In April 2016, Cambodian authorities arrested five staff members of the Cambodian Human Rights and Development Association on fabricated bribery charges. The ADHOC 5 spent 427 days in pre-trial detention before being released on bail, and were eventually given suspended five-year sentences in September 2018 for what Amnesty International called “a political outcome to a political case.” Ten years later, the international system that condemned those arrests has systematically withdrawn the funding that sustained the organizations working against such abuses. The mechanisms are different. The direction is the same.
The standard account of Cambodia’s civic space contraction treats it as a story about an authoritarian government suppressing dissent. The Cambodian government’s role is documented and significant: the 2015 Law on Associations and NGOs imposed mandatory registration requirements and granted the Ministry of Interior broad discretionary power over civil society organizations. The dissolution of the opposition Cambodia National Rescue Party in November 2017, the expulsion of the National Democratic Institute in August of that year, and the closure of the Cambodia Daily the following month established a pattern of state-driven contraction that is not in dispute.
What has received less attention is what happened next. The donor withdrawals that followed did not reverse the contraction. They compounded it. The sequencing, when mapped against the contraction events, reveals a pattern that no individual withdrawal decision makes visible on its own.
The Sequencing
The major civic space contraction events clustered in 2016 and 2017. The ADHOC 5 arrests came in April 2016. The NDI expulsion, Cambodia Daily closure, and CNRP dissolution all fell between August and November 2017. Within days of the CNRP dissolution, the United States cut election funding to Cambodia and signaled further reductions.
The institutional donor responses came later and accumulated over time. The European Union decided in February 2020 to partially withdraw Cambodia’s Everything But Arms trade preferences, effective August 2020, suspending duty-free access for approximately 20 percent of Cambodian exports. The EU cited “serious and systematic violations” of International Covenant on Civil and Political Rights provisions on expression, assembly, association, and political participation.
Sweden, historically one of the largest bilateral donors to Cambodian civil society, began restructuring in June 2020 when it decided to refocus bilateral cooperation. In December 2023, the Swedish government announced the complete phase-out of bilateral development cooperation with Cambodia, effective December 31, 2024. Approximately $17 million in annual funding to more than 30 civil society organizations ended. The Swedish Ministry for Foreign Affairs cited the war in Ukraine and the need for strategic reallocation, though observers noted that the decision came after years of criticism of Cambodia’s democratic trajectory. Sweden’s section office in Phnom Penh closed on September 30, 2024.
The most recent withdrawal was the largest. On January 20, 2025, the Trump administration froze most US foreign aid globally. By May 2025, 30 USAID contracts in Cambodia had been permanently cancelled as part of an 83 percent cut to USAID programs worldwide. Affected projects ranged from child literacy and early childhood nutrition to tuberculosis treatment and demining operations. An analysis by the Business and Human Rights Law Group found that over $371 million in US aid had been distributed across 185 projects in Cambodia between January 2020 and January 2025. CamboJA reported in March 2026 that a year after the cuts, NGOs and unions had shut down programs, reduced staff, and were seeking alternative donors.
The chronology establishes a clear sequence: the core contraction events preceded the major donor withdrawals. Each subsequent withdrawal then reduced the organizational capacity available to resist or document the next phase of contraction.
The Academic Evidence
A study published in International Organization in Fall 2025 provides empirical grounding for this pattern. Lucy Right, Jeremy Springman, and Erik Wibbels found that after governments pass restrictive NGO legislation, the donors most committed to democracy promotion reduce their democracy aid by approximately 70 percent, rather than increasing support in response. The study specifically names Sweden’s withdrawal from Cambodia as an illustrative case, and notes that SIDA had previously identified “promoting an enabling environment for civil society” in response to restrictive legislation as one of its two highest priorities for 2016 to 2022.
The finding inverts the expected logic. Democracy-oriented donors do not push back against restrictive civic space laws by defending the organizations those laws target. They back down. The effect concentrates losses on NGOs and bypass channels rather than on recipient governments. The mechanism is structural: when restrictive laws make it difficult to work through local civil society partners, donors conclude that their programming is no longer achieving its objectives, and withdraw.
For Cambodia, this means the passage of LANGO in 2015 and the broader crackdown of 2016 to 2017 did not just directly suppress civil society. They triggered a secondary withdrawal mechanism that operated through the international donor system itself. Each donor departure then reduced the organizational infrastructure available for the next, creating the conditions under which further withdrawal appeared rational.
The Replacement Dynamic
The vacuum created by these withdrawals has not remained empty. In late February 2025, the United States cancelled two aid projects in Cambodia focused on child literacy and early childhood nutrition. Within a week, China’s aid agency announced funding for programs with near-identical objectives. Fortune magazine reported that at a congressional hearing, a Trump administration official was directly asked about the Cambodia case and the timing of China’s announcement, and dismissed concerns about Chinese influence.
This is not a hypothetical geopolitical consequence. It is a documented one. The competitive dynamic between US and Chinese development engagement in Cambodia has operated for years, but the scale and speed of the 2025 replacement represented a new threshold.
What Cambodia Has Done
Cambodia’s government has not been entirely silent on governance reform. In its fourth-cycle Universal Periodic Review submission in May 2024, the Cambodian Human Rights Committee reported a second phase of consultations on amendments to 16 articles of LANGO, an extension of the memorandum of understanding with the Office of the UN High Commissioner for Human Rights through December 2024, and continued bilateral human rights dialogues with Japan and the United Kingdom. The submission reported 6,279 registered local associations and NGOs as of December 2023.
These are official claims, not independent assessments of implementation. The UN Special Rapporteur on human rights in Cambodia described the post-2017 environment as dominated by the established elite, with continued detention and prosecution of political actors. The ASEAN Intergovernmental Commission on Human Rights, structurally constrained by a mandate that places “primary responsibility” for rights protection on member states, operates by consensus and appointment, and has produced no Cambodia-specific protective intervention on civic space.
The gap between Cambodia’s official reform language and the documented conditions is real. It is also not the only gap in the system. The gap between international concern and sustained institutional support is equally documented.
The Mechanism Nobody Maps
In the existing academic and policy literature, no study integrates the full sequence for Cambodia. The building blocks are distributed across separate research streams: aid dependency and civil society formation, government repression and NGO adaptation, donor responses to restrictive legislation, aid conditionality theory, and Cambodia’s own institutional record. Each stream documents a component. The compound interaction between them, in which state action triggers donor withdrawal, donor withdrawal reduces civil society capacity, and reduced capacity is then cited as evidence that further engagement is not achieving results, has not been formalized as a single Cambodia-specific model.
This matters because the feedback loop creates policy incoherence. The same international system that conditions trade preferences on civic space metrics simultaneously withdraws the funding that sustains civic space organizations. The EU’s partial EBA withdrawal cited restrictions on assembly and association. The donor withdrawals that followed reduced the organizational capacity of the very groups exercising those rights. These are not contradictory policies adopted by the same actor. They are structurally connected decisions by multiple actors within the same international system, each internally rational, producing a combined outcome that none individually intended.
The Thirty-Two Year Echo
In 1994, Michael Kirby, the first UN Special Representative for Human Rights in Cambodia, wrote in his report “Cambodia: Unequal Suffering, Unique Opportunity” that Cambodia “deserves more support from the international community than mere words.” He recommended the integration of NGOs into legislative consultation, the establishment of an independent national human rights institution, and guarantees for the security of civil society. Three decades later, LANGO governs NGO registration, no independent NHRI compliant with the Paris Principles has been established, and the donor infrastructure that funded the organizations Kirby sought to protect has largely withdrawn.
Kate Flower, writing in The Diplomat on the tenth anniversary of the ADHOC 5 arrests, posed the question directly: “If that system and the international community are no longer dependable, who will stand up for human rights?” The evidence assembled here suggests a prior question. The system was not designed to protect what it depends on. Its architecture makes the feedback loop not an anomaly, but an output.
