The CCP’s Overseas Footprint, by Region and Timeline: What Is Established, What Is Only Correlated, and What the Latin America Brief Leaves Out

FINDINGS PAPER · FCM TERMINOLOGY — Enhanced edition · 2 October 2026

A comparative study prepared against the Human Rights Foundation report of 29 September 2026, “The CCP’s Deepening Footprint in Latin America and the Stakes for Democracy in the Region.”

Terminology. This paper follows the FCM guideline. Mainland China is the geographic and administrative area governed by the Chinese Communist Party (CCP) through the People’s Republic of China (PRC), founded in 1949. ROC (Taiwan) is the Republic of China — the constitutional government that has administered Taiwan, Penghu, Kinmen, Matsu, and other free-area territories continuously since 1949, and that predates the PRC. Cultural China — language, kinship, and a multi-millennial civilization — is not either state. Conflating the three is itself an influence technique, and this paper does not do it.

Status. Enhanced draft for review, 2 October 2026. Open-source synthesis, not a classified assessment. Figures that move (ally counts, loan stocks, port concessions) are dated. Where a claim is widely repeated but poorly sourced, it is marked contested. Recommendations in §10 are policy proposals addressed to free nations; they are argued from the findings, not derived from them, and each carries its costs and limits.

Cite as. The CCP’s Overseas Footprint, by Region and Timeline — findings paper, enhanced edition, 2 October 2026.

---

Executive summary

The CCP runs a party-state repertoire — strategic finance, state-owned-enterprise contracts, United Front organization, training pipelines, media and narrative work, security-technology transfer, and a diplomatic price for recognizing ROC (Taiwan) — that is global and repetitive. Regions differ in mix, not in menu.

Five headline findings:

1. The actor is the party, not “China.” The United Front Work Department, the International Liaison Department, the propaganda system, policy banks, and co-opted diaspora bodies do distinct work. Collapsing them into one actor produces both false alarms and missed operations.
2. The timeline breaks at 2013 and 2016–19, not “after COVID.” Belt and Road (2013) scaled the repertoire; the 2016–19 recognition-switch cluster was its sharpest political yield. The post-COVID period restored elite mobility; it did not start the strategy.
3. Recognition switches are the hardest evidence. Ten formal allies have moved from ROC (Taiwan) to the PRC since 2016, on dated, public, attributable acts. Project scandals are case studies; poll shifts are weak evidence.
4. The threat differs by regime type. Developing states face contract opacity and recognition bargaining. Developed democracies face interference, technology transfer, and transnational repression — not debt.
5. Capacity is not conversion. Counts of 2,000-plus united-front-linked bodies in four democracies measure latent mobilization, not operations. Narrative success is the least-measured outcome in the entire field.

Five headline recommendations for free nations:

1. Defend ROC (Taiwan)’s diplomatic space with pre-negotiated development compacts for the twelve remaining allies and a standing coalition contesting the PRC’s expansive reading of UN Resolution 2758.
2. Mandate transparency for sovereign loan collateral and infrastructure concessions — publish the terms, require parliamentary approval for collateralized borrowing.
3. Require foreign-influence disclosure on the FARA / Australia-FITSA / UK-FIRS model, aimed at party organs and their cutouts, not at diaspora communities.
4. Screen critical infrastructure — ports, cables, space ground stations, telecoms — for control and extraterritorial-lease risk before concessions are signed, not after.
5. Publish an annual resilience scorecard that counts reversals of bad deals as wins, so defensive success becomes visible and repeatable.

---

Contents

1. Why this paper, and how it improves on the example
2. Method, limits, and how claims are graded
3. Timeline: six phases, one repertoire
4. Instruments, not slogans
5. Regional findings
6. ROC (Taiwan) diplomatic isolation as the cross-check
7. Numbered findings
8. What this paper will not say
9. Strategic conclusions
10. Policy proposals: countermeasures for free nations
11. Implications for a next investigative pass
12. What would prove this paper wrong
13. Sources and references

Appendix A. Glossary of party-state instruments

---

1. Why this paper, and how it improves on the example

The HRF paper, led by Sascha Hannig and published 29 September 2026, is a useful regional brief. It names five channels in Latin America: strategic financing; diplomatic and political consolidation; security rhetoric and assistance; narrative work paired with what it calls hybrid-warfare capabilities; and isolation of ROC (Taiwan) diplomatic channels. It draws on open sources plus 14 interviews in 12 countries [1]. It is also honest that causation between CCP activity and democratic decline is contested.

It is not yet an investigative findings paper. It is one region, a short look-back (five and ten years), and it treats “China” as a single actor. It under-specifies the party instruments that do the work — United Front Work Department, International Liaison Department, propaganda outlets, state-owned contractors, and “peaceful reunification” councils — and it does not test whether the same pattern appears, on the same dates, in Africa, the Pacific, Europe, and North America. A better paper does four things the example does not.

• Separate the actor. Party, PRC state, state-owned firms, co-opted diaspora bodies, and local elites are not one thing. A dam built by Sinohydro is not the same evidence as a United Front council issuing a statement against the ROC (Taiwan) president.
• Date the sequence. Influence is a timeline, not a mood. The useful breaks are 1971, 2000, 2013, 2015–17, 2018–19, 2020–22, and 2023–26.
• Use ROC (Taiwan) recognition as a cross-regional indicator. Formal switches are public, dated, and harder to spin than “discourse power.” As of 2 May 2026, Reuters still listed 12 states with formal ties to ROC (Taiwan) [2].
• Grade the claim. Established, correlated, or contested. The example’s phrase “enough evidence to state that China has intensified its efforts” is true and also too soft to brief from.

---

2. Method, limits, and how claims are graded

This paper uses process tracing on public acts: recognition switches, leader visits, port and space-station concessions, party-school and scholarship programs, United Front organizational expansion, and documented cases of transnational repression. It does not treat trade volume as influence. Mainland China is a top goods partner for dozens of democracies; that is commerce until a political condition is attached, a concession is opaque, or a local institution is repurposed.

Evidence grades

Every substantive claim in §§3–7 carries one of three grades:

• Established — multiple independent public sources converge; the act is dated and attributable (e.g., a recognition switch, a signed concession, a prosecuted agent).
• Correlated — the pattern co-occurs with CCP activity, but causation is not shown and local causes are plausible (e.g., elite-capture anecdotes, democratic-decline timing).
• Contested — the claim is widely repeated but poorly sourced, or the category itself is disputed (e.g., flattened counts of “overseas police stations”).

Limits

Loan and aid figures reported by recipient politicians are often opening bids, not contracts. “Overseas police stations” became a public category after Safeguard Defenders’ 2022 reports [11]; some sites were service centers, some were used for pressure, and host-government responses varied — the category should not be flattened. Dual-use space and cable sites are risk findings, not proof of espionage. No interview set is claimed here; a next draft should add structured interviews in the remaining ROC (Taiwan) ally states and in two switched states per region.

---

3. Timeline: six phases, one repertoire

The repertoire — elite financing, infrastructure contracts, diaspora organization, media, training, and a diplomatic price for recognizing ROC (Taiwan) — is older than the Belt and Road. What changed after 2013 is scale, central coordination, and the willingness to spend political capital in democracies, not only in already-aligned governments.

Finding T1 (established). The acceleration outsiders date to “after COVID” is real for travel and elite access, but the political breaks are 2013 and 2016–19. HRF’s COVID pivot is a recovery of mobility, not the start of the strategy.

---

4. Instruments, not slogans

Jamestown’s February 2026 mapping (Cheryl Yu) counted more than 2,000 united-front-linked organizations in the United States, Canada, the United Kingdom, and Germany alone [6]. The Australian Strategic Policy Institute and others have tracked more than 200 “peaceful reunification” bodies worldwide [7]. Those are capacity counts, not proof that each body is active.

---

5. Regional findings

5.1 Latin America and the Caribbean

This is the HRF region, and the public record supports the brief’s direction. It does not support treating the region as the leading edge. Formal switches since 2017 — Panama (2017), Dominican Republic and El Salvador (2018), Nicaragua (2021), Honduras (2023) — are the cleanest evidence. Remaining formal allies of ROC (Taiwan) in the region, as listed by Reuters on 2 May 2026 [2]: Belize, Guatemala, Paraguay, Haiti, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines.

Project evidence is uneven. Ecuador’s Coca Codo Sinclair dam (Sinohydro) is a documented case of defects, cost, and renegotiation, not a template for every loan. Peru’s Chancay port, inaugurated by Xi in November 2024 and linked to COSCO and a large loan package, is a logistics foothold on the Pacific, with local environmental and labor disputes that are real and separate from “hybrid war.” Argentina’s Neuquén deep-space station, agreed under a previous government, remains a secrecy and dual-use controversy. Venezuela and Nicaragua show the security channel more clearly than Chile or Uruguay do: surveillance and political alignment in already authoritarian systems. Colombia’s BRI cooperation document and scholarship offers, and Bolivia’s proposed lithium investment, are intents until money and offtake are public.

Finding L1 (established). Isolation of ROC (Taiwan) in this region is a 2017–2023 campaign with dated switches, not a diffuse “narrative.” Finding L2 (correlated). Elite capture stories (Chile’s congressional controversy; scholarship pipelines) are leads, not yet a regional causal model. Finding L3 (contested). HRF’s claim that CCP activity has “changed the political landscape” of established democracies overstates what its own interviews would certify.

5.2 Africa

Africa is the older theatre. Most switches happened before BRI. Burkina Faso (May 2018) left ROC (Taiwan); Eswatini is the last African state with formal ties, and its king restated that position at the UN General Assembly in September 2026 [3]. The Forum on China-Africa Cooperation, resource-backed finance, party-to-party training, and media carriage (StarTimes and content-exchange deals) are the standing instruments. “Peaceful reunification” councils in Niger, Cameroon, Tanzania, Chad, Angola, and others issued coordinated statements against President Lai Ching-te’s 20 May 2026 speech — a useful example of a global template, not of African public opinion.

Finding A1 (established). Party-to-party training and media carriage are thicker in Africa than in Latin America, and older. Finding A2 (correlated). Resource finance creates renegotiation leverage; it does not, by itself, explain democratic backsliding, which has local causes. Finding A3. Eswatini is a single-point diplomatic risk, not a regional trend.

5.3 Pacific Islands

Highest political yield per dollar. Solomon Islands and Kiribati switched in September 2019. Nauru switched on 15 January 2024, within 48 hours of the ROC (Taiwan) presidential election; Taipei called the timing deliberate [4]. A 2022 PRC–Solomon Islands security agreement, whose leaked draft allowed armed police deployments, is the clearest security-access document in the region. Remaining formal allies: Marshall Islands, Palau, Tuvalu.

Finding P1 (established). Recognition switches here cluster around ROC (Taiwan) electoral moments and are publicly tied to aid comparisons. Finding P2 (established, narrow). The Solomon Islands security pact is a status-of-forces risk that Latin America does not currently match.

5.4 Southeast Asia and the PRC periphery

Different from the “influence in democracies” frame. Mainland China is a permanent neighbor: dams on the Mekong, South China Sea coercion, and United Front priority against ROC (Taiwan) and, after 2020, Hong Kong exiles. Cambodia and Laos are alignment cases. Singapore, the Philippines, and Vietnam show the split the global story needs: deep economic ties alongside security hedging. The Philippines’ public breaks with PRC maritime pressure after 2023 are evidence that exposure does not equal capture.

5.5 Europe, North America, and other developed democracies

Here the channel is not recognition — none of these states has formal ties with ROC (Taiwan) as a China seat — but interference inside open societies. Recorded patterns: United Front organizational density; Confucius Institute and campus pressure, followed by closures in the United States and several European states; transnational repression cases against Hong Kong, Uyghur, Tibetan, and Falun Gong targets; elite and academic talent programs; and local-government “sister” diplomacy. Canada, Australia, the United Kingdom, and New Zealand have each held inquiries or expelled officers over interference allegations. The United States has prosecuted undisclosed agents and restricted military-linked technology.

Finding D1 (established). The developed-democracy problem is interference and technology transfer, not debt. Comparing it to Coca Codo Sinclair obscures both. Finding D2 (correlated). Narrative success is the weakest measured outcome. Capacity is not conversion.

5.6 Central Asia, South Asia, and the Middle East

BRI’s original overland map. Pakistan’s corridor, Central Asian gas and rail, and Gulf port and telecom deals are commercial-strategic. The political price is usually silence on Xinjiang and, secondarily, on ROC (Taiwan), not a recognition switch — these states recognized the PRC decades ago. Iran and Russia are partners of convenience with their own agendas; treating them as CCP instruments reverses the arrow.

---

6. ROC (Taiwan) diplomatic isolation as the cross-check

Formal allies of ROC (Taiwan) as of the Reuters list of 2 May 2026, still the operative public count into this drafting date [2]: Guatemala, Belize, Paraguay, Haiti, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Marshall Islands, Palau, Tuvalu, Eswatini, and the Holy See. Twelve. In September 2026 Paraguay, Eswatini, and Palau used the UN General Debate to contest the misreading of Resolution 2758 that keeps ROC (Taiwan) out of UN bodies [3].

Finding R1 (established). Since 2016 the PRC has taken ten formal allies from ROC (Taiwan), on Taipei’s own accounting around the Nauru switch [4][5]. The method is competitive inducement plus a one-China condition, aimed at small states. Finding R2. Losing a formal ally does not transfer administrative control. ROC (Taiwan) still governs the free area; the loss is a UN-system and symbolic loss. Finding R3. A permit-and-license policy for mainland fishing vessels in ROC (Taiwan) waters, discussed after the 1 October 2026 Dongsha expulsion, is a domestic-administration question. It is not a substitute for this diplomatic contest, and it should not be traded against recognition [10].

---

7. Numbered findings

• F1. The CCP runs a party-state repertoire that is global and repetitive: finance and SOE contracts, United Front bodies, training, media, security kits, and a diplomatic price for ROC (Taiwan). Regions differ in mix, not in menu.
• F2. The correct periodization is 1971, 2000, 2013, 2015–17, 2018–19, 2020–22, 2023–26. “After COVID” describes restored travel, not origin.
• F3. Formal recognition switches are the best cross-regional dependent variable. Project scandals are case studies. Poll shifts are weak evidence.
• F4. Democratic decline and CCP presence co-occur in several states and are not shown to be the same process. Venezuela’s trajectory is domestic authoritarian consolidation with an external patron. Chile’s institutions have so far reversed at least one sensitive authorization (a transpacific cable).
• F5. Developed democracies face interference and technology risk. Developing states face contract opacity and recognition bargaining. Briefing them as one “footprint” hides the policy response.
• F6. Cultural China is not the PRC. Papers and posts that slide from civilization to the 1949 state do the party’s substitution for it. ROC (Taiwan) is the prior Chinese state still governing free territory.
• F7. Capacity numbers (2,000-plus united-front-linked bodies in four democracies; 200-plus reunification councils) measure latent mobilization. They are not a count of operations.

---

8. What this paper will not say

It will not say that every mainland Chinese firm is a party cutout, that BRI debt explains global autocratization, that Latin America is lost, or that ROC (Taiwan)’s twelve allies are about to switch. It will not treat Resolution 2758 as a settlement of sovereignty; the resolution seated a delegation, and the PRC’s later reading of it as a decision on Taiwan’s status is a political claim, disputed by ROC (Taiwan) and by several of its allies on the record in September 2026 [3]. It will not use casualty slogans in place of instrument tracing.

---

9. Strategic conclusions

The findings above support six conclusions about the character of the contest — what kind of problem free nations are actually facing.

C1. The contest is over recognition and institutional capture, not trade. Mainland China will remain a major trading partner for most democracies; that commerce is not the threat. The threat begins where a political condition attaches: a port concession with opaque terms, a loan collateralized against sovereign assets, a media deal that buys editorial silence, a diplomatic switch purchased with an aid package. Policy that targets trade as such will fail on both economics and evidence; policy that targets the conditions can succeed.

C2. The party is the correct unit of analysis — and of response. Every instrument in §4 traces to party organs (UFWD, ILD, propaganda system) or to state firms executing party direction. Treating the PRC state, its firms, its diaspora, and its civilization as one undifferentiated “China threat” produces two errors at once: it alarms innocent communities and misses the actual operators. Countermeasures should name party organs the way sanctions regimes name specific entities — precisely, or not at all.

C3. Small states are the decisive terrain. Ten switches since 2016, aimed at states with small budgets and large infrastructure needs. The PRC’s method — competitive inducement plus a one-China condition — works because no free-nation alternative is pre-positioned when the offer arrives. By the time a switch is announced, the contest is over. The policy implication is temporal: alternatives must exist before the bid, not as commentary after it.

C4. Developed democracies and developing states face different problems and need different answers. Interference, technology transfer, and transnational repression inside open societies (D1) cannot be answered with debt-transparency tools; contract opacity and recognition bargaining (L1, A1, P1) cannot be answered with campus policies. A single “counter-influence strategy” that briefs both as one footprint will misallocate effort in both.

C5. The record contains defensive wins, and they are the most under-used evidence. Chile’s reversal of a sensitive cable authorization, Confucius Institute closures across the US and Europe, expulsions of interference-linked officers, Eswatini’s restated position at UNGA 2026, the Dongsha enforcement of 1 October 2026 — these are demonstrations that exposure plus institutional response works. A strategy built only on threat inflation, without a visible theory of successful defense, will not sustain democratic consent for its costs.

C6. ROC (Taiwan) is the load-bearing ally of the free world’s position. Twelve formal allies, a standing UNGA coalition contesting the 2758 misreading, and the demonstrated willingness of small states to hold the line (Paraguay, Eswatini, Palau, September 2026 [3]) make Taipei the proof that the PRC’s diplomatic campaign is resistible. Every switch the PRC buys is priced against the expectation that the next holdout will also fold; every holdout that stands raises the price of the next bid. Supporting ROC (Taiwan)’s diplomatic space is therefore not charity — it is the cheapest leverage in the system.

---

10. Policy proposals: countermeasures for free nations

Ten proposals, ordered from most time-sensitive to most structural. Each is argued from the findings above. Each names its price — a proposal that won’t admit what it costs isn’t a proposal, it’s a wish.

P1. Pre-position development alternatives for ROC (Taiwan)’s remaining allies

The next recognition bid will not be announced. It will arrive as a private offer to a finance minister, and by the time the switch is reported, the contest is over (R1, C3). The only move that matters is made before the bid: a pre-negotiated, publishable development compact — infrastructure, health, education — sized to what the PRC has historically tabled in that region, held ready by free-nation development-finance institutions for each of the twelve remaining allies. When the offer comes, the head of government should face a genuine alternative, not a lecture. And the calendar matters: keep the compacts ready around ROC (Taiwan) electoral moments, because that is when the switches cluster (P1).

The price: small-state leaders can auction the contest upward, and an unconditional compact subsidizes the same opacity it is meant to counter — so tie every compact to the transparency rules in P2. Donor publics will ask why these twelve states and not others. Say it plainly and keep saying it: every holdout raises the price of the next PRC bid (C6). That is not charity. It is the cheapest leverage in the system.

P2. Make sovereign finance legible

The strategic-finance instrument runs on opacity: collateral terms nobody sees, concessions nobody votes on, renegotiations announced after the fact (F1, A2). Three legislative rules would change the game. First, publish the collateral terms of any sovereign loan secured against public assets or resource revenues before disbursement. Second, require parliamentary approval — on the full contract, not a summary — for infrastructure concessions above a threshold value. Third, publish every renegotiation and collateral seizure within ninety days. Anchor all of it in the existing G20 and IMF–World Bank debt-transparency initiatives [17], so this reads as multilateral standard-setting rather than unilateral suspicion.

The honest limit: governments in genuine fiscal distress will resist conditionality on emergency finance, and they have a point. The rules must distinguish emergency liquidity from long-term collateralized lending, or the first crisis will route around them. And publication by itself fixes nothing — it makes bad deals contestable, which is where parliaments, journalists, and courts do the actual work.

P3. Disclosure regimes that name the party — not the diaspora

Covert party-directed activity should be legible; diaspora communities should not be policed. Both goals are served by the same instrument, if it is drafted with precision: an agent-of-a-foreign-principal registration regime on the established models — the US Foreign Agents Registration Act [12], Australia’s Foreign Influence Transparency Scheme [13], the UK’s Foreign Influence Registration Scheme [14]. The registrable trigger must be direction or funding by a foreign party-state organ, with published guidance naming those organs (Appendix A), so that compliance is knowable in advance and the net lands on the UFWD, the ILD, and their documented cutouts — never on students, artists, or hometown associations with no party direction (F6, P8).

The price: disclosure regimes burden legitimate exchange, and a sloppily drafted one chills exactly the academic and cultural contact free societies should want. The instrument list must be party-specific and reviewed on a schedule. And be honest about enforcement: sophisticated cutouts are slow to prosecute. The register’s real value is deterrence and investigative journalism, not conviction counts.

P4. Screen the concession, not just the acquisition

Investment screening was built for equity purchases. The PRC’s infrastructure plays are often concessions and long leases — Chancay’s berth control, Neuquén’s ground station, the Solomon Islands security annex (F1, L1, P2). Screen those too: extend CFIUS-style review [16], the EU screening framework [15], and allied equivalents to concessions, leases, and operating agreements in ports, cable landings, space ground stations, and telecoms. Review extraterritorial-lease clauses, berth-control rights, and data-access terms before signature. Publish the criteria so allied investors face the same bar.

The catch: screening is only as good as the technical capacity behind it, and most small states don’t have it — so share allied review capacity, or the rule is decorative. And keep the criteria narrow: control and data-access risk, not the nationality of the company pouring the concrete. Over-broad screening deters the legitimate investment these countries need.

P5. A dedicated posture against transnational repression

Diaspora activists — Hong Kongers, Uyghurs, Tibetans, Falun Gong practitioners — face the sharpest end of the interference problem (D1, §5.5), and general-purpose policing underserves them. Four moves. Dedicated reporting channels with published case statistics, so the phenomenon is counted instead of rumored. Expulsion of officers credibly linked to repression operations, as several democracies have already done. Targeted, evidence-based sanctions on individuals directing repression campaigns, where the evidence supports designation. And funded legal-defense and digital-security support for the targeted, prioritized by documented threat volume.

The price: attribution is often intelligence-held and unpublishable, which strains public credibility — so build the public case on the fully documented instances and say which is which. Sanctions on individuals, uncoordinated, are theater; they only bite when jurisdictions move together.

P6. Sunlight for the information environment — not censorship of it

The narrative instrument runs on hidden subsidy: syndication deals, training trips, paid inserts whose funding the reader never sees (F1, §4). Answer with sunlight, not censorship. Beneficial-ownership disclosure for media outlets above audience thresholds. Funding disclosure for content-sharing deals with foreign state media. Mandatory labeling of paid inserts from foreign state actors. And play offense where the propaganda system is weakest: fund independent Chinese-language journalism, because the audience Beijing most wants to capture is the diaspora audience independents already serve.

The trade-off, stated plainly per D2: audience conversion is the least-measured outcome in this field, so score this proposal on exposure of subsidy, not on claimed shifts in public opinion. And the ownership rules must apply to all foreign state media equally. A rule that reads as discriminatory loses the legitimacy it needs to work.

P7. Maritime enforcement by statute, not by signal

Gray-zone maritime pressure thrives on ambiguity — was that entry authorized, and by whom? The Dongsha expulsion of 1 October 2026 [10] points the way: put licensing on a statutory footing. Any permission for mainland fishing vessels in free-area waters must rest on public law — published quotas, AIS-reporting mandates, defined zones — never on political understanding. Unauthorized entry stays an enforcement event: documented, published, unexceptional (R3). Free nations with coast-guard capacity should deepen joint patrols and information-sharing, and publish interdiction statistics so the pattern is visible.

The price: enforcement against a vastly larger maritime militia can spiral. Rules of engagement and de-escalation channels must be set before the encounter, not improvised at the waterline. And small coast guards need allied material support, or this is posture rather than policy.

P8. Name the party. Spare the civilization.

This one costs nothing and constrains everything, which is why institutions will resist it. Write every designation, sanction, disclosure rule, and screening criterion against party organs and military-linked entities (Appendix A) — never against “Chinese” persons, firms, or culture as such. Say the distinction affirmatively, in legislation and in public, in this paper’s terminology: the party and the 1949 state are the subject; Cultural China is not (F6, C2). Then audit existing measures for over-broad language and narrow them.

The price is discipline. Precision demands better intelligence: naming the organ requires knowing the organ — its funding lines, its cutouts, its personnel — which is exactly the investigative investment §11 calls for. Vagueness is cheaper. It is also how you lose the communities whose cooperation you need.

P9. Count the wins: an annual resilience scorecard

Threat assessments accumulate; defensive victories evaporate (C5). An independent allied or multilateral body should publish a yearly, evidence-graded scorecard: recognition switches and holds; concession reversals — Chile’s cable decision is the template positive case; interference prosecutions and expulsions; disclosure-register filings; deals reviewed and rejected. Grade every entry the way this paper does: established, correlated, contested.

The institutional resistance: governments dislike publishing their near-misses, so the publisher must be independent with a protected methodology. And the scoring must resist becoming a beauty contest — weight holds and reversals, not press releases.

P10. A standing allied mechanism with three products — or nothing

Twelve capitals learning the same lesson separately is a strategy for losing slowly (C3, F1). A standing working group of foreign ministries, development-finance institutions, and intelligence-sharing partners should own exactly three standing products: the switch dataset (§11), the diplomatic calendar for the 2758 coalition (building on the Paraguay–Eswatini–Palau precedent of September 2026 [3]), and a shared watchlist of party-organ cutouts to feed disclosure regimes (P3). Fixed cadence. Publish what can be published.

The price, and the condition: coordination bodies decay into talking shops. Charter this one around the three products, sunset any strand that stops producing, and keep membership to states willing to publish. If it can’t hold to that, don’t create it.

11. Implications for a next investigative pass

• Build a switch dataset: ask, inducement, contract within 18 months, performance at 36 months. Honduras and Nauru are the freshest tests.
• Map reunification-council statements after a fixed set of ROC (Taiwan) speeches. Coordination is the finding; local authorship is the test.
• Separate SOE contract performance (defects, arbitration, environmental review) from security annexes. Publish the annex where it exists, as with the Solomon Islands draft.
• In remaining ally states, interview finance and foreign ministries on what was offered and refused. HRF’s 14 interviews did not concentrate there.
• For ROC (Taiwan) maritime administration: any license for mainland boats in free-area waters should be statute, quota, and AIS reporting, not a political concession. Unauthorized entry stays an enforcement event, as at Dongsha on 1 October 2026.
• Grade host-country resilience: reversal of a cable or observatory deal is a positive indicator and should be scored, not only the approach.

---

12. What would prove this paper wrong

A findings paper that cannot say what would falsify it is a sermon. Three tests:

1. The switch test. If the next three recognition switches occur without any inducement package — no aid, no project, no debt relief within eighteen months on either side — then the competitive-inducement model (R1) is wrong, and something else is driving the diplomacy.
2. The capacity test. If a rigorous, independent audit of united-front-linked bodies in two democracies finds that more than half engage in directed political activity (not merely cultural association), then F7’s “latent mobilization” framing understates the problem, and this paper’s caution was misplaced.
3. The conversion test. If a well-designed study demonstrates that PRC narrative operations did measurably shift an electoral outcome or a policy decision in a developed democracy — with the causal chain documented, not asserted — then D2’s skepticism about narrative success needs revision.

The paper’s friends should run these tests. Its critics are invited to.

---

13. Sources and references

[1] Human Rights Foundation, “The CCP’s Deepening Footprint in Latin America and the Stakes for Democracy in the Region,” led by Sascha Hannig, 29 September 2026. [2] Reuters, “Taiwan’s 12 remaining diplomatic allies,” 2 May 2026. [3] Focus Taiwan, 24 September 2026, on Paraguay, Eswatini, and Palau contesting the 2758 misreading at the 81st UN General Assembly. [4] The Guardian, 15 January 2024, on Nauru’s diplomatic switch. [5] Will Piekos, compilation of post-2016 diplomatic switches (inducement figures are reported asks, not audited contracts). [6] Cheryl Yu, “Harnessing the People,” Jamestown Foundation, 11 February 2026 (mapping of 2,000-plus united-front-linked organizations in the US, Canada, UK, and Germany). [7] The Strategist (Australian Strategic Policy Institute), 16 June 2026, on “peaceful reunification” councils worldwide. [8] Diálogo Américas, 18 November 2025, on United Front nodes in Latin America. [9] Radio Free Asia, explainer on the United Front Work Department, 3 November 2024. [10] Republic of China (Taiwan) Coast Guard Administration, public account of the Dongsha expulsion, 1–2 October 2026, via Taipei Times, Focus Taiwan, and Taiwan News. [11] Safeguard Defenders, “110 Overseas” reporting, 2022, on overseas police service centers. [12] United States, Foreign Agents Registration Act of 1938 (22 U.S.C. §§ 611–621). [13] Australia, Foreign Influence Transparency Scheme Act 2018. [14] United Kingdom, National Security Act 2023, Part 4 (Foreign Influence Registration Scheme). [15] European Union, Regulation (EU) 2019/452 establishing a framework for the screening of foreign direct investments. [16] United States, Foreign Investment Risk Review Modernization Act of 2018 (FIRRMA). [17] G20 Operational Guidelines for Sustainable Financing; IMF–World Bank debt-transparency initiatives; Blue Dot Network infrastructure-certification initiative.

This paper does not adopt HRF’s contested causal sentences [1]. It keeps the five Latin American channels, dates them, and sets them beside the same channels elsewhere.

Appendix A. Glossary of party-state instruments

For policymakers and drafters: the organs named in this paper, in one place.

• United Front Work Department (UFWD) — CCP Central Committee department tasked with co-opting non-party actors at home and abroad: diaspora bodies, business figures, academics, religious and community organizations. Upgraded to a central leading group in 2015; called a “magic weapon” by Xi in 2017.
• International Liaison Department (ILD) — the CCP’s party-to-party diplomacy arm; runs training programs for foreign political parties, notably dense in Africa.
• China Council for the Promotion of Peaceful National Reunification (CCPPNR) — founded in Beijing in 1988; the template for 200-plus overseas “peaceful reunification” councils that issue coordinated statements on ROC (Taiwan).
• Propaganda system / China News Service — state media and content-syndication networks; training trips and paid inserts in local outlets.
• Policy banks and SOE contractors — China Development Bank, Exim Bank; firms such as Sinohydro and COSCO Shipping Ports; the commercial carriers of strategic finance.
• Security-technology exporters — firms such as Hikvision, ZTE, and CEIEC supplying surveillance and telecoms systems to interior ministries.

Designations and disclosure obligations proposed in §10 attach to these organs and their documented cutouts — not to mainland-Chinese persons, firms, or culture as such (§P8).

---

End of enhanced draft · 2 October 2026