Ideas to Action:

Independent research for global prosperity

This brief summarizes three years of research under the project, “Rethinking Humanitarian Reform,” led by Jeremy Konyndyk, Patrick Saez, and Rose Worden, and funded by the aid departments of the United Kingdom and Australia. The project aimed to understand the incentives behind the humanitarian system and shift them to better prioritize the needs of affected populations. Read more at

Past reform agendas have failed because they ignored the underlying architecture

In 1996, the Joint Evaluation of Emergency Assistance to Rwanda spurred a movement to improve the quality and accountability of humanitarian aid through common standards and codes of conduct. Less than a decade later, in 2005, the Humanitarian Reform agenda was launched by Jan Egeland (then the United Nations (UN) Emergency Relief Coordinator. It followed a number of crises—notably Darfur and the Indian Ocean tsunami—in which the humanitarian sector had marginalized crisis-affected people while mounting disjointed and sub-par relief operations. Changes sought to pull the disparate components of the sector towards a more unified approach to leadership, coordination, financing, and accountability. They usefully established a more predictable international humanitarian coordination structures and a new contingency fund (CERF), under the leadership of the UN.

Just a few years later, significant shortcomings again emerged in the responses to the 2010 Haiti earthquake and Pakistan floods. Humanitarian donors and agencies launched a new round of reforms, this time branded as the Transformative Agenda, to deliver on the aspirations of the earlier reforms in large-scale emergencies. But while rapid-response protocols improved, that agenda likewise under-performed on other objectives. Following the fragmented response to the Syria crisis and the Ebola outbreak in West Africa, a fresh set of reforms was launched via the 2016 World Humanitarian Summit—notably a package of comprehensive commitments between donors and aid agencies to efficiency and effectiveness, known as the Grand Bargain. In it, participants committed to improve humanitarian financing by increasing direct support to local and national responders, reducing earmarking, and including people receiving aid in making decisions which affect their lives.  

The COVID-19 pandemic should have been a watershed moment for putting those commitments into practice. As we wrote in April 2020,[1] as the pandemic was gaining momentum,

“This is a crisis of truly global scale and it will place enormous constraints on traditional humanitarian operations: models accustomed to surging support toward geographically delimited crises must now tackle a geographically unlimited crisis and huge obstacles to surging personnel and resources…Humanitarians will have to rethink the way the response is planned, coordinated and financed…All this should press traditional humanitarian actors toward a deeper and more equitable partnership with, host governments and local actors – both because local leadership in this crisis is critical to success, and because it is an operational inevitability.”

But in the face of COVID-19 challenges, donor governments and international aid organizations didn’t lean into their reform commitments around financing and localization, but rather leaned away from them, at least initially. Big aid agencies raised money for themselves, rather than stepping back to facilitate funding for local partners. Donors were slow to release money, and when they finally did, they targeted funds toward those same large agencies rather than those closest to the frontlines.

The fact that COVID-19 prompted a pivot back to habitual practices—rather than accelerating a shift toward reform commitments—is a signal that the Grand Bargain reforms, like those before them, have not been transformative. Worse, they might have entrenched power imbalances, inefficiency, and a lack of accountability to people in crises. Every major disaster is a test of how far the humanitarian system has changed. Earlier crises forced a reckoning of the shortcomings of previous reforms; now, the humanitarian sector must learn from its response to COVID-19.

But the reckoning, this time, must be fundamentally different. For the underlying failure to transform is rooted deep in the structures that underpin global humanitarian assistance. The humanitarian sector remains unable to achieve the aspirations it set out for itself 15 years ago: providing aid that is coherent, timely, and accountable to those it serves. The obstacle is not that we have been setting the wrong goals, but rather that we have not altered the architecture tasked with delivering on them.

Past reform agendas have avoided meaningfully challenging that architecture. They have sought to make the humanitarian sector more cohesive and responsive to affected people without altering the fact that the sector’s power structures, bureaucratic incentives, and core business model all tilt toward donors and aid providers rather than aid recipients. The result has been wave after wave of normative commitments and technical guidance, but power, incentives, and resource flows have been left fundamentally unchanged. The resulting effect is better-coordinated supply but no fundamental move towards humanitarian action driven by those who need it.

This presents a hard choice to the powers-that-be in the humanitarian sector. Are they truly willing to shift their power? Or are they content to leave the power structures intact, and accept thereby that the reform commitments they have made over the last two decades will never be meaningfully realized?

Ultimately, it is a political and philosophical choice. Are humanitarian actors committed to building a system in which affected people will set the tone, rather than big institutions? Are those big institutions, the leaders who run and oversee them and the donors who fund them, willing to redefine their role and share their power? Will they support a sector in which institutions measure their impact in terms of partnerships and outcomes, rather than revenue? Can international institutions evolve toward enabling others’ success, rather than emphasizing their own delivery and flag-planting? Can global humanitarianism adapt to elevate the views of affected people, rather than the mandates of global institutions, as the organizing principles for humanitarian action?

If there is convergence toward these changes—which right now are affirmed in rhetoric but ignored in the sector’s revealed preferences—then there is a path. Changing these deeper fundamentals of the humanitarian sector will take political will. It will require overcoming a collective action problem in which many actors are dissatisfied with aspects of the present model, but reluctant to invest the effort—and risk—that would go into fixing it. It will take both a vision, and a political opportunity to enact that vision.

For the last three years, the Center for Global Development (CGD) has sought to propose actionable ideas to start realizing that vision. We have published a set of policy papers and blog posts critiquing the fragmented, supply-driven humanitarian status quo and outlining elements of a more systemic and demand-driven future. These products have been informed by twice-yearly, in-depth private convenings of senior leaders from both aid agencies and donor governments to discuss opportunities and constraints to implementing these ideas.

Collectively these writings attempt to articulate a set of changes that, in the words of the CEO of the International Rescue Committee (IRC), David Miliband, would help shift from the current humanitarian sector into a true humanitarian system. That is, to evolve from a landscape of individual institutions with overlapping values each pursuing their own institutional interests, and toward an alliance of networked institutions working in intentional complementarity toward shared priorities and outcomes. From a sector that is organized around the question of “what is my institution mandated to supply?” towards a network that asks, “how can we adapt ourselves to what crisis survivors are demanding?”

We proposed several major changes to the humanitarian architecture to facilitate such an evolution. Importantly, these shifts would span four levels of the humanitarian landscape:

  • adopting independent accountability mechanisms to enable a people-driven response;
  • using area-based models of frontline coordination;
  • remaking the financial business model to resource the humanitarian system as a public good provider; and
  • adapting the sector’s governance to more effectively steer its humanitarian impact.

Sector-wide change requires mutually reinforcing reforms across all levels. Field reforms will falter if financing practices do not shift to enable them; financing practices require accountability; and oversight that can only be delivered through different management and governance structures.

Change one: move to a people-driven response

The first major change, and the foundation on which all others are built, is to proactively put the perspectives and voices of crisis-affected people at the center of humanitarian decision-making to make aid providers more meaningfully accountable to those they serve. These ideas have been repeatedly codified in humanitarian reform commitments as far back as the early 1990s, but those commitments have yet to translate into widespread changes to humanitarian practice. As our 2019 paper – People-Driven Response – argues, the participation “revolution” promised in the Grand Bargain will not be achieved by the same players continuing to wield the same power in the same ways that they always have.[2] The aid community must move beyond technical and rhetorical approaches to accountability and begin instead reshaping the power and incentive structures that influence aid decision-making.

Doing so hinges on three imperatives:

  1. Feedback and accountability mechanisms cannot be mediated by the same agencies whose work they evaluate. To avoid such conflicts of interest, these mechanisms must operate independently of aid agencies, and be able to amplify the views of crisis-affected people directly to donors, the public, and aid governance structures (see figure one).
  2. The aid sector must pursue structural changes to elevate the feedback and perspectives of affected people into meaningful influence over aid decision making. This will mean incorporating those perspectives into all levels of aid decision-making, from project-level implementation processes to crisis-levels strategy discussions, all the way up to global-level governance processes.
  3. Aid organizations will need to pursue corresponding changes to their policies and operations, to build institutional cultures centered on accountability to the people they serve.

Figure 1. Independent feedback and audit mechanisms

Figure 1. Independent feedback and audit mechanisms

Change Two: Ensuring Inclusive Coordination

But simply hearing and amplifying the voices of affected people is not enough if they face a delivery architecture that cannot adapt itself to their perspectives and involvement. Our second paper – Inclusive Coordination – therefore focuses on outlining a new “area-based” model for humanitarian coordination and planning.[3] The current model – the cluster approach – was a step forward when it launched 15 years ago, but has struggled to overcome a range of chronic problems. It fragments humanitarian delivery and funding into technical siloes rather than applying a holistic approach to priority needs. Importantly, it has reinforced the dominance of large international organizations while marginalizing local frontline actors and governments. Our analysis of major humanitarian appeals found that the amounts requested for cluster leads range from 2,000 percent to 20,000 percent of the amounts requested for even the largest national partners.[4]

A reorganization of coordination and planning around frontline aid recipients is badly needed. A next-generation architecture should be created based on principles borrowed from area-based programming. It would treat needs holistically within a defined community or geography; provide aid that is explicitly multisector and multidisciplinary; and design and implement assistance through participatory engagement with affected communities and leaders. Three shifts would enshrine geographical area rather than sector as the essential organizing principle:

  1. Reorient the humanitarian program cycle around explicitly multisectoral interventions rather than boundaries of the individual clusters;
  2. Delink cluster leadership from financial intermediary roles; and
  3. Open coordination and planning to much greater participation and leadership by local organizations and communities.

This should not mean the end of the clusters: a hybrid model is possible, where capital-based clusters retain their technical, normative, and quality assurance role, while program cycle functions like needs assessment, program design and planning, resource allocation, monitoring and evaluation would be decentralized to subnational coordination hubs.

Figure 2.  Hybrid area- and sector-based coordination model

Figure 2.  Hybrid area- and sector-based coordination model

Change Three: Financing the Humanitarian Public Good

An integrated field delivery model cannot succeed if it remains fueled by an outdated financing architecture. Our paper on Financing the Humanitarian Public Good outlines proposals to start changing unhelpful incentives built into the dominant financing practices, and that haven’t been fundamentally addressed by reform processes.[5] Despite common appeals, resource mobilization remains fragmented between multiple agencies and pledging events. They translate into as many priorities, timeframes, and oversight practices as there are donors. The vast majority of humanitarian finance is spent on responding to needs that are predictable, for example climate or health risks that can be modelled, or protracted situations of displacement. By contrast, funding decisions are usually made after the event, and on a short-term, emergency basis. And because grants are mostly bilateral and earmarked, aid agencies are overly reliant on overheads to resource their core functions, which encourages negative competition. This translates into donors using large UN agencies as their main financing intermediaries rather than funding frontline organizations more directly. It also lengthens costly delivery chains and reinforces siloes according to global mandates rather than needs on the ground.

The core challenge here is to envision a model that is truly needs-based, while recognizing that limited donor bandwidth will still inherently require intermediaries for transferring donor funds on to frontline aid programs. Squaring that circle will require three important shifts:

  1. At the global level, aid must support crises more consistently according to the severity of needs and risks. Donors should pre-arrange and pool a portion of their humanitarian funds towards predictable crises on a multi-year, multisectoral basis. This pool would serve a central treasury function, allocating resources to crises according to objective assessments of priorities, not mainly political choices or pre-determined mandates. Other replenishment mechanisms such as GAVI or the Global Fund provide models for how to do this.
  2. At the country-level, the balance should shift towards using multisectoral pools as intermediaries, instead of the inefficient and siloed delivery chains through large agencies. This would help shape country operations according to needs, and support area-based coordination models. Existing UN and NGO pooled funds would need to evolve to deliver this ambition.
  3. But if agencies are used less as intermediaries, a different financing model is required to resource their core responsibilities. Donors and agencies should agree what constitutes core international functions and how much they should cost, and donors should allocate predictable core funding to them, rather than project overheads.

Figure 3.  Rearranging financing flows

Figure 3.  Rearranging financing flows

Change Four: More Effective Humanitarian Governance

And finally, shifting power and improving cohesion in humanitarian financing and delivery will also require a different approach to humanitarian governance. Governing institutions – such as member state boards of multilateral organizations and NGO boards of directors – have tremendous influence over the strategic direction of individual institutions and the sector writ large.

However, as our paper on Effective Humanitarian Governance outlines, these governing bodies are exclusive, organizing power and influence around powerful governments and individuals. There are few entry points for aid agencies’ downstream clients, or for host governments with the willingness and capacity to take the lead. At the system level, inter-agency coordination does not enable accountability for effective collaboration to deliver common outcomes. And there is no high-level forum in which to align behind a common understanding of priority needs and risk, and sector-wide effectiveness. While the for-profit world runs on a clear quantitative metric of profitability, the non-profit world runs on the more qualitative measure of mission impact. Whose voices define that mission, and whose perspectives assess that impact, are therefore profoundly important.

Moving towards a sector in which mission effectiveness is measured through partnerships rather than individual capabilities, and outcomes rather than fundraising totals, will depend on governance bodies promoting those priorities. Communities affected by crises should be more directly represented in governing boards and in how donors measure the performance of the organizations they fund. National and local humanitarian organizations who have proven their ability should be resourced to lead disaster response. Multistakeholder platforms at the global and country level should be established to assess needs, agree policy directions, and mobilize and align resources accordingly. This could include using constituency-based models of governance for the global replenishment and country-based pooling mechanisms discussed above.

Finally, the power to assess collective effectiveness should become an independent function systematically deployed alongside humanitarian operations, directly informing governing structures and funding relationships. The growing practice of independent accountability audits and perceptions surveys provides a useful basis on which to build.

Making it happen

The starting point for our research project was an assessment – which has become a leitmotiv in the sector – that humanitarian reform processes have failed to deliver transformative change, even as they delivered marginal improvements in collective practice.

The annual independent assessment of the Grand Bargain for 2020[6] noted an assortment of areas where progress was evident and found that the process provided a useful forum for aid groups to discuss and exchange ideas on changing their individual practices. But troublingly the report also observed that “substantive shifts in practice have not yet been realized in most areas” and that four years into the process, “there was still no strategy for rolling out the commitments to country and crisis levels.” The report goes on to note that the Grand Bargain workstreams “have remained largely focused on technical issues, with some success in this respect. But there was no corresponding political investment in addressing the long-standing challenges that continue to inhibit change.” This has in turn meant a “continued failure to take to scale many of the positive or illustrative practices that have emerged through the Grand Bargain process.”

For those who have lived through past reform efforts, there is a bit of déjà vu here. The Grand Bargain has ended up on a similar path to processes that preceded it: dissipating political engagement despite the best efforts of skilled leaders, mired in technocratic committees, and producing modest progress but nowhere near the intended degree of tangible change.

There are parallels here with other global challenges where transformational change is required – the most prominent being climate change. Borrowing from studies of ‘climate delay’ – the idea that, while climate change is no longer denied, stakeholders with powerful interests find reasons to delay necessary transformations – we have mapped below the types of arguments heard from various stakeholders during the course of our research to justify the slow progress of humanitarian system reforms. These narratives are powerful, and difficult to counter only through top-down models of change.

Figure 4. Humanitarian system transformation: delay narratives

Figure 4. Humanitarian system transformation: delay narratives

As scholars of the humanitarian system note[7], system-wide reform processes have used a standard, policy-led model of change: problems are identified and policies agreed to address them, based on more or less robust evidence, and specific actions are ‘rolled-out’ to implement those policies. As such, the change process looks on paper like a project that is easy to control.

This type of linear management often faces difficulties at the organizational level. These difficulties are multiplied in a sector that is made up of a multitude of different types of organizations with no central command and control, and where power and influence are exerted through a myriad of individual relationships and incentives. The diffuse nature of the Grand Bargain implementation process – centered at headquarter level, churning away simultaneous joint and individual commitment processes with varying degrees of political momentum, and relying on each signatory organization to voluntarily apply those commitments throughout its own operations – was not a recipe for success. Rehashing yet another process of normative commitments and technical fixes negotiated in Geneva, New York, and donor capitals is unlikely to yield dramatically different results. There needs to be a common governance platform to manage accountability for both individual and collective changes.  

The good news, however, is that change, even radical change, does happen in the humanitarian system. But it often happens on the ground, in unpredictable and non-linear ways, when the system is subject to external forces. The COVID-19 pandemic has provided such an impetus for change. Even though it has not immediately modified the practice of those who wield most power in the system, “in a myriad of ways, many still unforeseeable, the intensity of the present period is accelerating change,” as one of Médecins Sans Frontières’ (MSF) leaders put it.[8] While the traditional humanitarian operating model struggled to adapt, the role of governments and national systems in managing basic services and social protection has become center-stage, as have the activities of local community groups. In this context, and faced with stagnant aid budgets, international humanitarian organizations are having to rethink where their added value lies: substituting the state or advocating for inclusion of the most marginalized? Direct delivery or solidarity with local actors?

Instead of designing a new policy-led reform process at headquarters, humanitarian donors and agencies now have an opportunity to harness change that is already happening. This should start by identifying which adaptations are working that, if scaled-up or replicated, could help accelerate progress towards the objectives of the Grand Bargain 2.0 - a more efficient, effective, locally led and accountable humanitarian system. With that clear purpose in mind, bringing different stakeholders together to pilot a deliberate expansion of those changes in specific contexts will be key to test the viability, for instance, of pre-arranging finance at scale, new financing intermediaries for local actors, area-based coordination models or independent outcome measurement. Investing adequate resources to deliver such pilots will be key. Governance will need to be realigned. As ever, political and financial support from the main humanitarian donors will be critical to reach the critical mass required for such pilots to deliver tangible, systemic change.

Jeremy Konyndyk co-led the research that generated this brief until his departure from CGD in January 2021, and contributed to early drafts of the content and recommendations. The research and drafting process were completed following his departure from CGD. Patrick Saez and Rose Worden exercised final editorial control over the contents.




[4] Ibid.



[7] Knox-Clarke, P. (2017) Changing humanitarian action? ALNAP Working Paper. London: ALNAP/ODI.


Rights & Permissions

You may use and disseminate CGD’s publications under these conditions.