Loading current evidence…
Public records only. Your private brief and follow-up notes are not included.
Ask what this evidence establishes, what remains open, or what to check next.
Analyses written up from the corpus. Most regenerate with the pipeline; the local-protection review is hand-written.
How to use the terminal’s corpus-wide research tools, then return to the evidence workspace.
Sixteen strategic sectors on policy and trade indicators, with robustness checks and one pre-specified test; generated from code.
How the responsiveness decline was re-tested: estimands, rival explanations, decision rules, deviations.
How long local governments take to cite and act on central policy.
The crawl frontier: what the corpus cites and does not yet hold.
Five findings in two pages: substitution, trade, Switzerland, transmission, the 15th plan.
Claims against approvals in MIIT's settlement rounds, 2016–2019, and why claims were cut.
Hand-reviewed cases where local rules contradict central bans on local protection.
Eligibility clauses that admit only locally registered applicants: the screen, its hand check (56/60) and the rates by level.
Share of each directive copied into its provincial restatement, 2013–26 (exploratory): rising to 2018, lower in 2025–26 on few documents.
Keyword-screen leads: local documents below central bans on local protection.
Firm lists in policy documents: who is listed, for what, by ownership.
Trending terms, tree activity and announced documents from gov.cn news.
The pre-registered test: the decline is not statistically established (0.61, 0.28–1.07).