Image by ©2020 European Union
  • Report
  • 1 April 2021

Development actors at the nexus: Lessons from crises in Bangladesh, Cameroon and Somalia: Chapter 2



A key part of how development actors strategically engage in a crisis context is determined by who they partner with and whether (and the ways in which) they prioritise the livelihood and recovery needs of crisis-affected populations. The case studies explored how development actors work with governments in crisis-affected countries, especially regarding support at the policy level, in resource allocation and service delivery in targeting crisis-affected populations. They also investigated how bilateral and multilateral development donors partner with NGOs and UN agencies at the country level and the impact this has on their ability to support crisis-affected populations. This section summarises key findings and considerations that emerged from the case studies on effective partnerships in crises.

Share section

Key Findings

How do development actors work with governments, and how does this impact their ability to support livelihoods, recovery, development and peace in crises?

The different approaches to partnership that humanitarian, development and peace actors adopt can support but also potentially limit their ability to work collaboratively at the nexus. The case studies highlighted the importance of engagement with government institutions − central to good development practice − and how this can enable an effective presence in crisis settings and therefore lay the foundation for collaboration with humanitarian and peace actors. But partnerships with government may also pose challenges, where government institutions may be weak or party to conflict, and where nexus partners, particularly humanitarian actors, feel that such a partnership undermines humanitarian principles of neutrality and impartiality.

Most development partners − donor governments, multilateral organisations and funds, and MDBs − commit to work in partnership with national governments, including in crisis, conflict and fragile settings. Aid effectiveness principles for engagement in fragile states stress the importance of national ownership, aligning support with nationally led development strategies, moving towards the use of country systems for delivery, and investing in institutional capacity building.[1] In practice, development partners have found it challenging to scale up delivery through country systems and achieve policy alignment in fragile states, due to weaknesses in the policy and institutional environment, national capacities and levels of trust.[2] This was the case in Somalia where the nascency of the federal government and the weakness of public financial management systems has meant that many donors are still hesitant to provide direct budget support or financing through government systems. In some cases, globally set preferences for the choice of aid modality override country-level considerations.[3] In addition, there is a perception among some actors that development cooperation does not prioritise crisis-affected regions at the subnational level, target the most vulnerable or marginalised populations, or focus on the sectors/sub-sectors that would most likely benefit from them. This appears to be the case in Cameroon, where longstanding socioeconomic marginalisation and underinvestment in crisis-affected regions are recognised as drivers of structural vulnerability[4] and where ODA is concentrated in non-crisis regions. But, overall, substantiating these complaints is difficult due to data challenges, including limited spatially disaggregated data.[5]

MDBs’ ability to leverage their strong relationships with the government and deliver at scale has been described as game-changing. The World Bank has recently scaled up its engagement in fragile and conflict-affected contexts, refugee settings, disasters and in building resilience to other shocks, in many ways creating a paradigm shift in the scope, scale and nature of development cooperation in crises. This is evident across all three case study countries, with large-scale World Bank projects focused on medium- to long-term risk reduction, recovery and resilience in disaster or conflict-affected regions, refugees and host communities, national safety nets, and other areas. These are financed through specific windows tailored to crisis (discussed in the financing section, later) and also prioritised within the World Bank’s country partnership frameworks and performance-based allocation mechanisms. While there is some debate about whether MDBs’ orientation towards productive investment, private-sector-led growth and debt financing is appropriate in fragile and conflict-affected contexts, overall interviewees saw their role in protracted crises as overwhelmingly positive in driving change. In the three case study countries in particular, in addition to bringing large-scale finance to address structural challenges, the World Bank and, in the case of Bangladesh also the Asian Development Bank (AsDB), have played an important part in strengthening the role of the government in addressing crises through providing technical support and longstanding partnerships with the government.

Given their leverage, the support of large development actors, particularly MDBs and the EU, has led to shifts in government policy and practice towards taking a more comprehensive approach to protracted displacement in some contexts. In Cameroon, the World Bank in partnership with the United Nations High Commissioner for Refugees (UNHCR), through dialogue with the government, played a key role in promoting the shift from humanitarian programmes that provide direct assistance to refugees to an area-based approach that covers both displaced and host populations. This partnership also led to the inclusion of refugees in the national safety net system. In Somalia, durable solutions for internally displaced people and returnees are identified as a national development priority, and the EU, bilateral partners (UK and Denmark) and UN agencies (United Nations Development Programme (UNDP), International Organization for Migration and UNHCR) with support from the Regional Durable Solutions Secretariat, have played a key role in progressing this agenda at the policy and operational levels, although funding remains primarily humanitarian not developmental.

Development partners face the dilemma of how to balance an approach based on partnership, cooperation and mutual accountability with the government with responsiveness to crisis-affected populations in contexts where political commitment is wavering. In Bangladesh, the World Bank and AsDB have helped to broaden the response to the Rohingya refugee crisis to address the development needs of Cox’s Bazar district as a whole, which has been important in the context of mounting resentment by host communities. While the government has been firmly opposed to policy reforms that would signal support for long-term integration of refugees and has resisted advocacy from a humanitarian protection perspective, it has been much more open to dialogue with development partners focusing on the socioeconomic and developmental needs of the district. This has resulted in a shift in stance on certain issues – for example, the construction of infrastructure in the camps − that affect the district’s ability to sustainably manage the crisis as well as the wellbeing of refugees. In Cameroon, the World Bank, EU and UN are also working collectively, in dialogue with the government, to develop a comprehensive approach to internal displacement in the north and northeast, including through the development of the Recovery and Peace Consolidation Strategy for Northern and East Cameroon 2018−2022; however, this has yet to be endorsed by the government.

Partnerships with local governments through local governance programmes have been an important entry point to engage in crisis regions across the three countries. While the main focus is often on strengthening local government institutions and their capacity for service delivery, area-based approaches that engage multiple stakeholders in inclusive local development planning have been a promising approach to supporting municipalities while also enhancing coordination among local actors.[6] When carried out in a conflict-sensitive manner, participatory development planning can create space for dialogue, build trust and enable local actors to identify interventions that will support peace. In Cameroon, the National Community Driven Development Programme (PNDP), set up in 2004 and funded by the Agence Française de Développement, EU and World Bank, has facilitated participatory local development plans at the municipality level. Challenges remain with central government budget allocations to these plans and with their adaptability to changes in local needs. Nonetheless, the PNDP has provided an anchor for development actors to engage with local actors in the north and east, to develop (and in some cases finance) livelihoods and infrastructure solutions that benefit crisis-affected populations (e.g. the local employment programme in the north, water-piping). Similar area-based approaches with a strong peacebuilding focus have been used in Somalia where local government capacity and legitimacy is absent or weak. Despite efforts such as these to work with local authorities, there is a perception that development actors over-emphasise top-down policy and institutional reforms and partnerships with central government, with crisis-affected regions neglected or overlooked because power is highly centralised (e.g. in Cameroon and Bangladesh) or authority and reach is weak (in Somalia).

Partnership with the government is challenging or impossible in some contexts, particularly where it is an active party to conflict. These challenges can apply to central or local governments. In conflict-affected areas, where insurgents may target aid organisations seen to be in partnership with the government, humanitarian actors in particular emphasise the need to maintain neutrality and independence and not be linked to central or local governments. Development partners have struggled with how to engage with the government in Cameroon as the conflict with separatists in the English-speaking regions has escalated, and many have withdrawn from the region. They were slow to recognise the risk of conflict and to take action to prevent an escalation and, while their leverage may have been limited, the difficulty adapting or exploring alternative partnership approaches has been a challenge. Development actors also face partnership dilemmas where the state lacks authority and control, and ‘illegitimate’ militia have significant local influence, for example in al-Shabaab-controlled areas of Somalia. Focusing development cooperation only in areas of the country controlled by the government risks increasing inequality and marginalisation of conflict-affected areas, potentially fuelling drivers of conflict. In such contexts, development actors must look for alternative ways to reach vulnerable populations, such as working with local actors, multi-mandated UN agencies and NGOs that engage at both local and national policy levels, or through humanitarian actors that have access in the interim, in a way that lays the foundations for peacebuilding and a transfer to legitimate and responsive national systems in the longer term. Although there are examples of good practice, such as in Somalia, this is not yet an evidence-based approach, and there is a need to learn from and scale up these programming models and create instruments to systematically support non-governmental partners.

How do development actors work with non-governmental partners, and how does this impact their ability to support livelihoods, recovery and development in crises?

Although the largest share of development finance in protracted crises is channelled to recipient governments, development partners also work with a range of other partners. This can be a means to promote bottom-up, participatory development and to engage in crisis-affected regions, particularly where government structures are weak or unresponsive to local populations. The largest share of development finance in Cameroon and Bangladesh is channelled to recipient governments, which mirrors the wider trend in protracted crisis countries (Figure 2). In the decade to 2019, 38% of total developmental ODA was channelled to recipient governments in protracted crisis countries, followed by multilateral organisations (17%), NGOs and civil society (12%), donor governments (11%), unspecified public sector institutions[7] (5%), and private sector institutions[8] (2%). Nonetheless, funding channelled to recipient governments varied across the three case study countries, reflecting their different levels of fragility, from only 6% in Somalia, to 49% in Cameroon and 62% in Bangladesh. Likewise, 75% of developmental ODA was channelled to multilateral organisations, NGOs, civil society and the private sector[9] in Somalia, compared with 22% in Bangladesh and 11% in Cameroon.

Figure 2: Channel of delivery of developmental ODA to protracted crises countries, 2010−2019

Figure 2: Channel of delivery of developmental ODA to protracted crises countries, 2010−2019

During the period 2010 to 2019, more than a third (38%) of total developmental official development assistance was channelled to recipient governments in protracted crisis countries, followed by multilateral organisations (17%); NGOs and civil society (12%); donor governments (11%); unspecified public sector institutions (5%); private sector institutions (4%) and third country governments (1%). The remaining 12% went to other channels of delivery.

Source: Development Initiatives based on OECD DAC Creditor Reporting System (CRS).

Notes: Data is in constant 2018 prices. ODA is from DAC and multilateral donors only. Developmental ODA excludes ODA reported under humanitarian purpose codes 720 (emergency response), 730 (reconstruction relief and rehabilitation) and 740 (disaster prevention and preparedness). Protracted crisis countries are those with humanitarian response plans and other appeals for five consecutive years or more in the year of the disbursement.

While subnational data is often poor, evidence from the country studies suggests that government-led delivery and budget allocations to crisis-affected regions are often limited, as in Cameroon. Furthermore, research indicates that development strategies often lack an explicit focus on targeting the populations that are most left behind.[10] Political obstacles play a role, for example in Bangladesh where the government has been reluctant to allocate its own budget to services that primarily benefit refugees, or in Cameroon where there is political resistance to decentralisation. The weakness of local government structures and capacity for delivery is also a major obstacle, as in Somalia and Cameroon. Therefore, providing funding through other channels, while working in parallel to strengthen local government and the national administration, is key.[11] A multi-pronged approach is needed. Maintaining parallel service delivery might meet the needs of vulnerable populations in the short term, but it is ultimately unsustainable and may not aid a transition out of protracted crisis. There are examples where development partners jointly develop programmes with governments to encourage ownership and alignment of priorities, but where UN agencies, international NGOs, private contractors or bilateral partners directly execute the projects, for example the World Bank social safety net programme in Somalia (see Box 2). Where there is genuine work in parallel to strengthen government capacity to deliver, this approach can support a transition towards government-led service delivery while also responding to immediate needs. Cooperation between development and humanitarian actors to work jointly towards a transition to government-led service delivery in the case study countries is not yet systematic – and this remains a core challenge for operationalising the nexus felt globally.

Share box

Box 2

World Bank−government service delivery through humanitarian agencies

Building government capacity is a long-term process. This is particularly true in fragile or conflict-affected contexts where it may not be appropriate for development actors to implement programming or deliver services through government systems. A challenge for donors is how to address the immediate needs of vulnerable people in parallel with capacity-building efforts.

One example of how this is approached in Somalia is the World Bank’s Shock Responsive Safety Net for Human Capital Project, known in Somali as the ‘Baxnaano’ programme. In collaboration with the Ministry of Labour, the World Bank established the programme to lay foundations for longer term social protection systems. The programme supports vulnerable households to increase their income and improve their resilience to shocks through predictable access to cash transfers. The Shock Responsive Safety Net for Locust Response Project builds on this existing programme, with an additional focus on food security and livelihoods.

The safety net programme is delivered through UN agencies and international NGOs with pre-existing programmes, in partnership with the UN World Food Programme and UN International Children’s Emergency Fund (UNICEF). The programme is intended as a transitional measure until government structures are established and have the capacity to lead in government-held areas, and where it is appropriate in terms of safeguarding humanitarian principles. Its branding as a nationally led initiative has been key for building state legitimacy. Decentralising management to member state and local government levels will be crucial to develop the social contract between government and society for greater sustainability and impact.[12]

It was in Yemen that the World Bank pioneered the approach of maintaining the national social protection programme during active conflict by using UN agencies as delivery partners rather than the government.[13] Taking a flexible interpretation of its own policies, the World Bank was able to provide ‘unprecedented and innovative’ International Development Association grant financing to UNICEF and UNDP to provide emergency cash transfers to millions of Yemenis at risk of famine,[14] building on the national safety net programme formerly implemented by the Social Welfare Fund but suspended following the outbreak of conflict.[15] Maintaining national systems during crises by channelling World Bank funds to humanitarian agencies in this way offers several benefits: it may help to relaunch social protection systems once the conflict is over; it gives development partners the opportunity to learn from humanitarian partners on integrating ‘do no harm’ principles in their approaches; and it helps foster partnerships between development and humanitarian actors that will be beneficial in a post-crisis context when it comes to rebuilding the social protection system.[16]

Non-governmental service delivery remains dominated by international actors (UN agencies and international NGOs); local NGOs, local businesses[17] and other community representatives are not adequately engaged, funded or supported to play a leading role in the protracted crises contexts investigated in this study. Of the almost US$31 billion of developmental ODA that was channelled to NGOs and civil society in protracted crisis countries between 2010 and 2019, only 11% went to developing country-based NGOs; however, this data should be interpreted with caution as some funding to national NGOs that operate internationally might be excluded. There are some notable exceptions, for example the Bangladeshi-founded organisation BRAC, which has grown into a global player and has a major role in running the refugee camps in Cox’s Bazar, and extensive rural development programmes that incorporate disaster management. However, overall on the humanitarian side progress with the localisation agenda has been limited, and although there have been important pilot efforts, such as Start Fund Bangladesh, these remain small-scale pilots. Analysis of humanitarian funding in 2019 indicated that only 2.1% (US$444 million) was directed to local and national actors, decreasing from 3.5% (US$782 million) in 2018.[18] Most humanitarian funding to local NGOs is channelled through UN agencies, pooled funds and international NGOs. On the development side, some development partners have instruments that can directly support local civil society actors, particularly in the context of efforts to promote governance and accountability, such as the global European Instrument for Democracy and Human Rights,[19] which was used in Cameroon. Local NGOs identified multiple barriers in accessing greater funding, including difficulties in meeting onerous due diligence, monitoring and reporting requirements, inability to retain skilled staff by paying salaries at a similar level to international NGOs, and lack of support for overhead costs holding back organisational development. As a result of information asymmetries, a culture of risk aversion and limited administrative capacity to manage broad grant portfolios or fund local NGOs directly, there is often a reluctance on the part of donors to channel large amounts of money via ‘unknown’ entities and a preference for working with well-established partners.[20] Some donors also cited concerns that local NGO actors may have weaker governance and financial controls and can lack political neutrality or be vulnerable to corruption or other abuses. Overall, the case studies reinforce the findings of numerous other reviews that existing financing channels tend to maintain a power imbalance between international and local partners and do not provide the necessary quality technical, capacity-building and financial support to further the localisation of humanitarian and development efforts, with some exceptions such as pooled funds.[21]

Share section

Key questions and considerations for development actors

  • Partnership strategies in protracted crisis contexts need to be informed by regular context analysis to be able to identify potential risks and respond to changes in the context. How can donors ensure their multi-year partnership strategies are reviewed regularly and supported by mechanisms that provide more flexible, multi-year support to non-governmental partners including UN agencies and international NGOs, as well as instruments with a specific focus on local civil society (discussed further below)?
  • How can development partners better utilise the capacities of their diplomatic and political representations to redouble efforts to reach a political settlement to conflict where the government is a party to conflict or where there are political obstacles to achieving sustainable solutions to crises? In such contexts, dialogue and information sharing across the HDP nexus must be strengthened to ensure that the actions of one set of actors do not undermine those of another – even if collaboration may not be possible at the operational level due to the need to safeguard humanitarian space.
  • How can work to strengthen local governments in crisis-affected regions be deepened, while also engaging a broad range of community stakeholders in planning? This could offer development actors a way to address the ‘missing middle’ between top-level policy and institutional support and community-based approaches in order to scale-up delivery at the subnational level in contexts where government structures are weak and promote inclusiveness to foster long-term peace. Support to decentralisation can be an important aspect of such an approach, as can long-term investments in local government capacity through local governance programmes and inclusive area-based approaches that engage local authorities and a broader range of local actors in planning.
  • Centrally driven policy and institutional reform and partnership with the government needs to be balanced with efforts to work with and strengthen the capacity of local civil society and private sector actors in crisis-affected regions. In considering how to strengthen local partnerships beyond the government, scaling up long-term investment in local civil society and the private sector is critical. How can HDP actors work together to bring about a change in the way local actors are financed and supported? This could include collaborating to develop plans and intermediary funding mechanisms[22] – such as pooled funds or NGO consortia − to accelerate financial and technical support directly to local civil society actors in ways that mitigate donors’ concerns.


  • 3
    OECD research has found that global-level preferences are more important for many donors than country-level considerations such as “analysis of the political economy, absorption capacity and development trajectories” in determining whether ODA is provided through budget support, sectoral aid, projects under direct management or another modality. See OECD, 2020. Fit for Fragility: Practice to Policy. Available at:
    Return to source text