The coordination architecture

1Why coordinate?

Without coordination, aid is concentrated in easily accessible areas, creates duplication, leaves areas forgotten and applies different standards from one organisation to another. Coordination serves to cover all priority needs, without duplication, with common standards, and to speak with one voice to the authorities and donors.

It respects the primary role of the State: when the authorities have the capacity, international humanitarian mechanisms support them rather than replace them.

2Who coordinates?

• The Emergency Relief Coordinator (head of OCHA) at global level, with the Inter-Agency Standing Committee (IASC), which brings together UN agencies, the Red Cross (as an observer) and NGO networks.
• The Humanitarian Coordinator (HC) in the country, often also the UN Resident Coordinator, responsible for the international response.
• The Humanitarian Country Team (HCT): the strategic body around the HC (agencies, international and national NGOs).
• The Inter-Cluster Coordination Group (ICCG), facilitated by OCHA, which links the sectors.
• At local level: regional or district coordination groups, thematic working groups (cash, accountability, PSEA, access).

3The clusters

The cluster approach (2005 humanitarian reform) organises the response into sectors, each with a lead agency at global level, for example: health (WHO), nutrition and water, sanitation and hygiene (UNICEF), food security (WFP and FAO), logistics and emergency telecommunications (WFP), protection (UNHCR), shelter and camp coordination (UNHCR and the International Federation or IOM depending on the context), education (UNICEF and Save the Children).

The core functions of a country-level cluster: support service delivery (avoid duplication and gaps), inform strategic decision-making (assessments, analysis), plan and develop strategy (standards, response plan), advocate, monitor and report, prepare for future crises, not forgetting accountability to affected populations.

The lead agency is also the “provider of last resort”: it commits, as far as possible, to filling critical gaps.

4The humanitarian programme cycle

1. Needs assessment and analysis: Humanitarian Needs Overview (HNO), often with a joint intersectoral analysis of severity.
2. Strategic planning: Humanitarian Response Plan (HRP), with objectives, targets and budget by cluster.
3. Resource mobilisation: appeal, CERF, country-based pooled funds (CBPF), bilateral funding.
4. Implementation and monitoring: monthly cluster reports, dashboards.
5. Operational peer review and evaluation, then a new cycle.

For a project, being aligned with the HRP and registered with the cluster greatly helps in obtaining pooled funding.

5Activating, adapting and deactivating clusters: rules and procedure

Clusters are not a permanent structure that exists by default in every country in crisis. The IASC Reference Module for Cluster Coordination at Country Level sets out a simple logic: a cluster is activated when there is a coordination gap that existing mechanisms, national ones in particular, cannot fill. It is deactivated, or transformed, as soon as that gap disappears. An NGO representative who knows these rules understands why some sectors have a cluster and others only a working group, and can influence those decisions.

The activation criteria

Formal activation rests on two cumulative conditions:

• A marked deterioration of the situation or a significant change (sudden-onset disaster, new conflict, mass displacement) that creates new or more extensive needs in one or more sectors.
• Insufficient national or existing coordination capacity to meet those needs in line with humanitarian principles, either because the scale exceeds the authorities' means, or because the State is a party to the conflict or has no access to certain areas.

The procedure is as follows: the Humanitarian Coordinator (HC), after consulting the Humanitarian Country Team (HCT), proposes the activation of specific clusters, indicating for each one the proposed cluster lead agency. The proposal is sent to the Emergency Relief Coordinator, who submits it to the IASC Principals for endorsement. In practice, the decision is quick (a few days) in a sudden-onset crisis. The HC may also designate a lead agency other than the global lead agency if another organization is better placed in the country, or propose co-leadership with an NGO.

System-wide scale-up

For the most severe crises, since 2018 the IASC has had system-wide scale-up protocols (System-Wide Scale-Up), which replaced the former "Level 3 activation". Activation is decided by the Emergency Relief Coordinator, after consulting the IASC Principals, on the basis of five criteria: scale, urgency, complexity, existing response capacity and the risk of failure to meet the expectations of affected populations. For a limited period (in principle up to six months, exceptionally extendable) it triggers the deployment of experienced staff, the mobilization of accelerated funding mechanisms and strengthened leadership. Depending on the year, it has applied to crises such as those in Yemen, northern Ethiopia, Ukraine, and Türkiye and Syria after the 2023 earthquakes.

What is not a cluster

• Refugee situations fall under UNHCR's Refugee Coordination Model (RCM), which derives from its mandate. The groups there are called "sectors", are led by UNHCR together with the government and partners, and the planning document is a Refugee Response Plan (RRP), not an HRP. Where refugees and internally displaced persons coexist, the Joint UNHCR-OCHA Note on Mixed Situations (2014) sets out the division of roles between the HC and the UNHCR Representative.
• Government-led sectors: in many disaster-prone countries, the line ministry chairs national sector groups, which agencies support. There is no formal activation, but the logic of the core functions remains the same.
• Working groups (cash, accountability, PSEA, access, information management): they are not clusters and have no provider of last resort agency, but they often weigh just as much on the quality of the response. Since 2022, the IASC Cash Coordination Model has attached the cash working group to the inter-cluster coordination group and provides for co-chairing by OCHA or an agency together with a non-UN actor.
• The Areas of Responsibility (AoRs) of the protection cluster: child protection (UNICEF), gender-based violence (UNFPA), mine action (UNMAS), housing, land and property (coordinated notably by the Norwegian Refugee Council). They have their own coordinators and tools, but report to the protection cluster.

Deactivation and transition

The reference module provides that activation be reviewed at least once a year, and that each cluster have a transition plan from the moment it is created: which functions will be handed over, to whom (ministry, national sector group, development platform), on what timeline and with what capacity strengthening. Deactivation is a success when it rests on real national capacity; it becomes a problem when it is dictated by falling funding or political pressure, and leaves areas without coordination.

Key point: activating a cluster does not transfer the State's responsibility. It creates a temporary mechanism, justified by a coordination gap, which must plan from the outset how it will cease to exist.

Recent changes in the architecture

Several developments have changed practice since the early 2020s:

• Area-based coordination (area-based coordination): in large-scale crises, sub-national area coordinators bring all sectors together around a given territory, which facilitates intersectoral analysis and geographic prioritization.
• The merging of the needs overview and the response plan into a single document (often called the HNRP) in many countries since the 2024 cycle, with shorter and more prioritized planning.
• The "humanitarian reset" launched by the Emergency Relief Coordinator in 2025, against a backdrop of sharply falling funding: lighter coordination structures, fewer meetings, refocusing on life-saving needs and "hyper-prioritization" of response plans. The arrangements vary from country to country and continue to evolve: a professional always checks the configuration in force in the country where they work.

6Mapping the coordination of a response: a method for your first ten days

A project manager or coordinator arriving in a new country often loses a month working out "who decides what". A methodical mapping, done in the first ten days, avoids costly mistakes: a missed meeting where the priorities of a pooled fund were set, a project launched in an area already covered, a local authority bypassed. Here is a proven method.

Step 1: gather the reference documents (days 1 and 2)

• The latest humanitarian needs overview and response plan (HNO, HRP or HNRP), or the flash appeal if the crisis is recent.
• The terms of reference of the HCT, the inter-cluster coordination group and the clusters in which your organization works.
• The cluster contact lists and the country page on OCHA's platforms (ReliefWeb Response and the Humanitarian Data Exchange, HDX), with the most recent 4W dashboards.
• Country guidelines on civil-military coordination, access and armed escorts, if any exist.
• The HCT strategy on accountability to affected people and protection from sexual exploitation and abuse (PSEA).
• The country-based pooled fund documentation (operational manual, latest allocation strategy) if your organization is eligible.

Step 2: draw the real organization chart (days 3 to 5)

Official organization charts show the formal links. You need to add what the documents do not say: the actual frequency of meetings, the people who actually attend, the forums where decisions are really taken (often a cluster's strategic advisory group rather than the plenary meeting), and the informal channels (weekly meeting of international NGO country directors, national NGO forum). The following table serves as a working grid.

Step 3: meet the key people (days 5 to 10)

Plan short interviews (30 minutes) with the OCHA head of sub-office for the area, the coordinators of the relevant clusters, the humanitarian affairs focal point of the local authority and one or two national NGO managers. Prepare three questions: what are the priorities and the least covered areas, what local rules must be respected (authorizations, notifications), and how can our organization contribute beyond its projects (co-facilitation, data, technical working group).

Step 4: formalize your organization's representation

For each body, designate a principal representative and an alternate, define their mandate (what they can commit to without approval and what they must refer upwards), and plan a ten-line debriefing note after each meeting. An organization that sends a different person every month, without a mandate or follow-up, has no influence at all.

The timeline of the first weeks of a sudden-onset crisis

Key point: in a sudden-onset crisis, influence is won or lost in the first ten days. The organizations that share reliable data and real capacity early are the ones whose areas and activities appear in the flash appeal.

7The humanitarian programme cycle step by step: timeline, analysis and the role of NGOs

The humanitarian programme cycle (HPC) is not an OCHA administrative exercise. It is the moment when the figures for people in need, the priority areas and the budgets that will guide pooled funds and much of bilateral funding for a whole year are decided. An NGO that does not take part later discovers that its areas are not a priority.

The typical timeline of an annual cycle

The Joint Intersectoral Analysis Framework (JIAF)

Version 2.0 of the Joint Intersectoral Analysis Framework, adopted in 2023, structures the needs analysis. Remember the following elements:

• A five-phase severity scale: 1 minimal, 2 stress, 3 severe, 4 extreme, 5 catastrophic. People in phase 3 or above are considered "in need".
• An analysis by geographic unit (usually administrative level 2 or 3) and by population group (IDPs in sites, IDPs in host families, returnees, host population, refugees where relevant).
• Sectoral and intersectoral indicators, combined according to rules set in advance, with a review by experts who may adjust the result provided they justify their decision.
• An intersectoral people in need figure that is not the sum of the sectoral figures (the same person may have needs in three sectors): it is generally calculated from the sectoral maximum per geographic unit and per group, or from the MSNA household analysis.

From people in need to the target

The response plan never targets 100% of people in need. The cluster sets a target taking into account severity (priority to phases 4 and 5), access, partner capacity, the response of other actors (government, the International Red Cross and Red Crescent Movement, development actors) and expected funding. The budget is calculated either by project (the older method, still used in some countries) or through activity-based costing: number of people targeted multiplied by a reference unit cost, possibly adjusted for access difficulty.

Key point: people in need and target figures are negotiated on the basis of data and methods, not influence. An NGO that brings solid assessment data and a transparent method carries more weight than an organization that lobbies for "its" area without evidence.

Contributing usefully to the analysis: what a good representative does

1. Before the workshop, they check the representativeness of the data for their areas: sample size, collection period, areas excluded for access reasons.
2. They bring dated and sourced secondary data (monitoring reports, assessment results, health facility data), not impressions.
3. They flag the groups that are invisible in average data: persons with disabilities, female-headed households, minorities, IDPs in host families scattered across towns.
4. They provide real unit costs (cost per person of a kit, of a rehabilitated borehole, of a cash transfer), which make a realistic budget possible.
5. They accept collective trade-offs once the decision has been taken, and document them in their own proposals.

8Full example: floods in a protracted crisis, from scenario to cluster costing

Fictional context: the province of Haut-Doma, in the Republic of Lomandi, has for six years been experiencing a low-intensity conflict that has displaced around 140,000 people. A sub-national WASH cluster already exists, co-led by UNICEF and an international NGO. In September, exceptional rainfall causes flooding in five districts. You are the WASH manager of the NGO "Solidarité Bassin", present in the districts of Tiloko and Sarré, and you represent your organization in the sub-cluster.

Step 1: the first 72 hours

The OCHA office in Haut-Doma convenes an emergency inter-cluster meeting. The HC decides not to request a system-wide scale-up (the scale remains provincial), but to launch an addendum to the response plan and a CERF request. The clusters already active are used: there is nothing to activate. That same day, your team sends the sub-cluster coordinator three precise pieces of information: 14 of the 22 water points in Tiloko are submerged, two of the three roads to Sarré are cut off, and you have a stock of 2,400 hygiene kits and 6 tonnes of chlorination products.

Step 2: the sector needs analysis

Ten days later, a joint rapid assessment, supplemented by data from the previous year's MSNA, enables the sub-cluster to estimate the proportion of the population in severity phase 3 or above for the WASH sector.

Step 3: setting the targets

The sub-cluster applies three rules discussed and endorsed in a meeting: priority to districts with severity 4 or above, consideration of the real capacity of the partners present, and consideration of access. The reasoning is recorded in the minutes.

Step 4: costing the response

The cluster uses a cost per person for an emergency WASH package (water supply, chlorination, hygiene kits, emergency latrines in sites, hygiene promotion). The reference cost, calculated from the budgets of four partners, is 24 dollars per person for three months. For Baleko, a coefficient of 1.3 accounts for additional logistics and security costs.

• Standard-cost districts: 257,880 people x 24 dollars = 6,189,120 dollars.
• Baleko: 31,500 people x 31.20 dollars = 982,800 dollars.
• Financial requirement of the WASH sector for the addendum: 7,171,920 dollars, rounded to 7.2 million in the document.

Step 5: the intersectoral figure

At the inter-cluster coordination group, a participant proposes adding up the sectoral figures for Tiloko. The following table shows why this is a mistake.

The figure retained (148,000) is slightly higher than the sectoral maximum, because the household analysis shows that some households are in phase 3 owing to a combination of moderate needs across several sectors.

Step 6: what Solidarité Bassin does

• It reports to the 4W its planned activities in Tiloko (35,000 people) and Sarré (22,000 people), specifying the sub-districts, to avoid duplication with the two other partners.
• It indicates that it could support the national partner present in Baleko with a supply of chlorination products, which is recorded as an action point in the minutes.
• It submits a proposal to the pooled fund aligned with the cluster's targets and reference cost, justifying a slightly higher cost in Sarré by the fact that supplies must travel along a single road.
• It shares its monitoring data every month, which will allow it to demonstrate to the donor the coverage achieved against the sector target.

What the example shows: every figure in the plan can be traced back to a source, a rule and a collective decision. That traceability is what donors and evaluators examine first.

9Common mistakes, contextual variations and the evaluators' perspective

Common mistakes in using the architecture

Variations by context

• Sudden-onset crisis: coordination is built in a few days; everything is provisional. The priorities are speed, information on access and geographic distribution. Cluster standards are set within a few weeks, first based on Sphere, then adapted.
• Protracted crisis: the main risk is bureaucratization (routine meetings, reporting for its own sake). The key issues become the transition to national services, the link with development and peace actors (the "nexus"), and the gradual scaling down of humanitarian coordination.
• Refugee and mixed situations: check who coordinates what (UNHCR for refugees, clusters for IDPs) and avoid double reporting by harmonizing tools.
• Urban contexts: municipal authorities, service operators (water, electricity) and the private sector are central; clusters must adapt to urban administrative boundaries and dispersed populations, and area-based coordination is often more relevant than sectoral coordination alone.
• Limited-access contexts and remote management: coordination often takes place from a capital or a neighbouring country (cross-border hub). The risk is making decisions without a voice from the field; national NGOs present in the areas must have a real place, including in hybrid meetings.
• Strong States: when the government leads the response, international mechanisms take on a supporting role (sector groups co-chaired with ministries) and must negotiate the space needed for impartiality.

What donors and evaluators look at

Inter-Agency Humanitarian Evaluations (IAHEs), conducted by the IASC after major crises, and coordination evaluations regularly examine the same points:

• The relevance of the architecture: did the activated clusters match the needs, and were they adapted and deactivated at the right time?
• The quality and speed of the needs analysis, and its actual use in allocation decisions.
• Coverage: were the most severely affected areas and the most vulnerable groups reached, or did aid concentrate on easy areas?
• The place of national and local actors, and of affected people, in decision-making.
• The cost of coordination (time spent in meetings, number of structures) relative to its added value.

On the donor side, ECHO, BHA and FCDO routinely ask in their proposal forms how the project fits into the response plan and how the organization takes part in coordination. A generic answer ("we participate in the clusters") scores poorly; a precise answer (forum, frequency, role, example of a decision influenced, data shared) is valued.

10Interview questions and key references

Interview questions

1. "Explain the coordination architecture in a country where a Humanitarian Coordinator has been appointed."

A good answer distinguishes the strategic level (HC and HCT), the intersectoral operational level (inter-cluster coordination group led by OCHA) and the sectoral level (clusters, areas of responsibility, working groups), then the sub-national levels. It points out the primary responsibility of the State, the provider of last resort role of cluster lead agencies and the difference with the Refugee Coordination Model.

2. "When should a cluster be activated or deactivated?"

The candidate cites the two criteria (deterioration or significant change, insufficient existing coordination capacity), the procedure (proposal by the HC after consulting the HCT, endorsement at IASC level), the periodic review and the requirement for a transition plan. They illustrate it with a real or realistic example.

3. "How should your organization contribute to the HNO and the HRP?"

The answer describes the cycle timeline, the contribution to assessments (MSNA, sectoral assessments), participation in JIAF analysis workshops with sourced data, the provision of unit costs and the alignment of proposals with the endorsed targets. It mentions the recent merging of the two documents in many countries and the prioritization effort.

4. "The people in need figure for your sector in your district seems underestimated to you. What do you do?"

The answer first checks the data (sample, period, excluded areas), gathers additional evidence, discusses it with the cluster coordinator and IMO before the workshop, and presents a reasoned proposal. It accepts the collective decision and documents the disagreement if necessary. It does not lobby the donor directly, bypassing the cluster.

5. "Why can't the people in need figures of the different sectors be added up?"

The answer explains double counting, the logic of the sectoral maximum or of household-level analysis, and gives an order of magnitude to illustrate it.

6. "What are the limitations of the cluster approach?"

A mature answer cites sectoral fragmentation (hence area-based coordination and intersectoral analysis), the heaviness and time cost, the still insufficient place of local actors, dependence on the funding of cluster lead agencies and the risk of confusion between the coordination role and the agency's interests. It proposes concrete solutions rather than general criticism.

Key references

• Reference Module for Cluster Coordination at Country Level: Inter-Agency Standing Committee (IASC).
• System-Wide Scale-Up Protocols (System-Wide Scale-Up Protocols): IASC.
• Standard Terms of Reference for Humanitarian Country Teams: IASC.
• Humanitarian programme cycle reference documents and annual guidance for the humanitarian needs overview and response plan: OCHA.
• Joint Intersectoral Analysis Framework (JIAF 2.0): IASC and JIAF project partners.
• Multi-Cluster/Sector Initial Rapid Assessment (MIRA) guidance: IASC.
• Refugee Coordination Model and Joint UNHCR-OCHA Note on Mixed Situations: UNHCR and OCHA.
• Cash Coordination Model: IASC (2022).
• The Sphere Handbook (2018 edition) and the Core Humanitarian Standard on Quality and Accountability (CHS, 2024 edition).
• Inter-Agency Humanitarian Evaluations (IAHE): IASC; ALNAP publications on the State of the Humanitarian System.