Executive summary

Mozambique’s recurring civil disorder since independence in 1975 unfolds in three interlinked arcs: (1) a Cold War–era civil war (1977–1992) driven by external sponsorship and domestic governance missteps; (2) a post-2013 relapse of FRELIMO–RENAMO tensions resolved with the 2019 Maputo Accord; and (3) a distinct jihadist-linked insurgency in Cabo Delgado (since 2017) fueled by local grievance, predatory governance around extractives, and transnational militant networks. These arcs are connected by persistent state–society distrust, uneven regional development, and the political economy of rubies and offshore gas, all compounded by the 2016–2017 “hidden debt” scandal that eroded public finances and legitimacy. Stabilization will require deeper political inclusion, accountable resource governance, and sustained—but appropriately conditioned—regional security support. 

1) Historical foundations: from liberation to polarized state-building

Mozambique’s anti-colonial struggle (1964–1974) culminated in FRELIMO’s victory and independence in 1975. The new state adopted a centralizing Marxist-Leninist project that expanded state farms, weakened customary authority, and pursued compulsory villagization to deliver services and security. While modernizing in intent, these policies alienated many rural communities and sharpened regional cleavages, especially where livelihoods and authority structures were upended. 

At the same time, Rhodesia’s intelligence service and, subsequently, apartheid South Africa cultivated and armed a counter-insurgent vehicle—RENAMO—both to punish FRELIMO for backing Zimbabwe’s liberation movement and to destabilize a hostile neighbor. The early role of Rhodesian CIO chief Ken Flower and the later handover to South African patrons are well attested in primary accounts and scholarship. 

Result: By the late 1970s, Mozambique exhibited the classic ingredients of a proxy war layered atop local grievances: a centralizing elite; marginalized peripheries; and neighbors intent on strategic disruption.

2) The first civil war (1977–1992): course and settlement

Course. RENAMO evolved from a covert Rhodesian asset into a nationwide insurgency under Afonso Dhlakama, striking infrastructure, coercing rural populations, and exploiting local discontent with villagization and heavy-handed security. FRELIMO relied heavily on Soviet-bloc support and later assistance from frontline states; Zimbabwe intervened directly to secure its corridors (notably Beira). The conflict devastated rural economies and social services. 

Settlement. With the Cold War ebbing and South African/Rhodesian sponsorship collapsing, Church-mediated talks in Italy produced the Rome General Peace Accords (Oct 4, 1992). The deal provided for a ceasefire, multiparty elections, demobilization/integration, and a UN mission (ONUMOZ). Implementation is widely judged successful by comparative standards. 

Legacy dynamics. The peace institutionalized RENAMO as an opposition party with regional strongholds; competitive but often contentious elections followed. Yet the war’s social geography—center vs. periphery, FRELIMO vs. opposition-leaning rural regions—remained salient. 

3) Post-2013 tensions and the 2019 Maputo Accord

Disputes over election credibility, decentralization, and security guarantees triggered RENAMO remobilization (2013–2018). Renewed negotiation yielded a two-step deal in August 2019: the Agreement on the Definitive Cessation of Military Hostilities (Aug 1, 2019) and the Maputo Accord for Peace and National Reconciliation (Aug 6, 2019), linked to a DDR program and constitutional changes on provincial governance. Implementation has been uneven but largely held, even as local protest movements against perceived electoral fraud flared again in 2024. 

4) A distinct conflict system: the Cabo Delgado insurgency (since 2017)

Origins. The northern province of Cabo Delgado—remote, Muslim-majority districts with chronic under-investment—saw the emergence of a militant movement locally called Al-Shabaab (unrelated to Somalia’s group), later aligned with Islamic State (IS-Mozambique/ASWJ). Drivers include youth exclusion, abusive policing, land/resource dispossession around ruby (Montepuez) and LNG mega-projects, and predatory local governance. The 2016–2017 “hidden debt” scandal compounded fiscal crisis and undermined state legitimacy as the insurgency coalesced. 

Escalation and externalities. Attacks peaked 2020–2021 as militants overran Mocímboa da Praia and threatened Palma, forcing TotalEnergies to suspend a $20 bn LNG project in April 2021. Humanitarian impacts have been severe, with displacement surpassing one million and recurrent spikes in 2025 despite security gains. 

Security response. From mid-2021, Rwanda deployed forces at Mozambique’s request; SADC launched the SAMIM mission. Security pressure reduced insurgent control, but SAMIM completed its withdrawal on 15 July 2024, leaving a larger Rwandan footprint backed by European funding and private-security contracting linked to the LNG restart ambitions. Violence persists in rural districts with periodic surges. 

Resource governance flashpoints. Montepuez’s world-class rubies remain a magnet for artisanal miners, criminal networks, and heavy-handed security—periodically igniting deadly clashes (2017–2025) and feeding narratives of dispossession that insurgents exploit. 

Bottom line: The Cabo Delgado insurgency is not a simple “foreign jihadist” import; it is a hybrid rebellion where local grievance and opportunity structures intersect with transnational jihadist branding and financing. Sustained coercive pressure can contain it, but only governance change can durably resolve it. 

5) Cross-cutting drivers of disorder

Center–periphery trust deficit. Early villagization and sidelining of customary authority created enduring mistrust toward state agents, regularly exploited by insurgents and opposition actors alike.  External destabilization and proxy dynamics. RENAMO’s genesis as a Rhodesian/South African instrument entrenched a politics of militarized opposition and retaliatory state security practices, normalizing coercion as bargaining.  Political economy of extractives. Mega-projects (LNG) and high-value minerals (rubies) have generated visible rents with limited local inclusion, land displacement, and securitized enclaves—fertile ground for criminality and insurgent propaganda.  Governance and fiscal crises. The hidden-debt affair shrank fiscal space, spurred donor suspensions, and weakened service delivery precisely as Cabo Delgado required scaled investment—corroding state legitimacy nationwide.  Electoral credibility gaps. Contested results and perceived partisan security forces periodically catalyze protest waves and opportunistic violence, including against extractive assets. 

6) The current picture (late-2025)

North: Rwandan forces (~4,000 by mid-2024) and Mozambican units hold key nodes; militants retain rural mobility with episodic displacement spikes (e.g., 46,000 newly displaced in July 2025). SAMIM has departed; EU funding and commercial contracts deepen Rwanda’s role; LNG restart remains conditioned on security and community acceptance.  Center/South: The 2019 Maputo Accord continues to structure FRELIMO–RENAMO relations; violence linked to local politics and criminal economies (illegal mining/logging) remains a risk. 

7) Scenarios (2026–2028)

Managed containment (most likely). Continued Rwandan and bilateral support keep insurgents dispersed; LNG cautiously resumes; low-level rural violence persists absent governance reform. Risks: over-reliance on external forces; grievance recycling around extractives.  Renewed escalation (plausible). A shock—electoral crisis, security force abuses, or predatory mining crackdowns—triggers local mobilization and insurgent resurgence, again targeting Palma/Montepuez corridors and community leaders.  Gradual de-escalation (aspirational). Credible community engagement, transparent revenue-sharing, and rights-respecting security operations reduce recruitment; DDR and livelihoods programs for defectors scale effectively. (Inferred pathway based on conflict diagnostics.) 

8) Policy recommendations

For the Government of Mozambique

Institutionalize inclusive local governance in the north. Restore and formalize consultative roles for customary and religious leaders; ensure district-level budget transparency and participatory planning tied to extractive revenues.  Rights-respecting security sector reform. Codify clear rules for military–police–private security coordination around mines and LNG facilities; embed independent complaint mechanisms and civilian harm tracking.  Deliver visible services fast. Ring-fence a portion of LNG and ruby-related fiscal flows for clinics, schools, and roads in Cabo Delgado; publish quarterly execution reports to rebuild trust after the debt scandal. 

For international partners (AU/SADC, EU, Rwanda, bilateral donors)

Conditioned security assistance. Tie support to protection-of-civilians benchmarks and community-approved operations; maintain external forces as mentors and enablers, not substitutes for local legitimacy.  Anti-corruption and fiscal transparency. Continue post-“hidden debt” reforms—e-procurement, asset disclosure, debt registry integrity—and pair budget support with independent monitoring.  Conflict-sensitive extractives. Require companies to adopt benefit-sharing compacts, inclusive land compensation, and open data on social payments; third-party audits for security-provider conduct. 

For extractive companies (LNG, rubies)

Community partnership as a security strategy. Expand local hiring/training, grievance redress, and shared infrastructure; coordinate with state/NGOs on livelihoods to undercut artisanal-criminal recruitment.  Human-rights compliance in site security. Align with the Voluntary Principles on Security and Human Rights; prohibit abusive joint patrols; publish incident reporting. 

9) Conclusion

Mozambique’s civil disorder is best understood as layered conflict: Cold War proxy war legacies, unresolved political inclusion, and a new insurgency fed by the political economy of extraction and a damaged social contract. Security operations have bought time; only governance reforms that share power and rents—and that respect communities—can transform that time into peace. The policy window remains open but will narrow if economic recovery proceeds without legitimacy or if coercion again substitutes for consent. 

Select references

General Peace Agreement (Rome, 1992) & implementation assessments.  Maputo Accord & DDR (2019).  Historical drivers: villagization, customary authority, and RENAMO’s origins.  Cabo Delgado conflict analyses and updates.  Extractives, security contracting, and governance.  Hidden-debt crisis and donor responses.