Why Peacebuilding Stays Underfunded: What the UN’s New Overview Gets Right, and Misses

Image: Domenichini Giuliano / Shutterstock

This policy brief examines the underfunding of peacebuilding and conflict prevention. The United Nations’ first Peacebuilding Overview, Building Peace in a Changing World: Investing in Peace When the World Pays for War (2026), documents a stark imbalance: over the past two decades the world has spent roughly one dollar on peace-related aid for every hundred dollars of military expenditure. The report attributes part of this gap to peacebuilding’s weaker shared data systems and evidence practices. That diagnosis is important but incomplete. Governments underfund prevention less because they lack proof that it works than because military spending responds to visible threats and organised constituencies, while successful prevention is harder to see and attribute. Better evidence can improve choices and accountability, but it will not by itself shift resources. That requires predictable financing, genuinely inclusive national ownership, responsibility for implementation and evaluation systems that do not privilege the organisations best equipped to measure themselves.

Contents

• Abstract
• 1. The evidence gap and the deeper incentives problem
• 2. What the report gets right: An operational definition
• 3. National ownership: A shift that still needs testing
• 4. Inclusive coalitions: Promising, but not yet proven
• 5. Evidence systems: Who gets to measure, and who gets measured?
• 6. Recommendations
• Conclusion

Abstract

The United Nations’ first Peacebuilding Overview [1], Building Peace in a Changing World: Investing in Peace When the World Pays for War (2026), documents a stark imbalance: over the past two decades the world has spent roughly one dollar on peace-related aid for every hundred dollars of military expenditure. The report attributes part of this gap to peacebuilding’s weaker shared data systems and evidence practices.

That diagnosis is important but incomplete. Governments underfund prevention less because they lack proof that it works than because military spending responds to visible threats and organised constituencies, while successful prevention is harder to see and attribute. Better evidence can improve choices and accountability, but it will not by itself shift resources. That requires predictable financing, genuinely inclusive national ownership, responsibility for implementation and evaluation systems that do not privilege the organisations best equipped to measure themselves.

1. The evidence gap and the deeper incentives problem

The Peacebuilding Overview, produced by the Peacebuilding Impact Hub of the Peacebuilding and Peace Support Office, is the first United Nations publication of its kind. Drawing on government, civil society, academia, the United Nations system and other multilateral institutions, it seeks to clarify peacebuilding support, strengthen evidence, promote nationally owned approaches and improve impact measurement. It appears as conflict rises and aid contracts.[2]

The Overview arrives at “a time of serious global disruption”. UCDP recorded sixty-one state-based armed conflicts in 2024, the highest number since 1946, while nearly two billion people live in fragile and conflict-affected settings. Its central statistic is stark: global military expenditure between 2002 and 2024 reached US$45.23 trillion, against US$446.8 billion of official development assistance allocated to peace. Peace-related ODA amounted to 0.99 per cent of military expenditure. This brief asks why that imbalance persists despite the growing case for prevention and peacebuilding.[3]

The categories are not fully comparable. Military expenditure includes deterrence and territorial defence, while peace-related ODA is far narrower and excludes much domestically financed prevention, mediation and institutional resilience. The ratio is therefore not a simple cost-benefit comparison. Even so, its scale reflects a profound imbalance in financing priorities.

The report places considerable weight on the evidence gap: peacebuilding lacks the shared data practices needed to demonstrate results, learn systematically and communicate value. It also acknowledges that evidence must be translated into arguments capable of moving policymakers and citizens.[4]

The problem goes deeper. Military budgets respond to identifiable threats and have ministries, armed forces, industries, legislators and electorates organised around them. Successful prevention may avert conflict or reduce its severity and duration, but those results are harder to quantify and attribute than the consequences of violence.

This creates an asymmetry of constituency and attribution. Violence produces visible events and organised institutions ready to demand resources. Successful prevention produces no comparable event. Its beneficiaries are dispersed, and few constituencies organise around violence that did not occur.

Responsibility for failed prevention is also easily dispersed. Effective prevention, by contrast, may require governments and international institutions to confront politically sensitive drivers of conflict, including exclusion, corruption, abusive security practices and sometimes their own conduct or that of allies.

The financing problem is therefore not simply measurement. Political systems finance responses to visible disorder more readily than the preservation of conditions in which disorder does not arise. More evidence may improve peacebuilding, but there is little reason to assume that evidence alone will overcome that incentive structure.

2. What the report gets right: An operational definition

The report’s most durable conceptual contribution may be its distinction between peacebuilding, understood as the nationally owned process undertaken by governments and their citizens, and peacebuilding support, the external assistance that accompanies it.

The Overview proposes seven operational elements. Peacebuilding support is politically anchored; addresses root causes; facilitates inclusive national ownership; applies across conflict phases; focuses on political processes, safety and security, rule of law and human rights, core government functions, basic services and the economy; aims for sustainable institutional impact; and catalyses development and wider financing.[5]

This is a useful operational definition rather than a claim to finality. It provides a common basis for analysis, policy, resource allocation and measurement, while allowing its elements to be tested against experience.

3. National ownership: A shift that still needs testing

The report traces a significant shift towards nationally owned prevention through the Secretary-General’s 2023 A New Agenda for Peace and the 2024 Pact for the Future. Member States are encouraged voluntarily to develop national approaches to the drivers of conflict within their own societies.

Since 2023, Canada, São Tomé and Príncipe, Kenya, Mauritania, Norway and Timor-Leste have voluntarily gone before the Peacebuilding Commission to present national prevention strategies and efforts, share experiences and challenges, and seek international support where needed.[6]

Kenya shows what sustained preventive infrastructure can achieve. Ahead of the 2010 constitutional referendum, national institutions and UNDP established UWIANO, an early-warning and response platform linking national and subnational actors. It identified and pre-empted nearly 150 incidents of violence; the referendum proceeded without reported violence. The Overview puts the early-warning cost below US$1 million, against about US$3.6 billion in losses from the 2007–2008 post-election violence.[7]

The comparison cannot establish a simple causal counterfactual: the peaceful referendum had many causes. It does show, however, that functioning preventive infrastructure can operate at a fraction of the cost of violent crisis and help preserve the political space for national goals.

National ownership cannot be assumed merely because a government presents a strategy internationally. A government may itself be implicated in the exclusion, repression or conflict dynamics the strategy addresses. The test is whether civil society, women, young people, minorities and politically marginalised communities had a meaningful role in defining risks and responses.

The Overview itself insists that governments cannot act alone and concedes that without means of implementation the paradigm shift towards nationally led prevention will remain “largely rhetorical”.[8]

National ownership should therefore be judged by what follows presentation: whether domestic resources are committed; institutions change practice; excluded constituencies gain a meaningful role; financing follows; and the strategy survives when prevention becomes politically inconvenient.

4. Inclusive coalitions: Promising, but not yet proven

The Overview also promotes Inclusive Coalitions: networks deliberately constructed across lines of conflict. It identifies credibility, local knowledge, action orientation, principled pragmatism and authenticity as characteristics associated with effectiveness.

The limitation is that the chapter’s perspectives, case studies and tools are drawn primarily, and in key respects exclusively, from Search for Common Ground. That does not invalidate the model, but it limits claims of generalisability.

Inclusive Coalitions should therefore be treated as a plausible model requiring independent testing across organisations, societies and conflict types. The question is whether the same characteristics appear elsewhere and predict better outcomes.[9]

The report cites a US$1.5 million Peacebuilding Fund investment in Guatemala’s Polochic Valley. A synthetic-control evaluation estimated that the intervention contributed to an eighty per cent reduction in potential conflicts in targeted areas.[10]

That is an impressive finding, but a statistically constructed counterfactual is not the same as an observed untreated comparison group. The result deserves attention without carrying more causal certainty than the method permits.

Borno State illustrates a different measurement problem. The report records more than 250,000 people surrendering from Boko Haram and the Islamic State–West Africa Province between 2021 and 2024, but only about 7,300 were fighters; most were family members or other conflict-affected people.

The report also notes concerns about women and girls, limited data on whether reintegrated people remain in their communities, and the absence of a unified tracking framework.[11]

5. Evidence systems: Who gets to measure, and who gets measured?

The fifth chapter distinguishes conventional monitoring and evaluation from impact evaluation, which seeks to establish whether an intervention caused meaningful change. It highlights the Peacebuilding Fund’s PeaceFIELD initiative and World Bank approaches aimed at greater methodological rigour in fragile settings.

That matters. Peacebuilding actors seeking greater public investment should demonstrate what it achieves. But evaluation is not institutionally neutral.

Rigorous impact evaluation requires baseline data, longitudinal tracking and specialist capacity that many locally rooted organisations lack. Prevention poses an additional problem: success may mean that anticipated violence does not occur. The most important result may be an event that never happened.

The Overview recognises these constraints: conflict settings complicate data collection and causal analysis, politically sensitive surveys can create risks, and impact evaluations are resource intensive.

Without compensating for these differences, an evidence agenda may reward large international organisations with measurement infrastructure while penalising smaller local actors with deeper contextual knowledge and community legitimacy. What donors measure shapes what organisations manage, what programmes survive and whose knowledge counts.

The report partly addresses this through efforts to “decolonise” monitoring and evaluation, including work by Peace Direct to shift knowledge production and decision-making towards community peacebuilders and local languages, concepts and priorities.[12]

The same caution applies to artificial intelligence. The report cites NaijaXLM-T, a Nigeria-specific model that substantially improved hate-speech detection over generic models and, with human moderators, reportedly increased effective moderation of inflammatory content from 35 to 80 per cent.

The example is promising because it is context-specific. The report is clear that AI should support, not replace, human judgement and must account for national ownership, language, trust, access and data asymmetries.[13]

The question is therefore not whether peacebuilding should use stronger evidence and technology. It should. The question is who designs the system, controls the data, can challenge its conclusions and has a fair opportunity to demonstrate value.

6. Recommendations

INCREASE PREDICTABLE FINANCING FOR PREVENTION

Increase the assessed contribution to the Peacebuilding Fund and sustain voluntary contributions alongside it. Member States should build on the General Assembly’s decision to provide US$50 million annually in assessed contributions for five years beginning in 2025. Assessed funding should provide a floor, not a ceiling, and should not substitute for voluntary contributions.[14]

Protect peacebuilding allocations as aid budgets contract. Donor governments should resist allowing prevention and peacebuilding to absorb disproportionate cuts as official development assistance falls. The Overview records a contraction of 23.1 per cent in global aid in 2025.[15]

Connect development finance to national prevention priorities. Multilateral development banks and other international financial institutions should align concessional and development finance more explicitly with nationally identified prevention priorities.

Expand multi–year financing for local peacebuilders. Donors and United Nations entities should use multi-year and intermediary arrangements that allow locally rooted organisations to build institutions, retain staff and generate evidence over time.

MAKE NATIONAL PREVENTION STRATEGIES CONSEQUENTIAL

Track what happens after national presentations. The Peacebuilding Commission should publicly record the financing commitments, policy changes, institutional reforms and follow-up arrangements that emerge from national prevention presentations.

Examine whether national ownership is genuinely inclusive. The Commission should use its convening role to hear civil society and affected constituencies where appropriate. Government ownership and national ownership are not the same thing.

Encourage voluntary peer review of implementation. Member States that present national prevention strategies and efforts should be invited to return periodically to report on implementation, obstacles and lessons.

BUILD AN EVIDENCE SYSTEM THAT STRENGTHENS THE FIELD

Finance the production of evidence as a public good. The Peacebuilding Fund, donors, international financial institutions and the Peacebuilding Impact Hub should finance baseline data, longitudinal tracking, independent evaluation and local research capacity, with findings made accessible across the peacebuilding community rather than retained within individual projects or organisations.

Correct for the measurement advantage of large organisations. Funders should pair higher evidentiary standards with technical support, proportionate requirements and financing for local evaluation capacity.

Test Inclusive Coalitions independently. The Peacebuilding Impact Hub and research partners should commission comparative research across organisations, countries and conflict settings.

Strengthen long–term monitoring of Borno reintegration. Nigerian authorities, working with relevant United Nations entities and independent partners, should develop a unified framework that distinguishes former fighters from family members and other conflict-affected populations and gives particular attention to women and girls.

Establish common safeguards for artificial intelligence. United Nations entities and partners using AI for early warning, conflict analysis or content monitoring should adopt common standards governing data control, privacy, representativeness, human oversight and protection against political misuse.

Conclusion

The Peacebuilding Overview is right that better definitions, comparable indicators and rigorous evaluation can improve programme choices and accountability. But evidence cannot be the whole answer. Even projections of large savings from prevention have failed to shift aid budgets materially because the problem also lies in the incentives governing public expenditure.

Prevention remains disadvantaged because its successes are difficult to attribute, effective action may require governments to confront powerful interests or their own contribution to conflict risks, and responsibility for failure can be dispersed. Military expenditure has ministries, industries and organised constituencies behind it. Prevention has no equivalent constituency organised around the calm it preserves.

There is also a risk that demands for ever more sophisticated measurement shift resources towards organisations best equipped to evaluate themselves, undermining the national and local ownership the Overview rightly emphasises.

The task is therefore larger than proving that peacebuilding works. It is to connect evidence to institutions capable of preserving peace, national strategies to financing and implementation, and rigorous evaluation to local knowledge and capacity. Closing the gap will require predictable financing, clearer responsibility for follow-through and evidence systems designed to strengthen the field rather than rank it.

Notes

[1] Author’s note: The author served as an Associate Editor of the Peacebuilding Overview discussed in this brief. The analysis and recommendations presented here are his own.

[2] United Nations, Building Peace in a Changing World: Investing in Peace When the World Pays for War, vol. 1 (New York: United Nations, 2026), 2, 8–10.
https://www.un.org/peacebuilding/sites/www.un.org .peacebuilding/files/documents/58633_peacebuilding_overview_repo rt_web_v2.pdf

[3] Shawn Davies et al., “Organized Violence 1989–2024, and the Challenges of Identifying Civilian Victims,” Journal of Peace Research 62, no. 4 (2025): 1223–1240, https://doi.org/10.1177/00223433251345636; United Nations, Building Peace in a Changing World, 4, 13, 63, 67, 76; Jordan Ryan, “A Hundred Dollars for War, One for Peace,” Toda Peace Institute Global Outlook (2026), https://toda.org/global-outlooks/a-hundred-dollars-for-war-one-for-peace/

[4] United Nations, Building Peace in a Changing World, 5, 15–16.

[5] Ibid., 8–9, 18–20.

[6] United Nations, A New Agenda for Peace, Policy Brief 9 (New York: United Nations, 2023), https://dppa.un.org/en/a-new-agenda-for-peace; United Nations General Assembly, Pact for the Future, A/RES/79/1 (New York: United Nations, 2024); United Nations, Building Peace in a Changing World, 32.

[7] United Nations, Building Peace in a Changing World, 36. The report states that UWIANO’s national and subnational warning systems identified and pre-empted nearly 150 incidents and that the early-warning efforts cost less than US$1 million.

[8] Ibid., 31–33.

[9] Ibid., 38, 42–44; Search for Common Ground, “Peace Impact Framework” (2023), https://cnxus.org/peace-impact-framework/.

[10] United Nations, Building Peace in a Changing World, 26–29; S. Martinez, Effects of Promoting Management of the Social, Political, and Institutional Environment on Conflict in the Polochic Valley, 3ie Impact Evaluation Report 148 (2025),
https://doi.org/10.23846/PFIE148.

[11] United Nations, Building Peace in a Changing World, 47–48.

[12] Ibid., 26–29, 52–59. See especially the discussion of PeaceFIELD, methodological constraints in fragile settings, and efforts to shift monitoring, evaluation and learning towards community peacebuilders.

[13] Ibid., 6, 59–62.

[14] Ibid., 15, 63. The report states that the General Assembly approved US$50 million per year for five years in assessed contributions to the Peacebuilding Fund beginning in 2025.

[15] Ibid., 5, 13–15; Organisation for Economic Co-operation and Development, “A Historic Decline in Foreign Aid: Preliminary 2025 ODA Data” (Paris: OECD, 2026), https://www.oecd.org/en/data/insights/data-explainers/2026/04/a-historic-decline-in-foreign-aid-preliminary -2025-oda-data.html.

---

The Author