Abstract
The imprisonment of an activist and the treatment of that person’s relatives are distinct subjects of inquiry. This article sets out a method for documenting reported pressure on relatives in mainland China while protecting families from further harm. It does not claim to have established a national case count or a representative prevalence rate.
Research question and definitions
The central question is whether a documented action against a relative can be connected to the detainee’s protected speech, association or other politically relevant conduct. The term “political prisoner” is operational rather than self-proving: every study record should document the detention, the relevant charge or proceeding, and the basis for political relevance. The family-level incident must be established separately.
Unit of observation and inclusion threshold
A case is a distinct family-level event with an approximate date, location, reported action, source, and evidence classification. A prisoner record is not itself evidence of an incident against a relative. Multiple reported restrictions affecting one family should be linked under one family identifier rather than counted as unrelated families.
Mechanisms requiring investigation
Potential mechanisms include interference with visitation or communication, questioning or threats, employment or housing pressure, education-related discrimination, movement restrictions, online harassment, and pressure on relatives abroad. These are hypotheses to test, not findings that each mechanism occurred. Researchers must distinguish ordinary consequences of incarceration from a targeted retaliatory measure and identify the basis for attributing the action to an official or other actor.
Evidence hierarchy
A dated official record or directly inspectable contemporaneous document can support a narrow documentary finding. First-person testimony supports an attributed account; an independent source may corroborate it if that source has genuinely separate access to the underlying event. An NGO statement establishes that an allegation was made, but does not by itself establish the underlying act. Repeated republication of one report is not independent corroboration. Unverified leads should remain outside any public confirmed-case tally.
Case-record schema
For each incident, record a case ID, anonymized family ID, detention-case reference, event date and reporting date, alleged action, alleged actor, original source and language, corroboration, conflicting evidence, authority response where available, current status, confidence tier, and disclosure restrictions. Preserve the exact language of relevant testimony in a secure research archive; public paraphrases should not exceed what the record supports.
Source and selection bias
Public testimony and advocacy reporting are likely to overrepresent families who can reach international networks or who are willing to speak. Silence does not establish the absence of harm; publicity does not establish its frequency. A purposive case inventory can illuminate mechanisms but cannot produce a defensible national prevalence estimate without an appropriate sampling design.
Privacy and informed consent
Public records may contain identifying details that should not be republished. Withhold children’s identities, school names, addresses, travel itineraries, unpublished medical information and contact details unless a necessity and safety assessment clearly supports disclosure. Separate the confidential source key from any public case matrix. Offer correction and withdrawal-review channels, and respect a family’s refusal to participate.
Legal and remedial analysis
Assess domestic procedural protections and international human-rights standards separately, with attention to the legal status and territorial application of any treaty invoked. Relevant protective demands include reliable custody information, meaningful access to counsel, appropriate family communication, non-retaliation for lawful advocacy, impartial investigation of credible abuse allegations and effective remedies. Do not infer a specific legal violation without a documented act and applicable legal standard.
Findings and limits
The supported conclusion of this article is methodological: the treatment of prisoners’ relatives requires independent observation and corroboration, rather than inference from the prisoner’s status. This is not a verified case catalogue. No named cases, national totals or allegations against identified officials are asserted. A case-evidence account would require a separate, source-verified annex; none is asserted here.
Primary record
The documented public occasion for this method is the hearing “The Faces of Repression: Prisoners, Families, and Communities Targeted by the CCP,” held by the U.S. Congressional-Executive Commission on China on 22 September 2026: cecc.gov hearing record.
Editions in preparation in Simplified Chinese, Traditional Chinese, French, German and Spanish will be linked here upon publication.