Humanitarian and development programming has largely been carried out in parallel, focusing on refugee and host communities, respectively. Most humanitarian assistance focuses on the refugees who are hosted in the camps. As outlined above, the JRP to the Rohingya Humanitarian Crisis includes a strategic objective on support to host communities in the two sub-districts surrounding the camps. Following its revisions due to Covid-19, it targeted a greater number of people in the host community (953,000 people) than refugees (860,000 people). Still, meeting the different needs of the host and refugee communities remain a challenge (Figure 9). As a result, shorter term humanitarian relief is insufficient to sustainably address the structural development needs of the host community. Development assistance provided by MDBs and a number of government donors (see the ‘International financing landscape’ sub-section) seeks to address this by improving infrastructure, local government capacity for public service delivery, and social safety nets, but it largely takes place outside of the JRP. The delay in the approval of the DDGP and a lack of subnational reporting on development activities make it difficult to integrate programming in Cox’s Bazar district across the nexus.

In the absence of a joint and long-term strategy on integrated programming in the district, joint programming consolidates donors’ and implementers’ efforts at the activity level to strengthen collective impact. It can also formalise cooperation across sectors and across humanitarian and development agencies, as seen in the Safe Access to Fuel and Energy Plus Livelihoods (SAFE Plus) project launched by the FAO, IOM and WFP (Box 3). This project implements an integrated response to fuel and livelihood needs for refugee and host communities, while reversing environmental degradation. Joint programming is an effective way to prevent implementing agencies in the same location from overlapping in their activities, potentially outside their remit. Several interviewees reported that due to a lack of understanding of who is operating in Cox’s Bazar district in different sectors, some development donors require their implementing partners to provide a ‘complete’ package of activities – for instance, to expand to the distribution of food packages or the provision of agricultural inputs. These demands disincentivise cooperation between agencies; while individual agencies are able to secure a greater amount of funding for multi-sectoral activities, it can increase the number of actors in each sector (with some operating outside their recognised area of expertise) and thereby lead to a fragmented response. Multi-donor funding to consortia of implementers – which can include the UN, international NGOs and national and local NGOs, alongside government departments – consolidates funding and incentivises cooperation between implementers according to their sectoral expertise.

Restrictions on longer term assistance to refugees are an obstacle to achieving durable solutions for displaced people. Activities involving cash transfers and skills development that are carried out in the camps are limited in their scope and each have to be negotiated with the government. These initiatives are not sufficient for Rohingya refugees to become independent from humanitarian relief. There is not enough space in the densely populated camps for the refugees to plant enough crops to be self-sufficient, hence they indefinitely rely on food distributions. Actors such as the UN and MDBs are working with the government to find sustainable solutions for different sectoral needs (e.g. water supply, electricity, shelter, infrastructure), but progress on these issues is slow given the continued emphasis on repatriation. The government’s approval of piloting the Myanmar curriculum for secondary Rohingya students is, however, a sign of such progress in the education sector. Conversations around expanding the type of assistance provided to the Rohingya refugees are politically sensitive, as the perceived imbalance between support provided to refugee and host communities already causes tensions.

Several development programmes in Cox’s Bazar district seek to improve social cohesion, but there is no shared understanding of how to measure success. Actors in Cox’s Bazar district attempt to alleviate tensions between host and refugee communities in two ways: prioritising the improvement of intercommunal relationships, and addressing host communities’ needs and thereby alleviating tensions created by sentiments of unfair distribution of resources. As shown in the ‘International financing landscape’ section, the volume of direct financial support to programmes that primarily target social cohesion in Cox’s Bazar district seems small compared with total humanitarian and development assistance. An example of such programmes is Canada’s funding to the Global Network for Women Peacebuilders to empower Rohingya refugees, Bangladesh and Burmese women, and youths to lead and participate in community-based peacebuilding.^[5] Similarly, USAID funds an education and employment programme targeting host community youths, aiming to reduce aggression and increase empathy towards refugee youths. It is much more difficult to quantify or even identify the effects on social cohesion of other development assistance provided to host communities. Programmes in the district might mention social cohesion in the project plans as an overarching ambition, but they do not contain indicators of success because it is not the primary objective. Interviewees shared that there is a recognition in Cox’s Bazar district that despite most development actors’ desire to improve social cohesion there is no shared understanding of the concept or how to measure improvements. In the absence of this agreement there is a risk that many actors can proclaim to have contributed to social cohesion and the nexus in different ways, without concrete evidence of their impacts on intercommunal relationships. Research on social cohesion in Cox’s Bazar district is emerging through efforts of Brac University in Dhaka, for example,^[6] but it has yet to form part of most programming approaches. Within the new phase of the SAFE Plus project (Box 3), FAO, IOM and WFP are already in the process of developing indicators to measure the project’s impact on social cohesion.

Over the last several decades, Bangladesh has moved from focusing on emergency responses to longer term disaster management and DRR. The range of programmes contributing to this process is broad, including the construction of coastal embankments, multi-purpose cyclone shelters or training on disaster-resilient agriculture. UNDP also advises the private sector on where to direct investment taking into account disaster risk. The government is proactive on disaster management through its policies and budgetary allocations as outlined above, providing a framework for development actors’ activities. Challenges remain, however, around infrastructure projects and urbanisation, which may increase disaster risk and vulnerability. Among the factors contributing to the severe monsoon floods in 2020 were the construction of roads across flood plains and changes to land cover associated with agriculture and urbanisation.^[7] In its National Plan for Disaster Management 2021–2025 the government recognises the importance of disaster risk-sensitive development,^[8] but this has yet to be integrated beyond targeted DRR efforts with all relevant development programming. This includes both an assessment of the disaster risks that might affect implementation over the course of the programme and ensuring that the programme itself does not amplify existing risks or create new ones. Humanitarian actors in Bangladesh already include considerations of disaster risk in their contingency planning.^[9] The UN Office for Disaster Risk Reduction together with OCHA, the UNRCO and the Ministry of Disaster Management and Relief started a joint project in 2020 to consolidate information systems on disasters, seeking to join up data on DRR and humanitarian responses. Early reflections from this pilot show that multi-mandate implementers of development and humanitarian assistance – including the UN and local and national NGOs – have already been engaged in DRR and are open to incorporating it more in their programming. Humanitarian and DRR actors in Bangladesh that have more experience with risk assessments can support implementing agencies of development activities with risk-sensitive planning and programming; it is unclear to what extent this knowledge transfer already takes place. An example of knowledge sharing around disaster risk is the Land Slide Early Warning System recently developed by FAO in Bangladesh, in cooperation with IOM and the Ministry of Environment, Forest and Climate Change. This system includes a landslide susceptibility map of Cox’s Bazar district, which enables risk-informed development planning and will warn relevant stakeholders of impending landslides to allow for early action. Some development donors (e.g. JICA) already mainstream DRR throughout their own development portfolios and might be able to share their learnings and best practice with other bilateral donors.

Successful anticipatory action pilots in the 2020 humanitarian monsoon response provided an opportunity for development actors to scale up and better integrate disaster preparedness, response and recovery. Since 2015, the Red Cross and Red Crescent Movement and WFP have tested and developed forecast-based financing in Bangladesh. Additional actors have recently progressed this agenda – including Start Network, CARE Bangladesh, Concern Worldwide, Islamic Relief Worldwide and others – and implemented an anticipatory response to the 2020 monsoon floods. In light of this pre-existing engagement on forecast-based financing in Bangladesh, the country was chosen by OCHA in 2020 as one of five anticipatory action pilots for the UN OCHA’s Central Emergency Response Fund (CERF) (Box 4). Early evidence shows the pilot was successful in enabling a faster, more cost-efficient and effective response while protecting development gains.^[10] Still, OCHA notes that the pilot was relatively small compared with overall needs for disaster response in Bangladesh.^[11] This raises questions on how to further scale up funding for anticipatory action, given the risk of diverting limited humanitarian funds from existing needs.^[12] Even though the anticipatory activities were implemented as part of the humanitarian response to the 2020 monsoon floods and were financed from humanitarian funds, they provided a possible entry point for development actors. Those currently engaged in disaster management can learn from CERF’s experience to establish similar funding mechanisms to support anticipatory action. Expanding the donor base would allow anticipatory action to scale up quickly and reduce the risk of diverting scarce humanitarian resources. Another argument in favour of involving development actors is the need for a longer term process to engage relevant government agencies and thereby support co-ownership, requiring sustained technical support and financing that humanitarian funds are unlikely to provide. Given the ability of anticipatory action to protect development gains, in particular by supporting agriculture and livelihoods, its scale up could also support a faster recovery following disasters.

While progress has been made in strengthening national social protection systems, with renewed efforts in response to the Covid-19 pandemic, they are not yet comprehensive in their coverage of the most vulnerable. As outlined in the ‘Domestic public resources’ sub-section, there are over 100 social protection programmes in Bangladesh across more than 20 line ministries. These target the poorest households and various other vulnerable groups, such as the elderly, widows and people with disabilities. Issues on targeting and coverage for these social safety nets were identified above (see ‘Domestic public resources’ sub-section). International development donors, including the World Bank, ADB, the UK and the EU have supported the government to strengthen national social protection systems over the last decade. In the context of Covid-19, these efforts were scaled up significantly to expand social assistance to the most affected people. This includes workers – in particular women – in export-oriented industries,^[17] small and medium enterprises, workers in the informal sector^[18] and the poorest and most vulnerable households. Even though Rohingya refugees are not targeted by national social safety nets, the World Bank approved an additional financing project to address the socioeconomic resilience of vulnerable host and Rohingya refugee communities. This engaged both communities in community services and cash-for-work schemes and has targeted 85,000 Rohingya refugees and so far supported 175,000 vulnerable households across all its components.^[19] The various initiatives to scale up social protection provide an opportunity to expand assistance to more people and improve delivery systems. As the Country Director of ADB in Bangladesh Manmohan Parkash says, “in a post Covid-19 era, the social protection schemes can be mainstreamed, with increased budget allocation, wider coverage, better targeting, and strengthened administration and effective delivery to the beneficiaries using digital technologies”.^[20] As identified above in the ‘Domestic public resources’ subsection, existing issues with targeting and coverage of social safety nets are an obstacle to reaching people most affected by Covid-19 despite scaled-up international assistance. The speed of recovery for the most vulnerable people in Bangladesh from the socioeconomic impacts of the pandemic will reveal how well the government is able to fill gaps with support from development actors.

A wide range of ‘non-essential’ development programming was put on hold by the government’s regulatory efforts to contain the spread of the pandemic, exacerbating vulnerabilities. Initially, this included all programming except for lifesaving activities, such as humanitarian assistance provided to Rohingya refugees, in the aftermath of natural hazards or as part of the pandemic response. For instance, most education and skills development initiatives in the refugee camps were paused, even though the Government of Bangladesh allowed (after some delays) the introduction of home-based, caregiver-led education to prevent further learning loss due to Covid-19. According to the ISCG, this caused a shortfall in informal education and skills development for women and girls in particular.^[21] Piloting the Myanmar curriculum with Rohingya secondary students, approved in early 2020, was also delayed due to Covid-19. Social cohesion between host and refugee communities faced a double burden as socioeconomic repercussions of the pandemic for vulnerable Bangladeshi households heightened tensions and community engagement programmes to sustain peace had to be put on hold. Agricultural training and the provision of inputs (e.g. as part of the SAFE Plus programme) were also disrupted. Development actors had to convince the government that resuming support to food production should be considered an essential function to prevent future food insecurity. The Covid-19 pandemic and related regulations therefore point to the close relationship between current development efforts and potential future humanitarian needs.