistics, speeches, tender notices, local regulations, scholarly articles, and corporate materials. A 2024 MERICS report describes the trend more accurately as a move toward a curated information space. The problem is selective visibility: some information remains abundant, while other materials become delayed, degraded, geoblocked, priced out, legally uncertain, or technically harder to preserve.

The shift is visible in databases routinely used by analysts. Wind Information is a leading Chinese financial-data provider and a core source for market, macroeconomic, and corporate information. China National Knowledge Infrastructure (CNKI) is the main gateway to Chinese academic publications and research materials. Qichacha and Tianyancha aggregate company-registration, credit, and ownership information used for corporate due diligence, ownership tracing, and supplier mapping. Reuters reported in 2023 that Wind Information had restricted offshore access to certain business and economic data, including shareholding structures and information on ultimate controllers. The same report noted overseas restrictions on parts of CNKI, including dissertations, conference papers, legal data, and statistical data, and reported that Qichacha and Tianyancha had stopped serving offshore users for periods. Reuters later reported that Qichacha had passed a data export security assessment and was set to launch an overseas version of its database. This suggests adjustment rather than simple closure, but it also shows that access to corporate and research information is increasingly mediated through compliance, security review, and differentiated user regimes.

Official statistics illustrate a related problem. In 2023, China suspended publication of youth unemployment data after record-high readings. When publication resumed, the National Bureau of Statistics released a revised series excluding students. The point is not that politically sensitive indicators are necessarily false. It is that availability, definition, and comparability can change precisely when outside observers need stable series. Similar frictions appear less visibly elsewhere: in local websites, court databases, policy repositories, procurement platforms, archives, fieldwork, and interviews. Broken links, registration barriers, keyword filtering, geoblocking, and caution among interlocutors can restrict evidence as effectively as formal prohibition.

Why Europe Should Care

For European policymakers, curated evidence is not a niche research problem. It shapes the quality of policy judgment. The EU’s own China framework identifies China simultaneously as a partner, economic competitor, and systemic rival, while its economic security strategy addresses risks to supply chains, infrastructure, key technologies, and economic dependencies. Each of these fields requires granular China evidence. In the end, de-risking assumes that risks can be empirically identified, compared, and distinguished from political noise.

The practical consequences are concrete. Investment screening needs reliable information on ownership, control, beneficial ownership, and local business networks. Export controls and sanctions require evidence about end users, subsidiaries, procurement links, and military-civil or dual-use technology ecosystems. Supply-chain mapping depends on local industrial clusters, customs data, supplier relationships, and environmental or labor-risk records. Research security requires universities to distinguish routine academic cooperation from arrangements that create unwanted technology transfer, foreign interference, or integrity risks. Without sound evidence, Europe risks both underreaction and overreaction: genuine vulnerabilities may remain unaddressed, while fragmentary information may encourage excessive suspicion.

The burden will not fall evenly across Europe. Large member states, EU institutions, major universities, and well-funded firms can buy commercial data, archive older material, maintain language expertise, and commission legal advice. Smaller administrations, media outlets, universities, and many CEE policy communities often work with thinner expert networks and project-based funding. If access becomes more stratified, Europe’s internal asymmetry of China knowledge will grow. Some capitals may build cautious assessments from multiple sources – others may depend on second-hand summaries, commercial dashboards, diplomatic reporting, or politicized narratives.

Such asymmetry can fragment policymaking at the EU level. Foreign direct investment screening already relies on information sharing between member states and the Commission; following the Council’s adoption of the revised EU FDI screening regulation in June 2026, coordination is set to become more consequential. Export-control enforcement, anti-coercion policy, critical-infrastructure protection, sanctions design, and strategic communication also require a shared evidentiary base. If governments assess China with unequal data access and uneven analytical capacity, common positions will rest on uneven assumptions. Evidence asymmetry can thus become policy asymmetry.

Building Evidence Resilience

Europe should not respond to curated access with speculation or blanket suspicion. It needs evidence resilience: the capacity to preserve vulnerable sources, diversify methods, state uncertainty explicitly, and protect interlocutors and research partners. Analysts should triangulate official documents with procurement and court records, corporate filings and trade data, satellite imagery, archived web material, and carefully handled interviews. The aim is neither to replace Chinese sources with external substitutes nor to confuse open-source intelligence with area knowledge. Rather, it is to be explicit about what each source can show, what it cannot show, why it remains visible, and what alternatives it excludes.

Institutions matter as much as methods. Universities, libraries, think tanks, ministries, and EU bodies should preserve vulnerable Chinese webpages, statistical tables, catalogs, local policy documents, and database documentation within legal and ethical limits. They should build shared bibliographies, versioned datasets, web archives, and governed repositories, with clear metadata standards. Such infrastructure would reduce dependence on individual hoarding and opaque commercial brokers. It should serve policymakers as well as scholars, especially when a briefing depends on knowing when a source changed, disappeared, or became non-comparable.

Training must also change. China expertise now requires digital source criticism, awareness of China-related legal risks, data ethics, preservation skills, and uncertainty reporting alongside language competence, historical depth, and institutional knowledge. Cooperation with scholars and interviewees in China remains indispensable, but data-protection rules and political uncertainty require more careful consent, attribution, anonymization, secure storage, and risk assessment. European research security should remain open and proportionate. It should protect people and knowledge without turning collaboration into suspicion by default.

Curated access does not make China research impossible. It makes the link between source, inference, and policy judgment more fragile. Europe’s China policy therefore needs a clearer evidentiary discipline. Briefings should distinguish observation from inference, note when sources are missing, inaccessible, or non-comparable, and assign confidence levels. That is not an admission of weakness. It is what proportional de-risking requires: policy tools calibrated not only to what Europe knows about China, but to how stable and verifiable that knowledge is.