Introduction

The humanitarian sector–which comprises varied international non-governmental organisations (NGOs), local NGOs, UN agencies, community-based actors and civil society more broadly—has often been critiqued for being bureaucratic, hierarchical, and lacking accountability to affected populations [1,2]. Fassin argues that inequality is embedded within the humanitarian sector, as it is within Western societies more generally [3], and proposes that a “hierarchy of humanity” drives the “asymmetry” between those who give and receive aid and between international/foreign actors and local/national actors [4]. His framing of power dynamics has been used to understand the power of international agencies within the humanitarian response, which often assumes that international actors alone hold technical knowledge [5,6]. This is manifest, for example, in how local staff often work on the frontline while expatriates hold management and decision-making roles [7]. Other asymmetries are apparent in seniority [8] and gender, with cultural and social norms placing women lower in organizational hierarchies [9]. Gender, race, nationality, and seniority intersect to exclude certain groups from decision-making [10]. In recent years, since the World Humanitarian Summit in 2016 in particular, discussions about power dynamics in the humanitarian sector have become more prescient as part of the push towards localising and decolonising humanitarian aid [11,12].

In this paper, we explore the power dynamics within a humanitarian coordination mechanism responsible for mental health and psychosocial support (MHPSS)–the MHPSS Taskforce in Lebanon. MHPSS is defined as “… any type of local or outside support that aims to protect or promote psychosocial well-being and/or prevent or treat mental disorders” [13]. People affected by forced displacement experience high levels of psychosocial distress, and while often overlooked in the past, MHPSS services are increasingly recognised as essential elements of the humanitarian response. Since 2005, the humanitarian response has been organised in clusters, such as health, protection, and shelter [14], but MHPSS is considered a cross-cutting issue that should be integrated across all clusters. It is recommended that MHPSS responses are coordinated in working groups or taskforces, which are intended to bring together relevant clusters and ensure MHPSS considerations and activities are incorporated at all levels and sectors. While this positioning is intentional, there has been a critique of this structure leading to a lack of dedicated focus on MHPSS, additional challenges in coordination, and side-lining of MHPSS in favour of other cluster priorities [15,16].

Our study focuses on power dynamics within the MHPSS Taskforce in Lebanon and responds to growing critiques of humanitarian power dynamics described above and the increased demand for decolonising the systems and structures that shape humanitarian responses. To date, as far as we know, this is among the first study to focus solely on power dynamics within a humanitarian coordination body. Thus, it helps fill a gap as there is currently a lack of clear guidance to inform best practice in decision-making within coordination groups in this sector, although recently published guidance from the UN Inter-Agency Standing Committee (IASC) on MHPSS coordination does stress the importance of government co-leading coordination as well as building consensus within coordination groups [15]. Specifically, there is a lack of empirical evidence and guidance on how humanitarian actors should consider power hierarchies that operate within coordination bodies, despite the cluster system and its working groups/taskforces being common to humanitarian responses. As such, this study represents a new effort to explore how the power hierarchies within coordination bodies, including those that are nationally led, may influence the effectiveness and inclusivity of MHPSS responses. In one published case study of the MHPSS Working Group responding to the Rohingya refugee crisis in Bangladesh, participants in the national working group were mostly senior staff without direct involvement in service provision, so information and decisions were not always well communicated to field staff and meetings were less informed by field realities [16]. The authors also noted that the group’s outputs tended to focus on developing national technical guidance documents, strengthening capacity, and conducting assessments, with local issues being neglected [16]. They described the need for MHPSS coordination groups to involve community-based actors who know the context best and to establish field-based coordination structures with clear reciprocal links to national efforts—either through national representatives attending field-based meetings or vice-versa. These examples illustrate how power dynamics inherent in the humanitarian sector, including within coordination bodies, might limit the extent to which local/field realities are considered. There is a clear need to explore further how power dynamics inherent in the humanitarian architecture might undermine efforts to decolonise and localise humanitarian responses and how best to mitigate this.

Methods

We conducted semi-structured interviews with members of the MHPSS Taskforce and other Lebanese stakeholders between February and November 2021. We used Zoom due to the COVID-19 pandemic. We also collected background information from documents related to the Taskforce, specifically minutes from meetings and the 2016 evaluation.

While the interviews also examined barriers and enablers to the work of the Taskforce as part of the broader study, the material presented here focuses on power dynamics, responding to the overarching research question: how do power relations influence the practice of the MHPSS Taskforce in Lebanon? Topic guides (annexed) were informed by Gaventa’s power cube and Arnstein’s ladder of participation described above [22,23,27] and were developed in consultation with GOAL project partners, who included the National Mental Health Programme. We piloted the guides during workshops to ensure they fully reflected the issues being explored, that questions were appropriately phrased, and that English and Arabic translations were suitable.

Thirty-four (34) participants (28 women and 6 men) participated in semi-structured interviews. Participants were recruited using a list provided by the National Mental Health Programme. They were either members of the Taskforce (n = 30) or had worked closely with it (n = 4), representing a range of actors working on MHPSS in Lebanon. The gender composition of the participants reflects the fact that the majority of Taskforce members are women. They included local non-governmental organisations (NGOs) (n = 10) and international NGOs (n = 11), UN agencies (n = 8), governmental organizations such as the Ministry of Public Health (MoPH) and other ministries (n = 4) and consultants (n = 1). The majority of participants were Lebanese. We have labelled the quotes we use to designate the type of organisation of the participant only to preserve anonymity.

Interviews were conducted by the research team (made up primarily of TZ, RM, RA and BM, all based in Lebanon) in either English or Arabic, depending on the participant’s preference. They were audio recorded and transcribed verbatim into English by RM, RA, BM, other team members and external translators. Other research team members checked transcriptions for accuracy; ten per cent of the transcripts were back-translated to ensure the accuracy of the Arabic-English translation. Transcripts were then anonymised and uploaded into Dedoose, a cross-platform application for qualitative data analysis.

As part of our commitment to co-production, we adopted a collaborative analysis and writing process for this paper. The interviews were coded collaboratively in pairs by the research team (ML, TZ, RM, RA, BM, SC) based on a codebook and using the blind coding feature offered by Dedoose. We used deductive and inductive coding, meeting bilaterally in pairs and regularly in a group to discuss codes and share perspectives on the data. The process of collaborative coding is described in more detail elsewhere [31]. While Gaventa’s power cube and Arnstein’s ladder of participation influenced the development of topic guides, during analysis, we found that participant accounts closely matched Arnstein’s ladder but did not neatly map onto the power cube. As such, we adopted an inductive approach when analysing those parts of the data that were unrelated to participation. However, we draw on the power cube and other power theories where relevant in the Discussion section.

We also reflected on our positionality and power throughout the research process, including during analysis, as part of our approach to collaborative coding and analysis. All of our analysis team members were part of an international NGO in Lebanon, except for ML, who was part of an academic institution in the UK. As a team, we discussed how gender, education level, ethnicity, age, experience in the humanitarian sector and geographical location particularly affected how we engaged with the data as researchers and how we perceived participant reflections on power relations in the Taskforce. We also reflected on the complexities of researching a coordination group where leaders of the group were also involved in framing the study. Our study is also complicated by the positionalities of participants, whose views about power may be based on their own positioning in the humanitarian system, especially given the increased recognition of power hierarchies in the humanitarian sector in recent years [4–8].

We developed a list of key themes from the coded content from transcripts and used them to draft the paper in a small group involving ML, TZ, RM, SC, RA and BM. The initial themes were shared with small groups of 16 participants who agreed to join feedback sessions in August 2022. Two of the seven feedback sessions were held with the Taskforce leadership team. The feedback is reflected in the findings, demonstrating how interview participants were involved in the analysis process.

Findings

In seeking to understand how power relations may influence the practice of the MHPSS Taskforce in Lebanon, the following paragraphs outline the main themes related to power dynamics. We outline three key findings: 1) UN agencies and international NGOs are perceived to have more power and influence in decision-making in the Taskforce; 2) seniority, expatriate-local dynamics, language, gender, and the focus and format of meetings shape power dynamics to varying degrees, and thus influence voice and decision-making in Taskforce meetings; and 3) levels of participation of members in the Taskforce vary. In the first and second sections, we have presented findings based on the themes that emerged through our coding and analysis process rather than those based on existing theories of power for the reasons explained above. For the third finding on participation, we explicitly structure findings based on Arnstein’s ladder of participation as this framework aligned closely with the data emerging from participants.

Discussion

Power dynamics are present in any decision-making or coordination mechanism but are especially important for the field of humanitarian aid, where power asymmetries often place “international” actors with funding in positions of influence. As outlined earlier in this paper, the cluster system and humanitarian architecture more broadly create several challenges for the humanitarian response, especially in a setting like Lebanon. In recent years, alongside the increased recognition of the need to localise and decolonise aid, there has been greater attention to the challenges of these power dynamics within the humanitarian sector. This study is among the first to rigorously explore how these power dynamics play out and influence practice within a coordinating body through a case study of the MHPSS Taskforce in Lebanon. The MHPSS Taskforce is a unique example in the humanitarian sector because it is the only nationally-led coordination group in Lebanon–representing a good example of efforts to localise. Our findings show that many of the power dynamics in this Taskforce reflect hierarchies inherent within the broader humanitarian system, with implications for similar coordination groups within and beyond Lebanon. We discuss the implications of each key finding below: the role of funding and power, the dynamics between international and local actors, the dynamics between senior and junior actors, the clinical versus PSS hierarchies, and participation and decision-making.

In our study, the perception of NGOs is that UN actors and international NGOs have greater voice and influence in decision-making within the Taskforce than national/local organisations was tied to their access to funding (e.g. unlike the NMHP which does not have funding) and credibility within the humanitarian system. This perception is not shared by the NMHP leadership in its relationships with the co-chairs. The relatively greater decision-making power of UN and international NGO actors than local actors reflects global levels of power in decision-making outside the state, as conceptualized in Gaventa’s power cube [22,23]. Barnett and Duvall’s taxonomy of power (developed for international relations but also applicable here) is also useful in framing funding and credibility as being structural–built into the system [32]. In reflecting on the implications of these findings, we also find Brandenburger & Nalebuff’s concept of “coopetition” important in thinking about how perceptions of power of different actors might be connected to competition and lack of trust in the humanitarian system. Coopetition describes the simultaneous presence of cooperation and competition [33]. In humanitarian emergencies, different organisations are required to develop trust with each other quickly for collaboration [34], while competing with each other for funding and attention [34,35]. Research in humanitarian settings has found both relationships of trust and mistrust among different organisations that are required to collaborate within a competitive environment [36].

Our finding about funding power aligns with existing critiques of the humanitarian funding structure that disproportionately directs funds towards international rather than local actors [37,38]. The unique aspect in Lebanon is that that the Taskforce is nationally-led, representing a new way of coordinating the humanitarian response. However, if local actors lack funding and human resources, as global literature suggests, they may have less influence on humanitarian decision-making. This is particularly relevant in the case of the Taskforce since lack of funding for the MHPSS Taskforce and action plans has been a major challenge. Despite this, international NGOs and UN agencies were still perceived to have more power in decision-making than other national organizations who are members of the Taskforce, even though the government, a local actor, is a co-chair. Despite the push towards localising funding in the broader humanitarian sector, this remains an area where little progress has been made [39], with implications not only for the MHPSS Taskforce but other decision-making bodies within the humanitarian system [38]. We also find more clarity may be required for Taskforce members on how the group operates and makes decisions to help tackle perceptions about decision-making processes within the Taskforce. Clearer communication may create opportunities for members to feel like their participation is more than being informed or consulted but includes opportunities higher on Arnstein’s ladder of participation, such as partnership [27].

The findings in the paper that junior actors tend to share field experiences, while senior actors are more engaged in decision-making on behalf of their organizations builds on existing studies [18] and has important implications for coordination in Lebanon and the humanitarian system. In the humanitarian system, decisions about who attends meetings occur at the organisational level and could be influenced by multiple factors, including politics and time availability. Additionally, the potential for hierarchies between local humanitarian workers and “expats”, which Pascucci found in her research in Lebanon and Jordan [7] and which has been found in other settings to privilege expatriate voices [10,38,40], adds another layer to the issue of seniority. Being in a more senior position of authority in an organisation and being an expatriate may provide additional privilege, potentially leading to local knowledge and expertise being undervalued in favour of outsider perspectives. However, our findings complicate existing studies on these power dynamics, as some expatriates have worked in Lebanon long enough to have an appropriate grasp of the context, resulting in a continuum rather than local and expatriate being set categories. Local-expatriate dynamics between individuals are also complicated by the NMHP (as chair of the Taskforce) being a local actor (a more unusual but recommended setup for coordination groups). Thus it is challenging to draw comparisons to other contexts, however the lack of funding power of the NMHP may lessen this influence. Reflecting on the findings about local-expatriate power dynamics, we also recognise that for the response to Syrian refugees, the actors are rarely local or from the same community (i.e. Syrian) but might most often be Lebanese. These nuances emphasise the importance of ensuring the participation of a range of voices–including Syrian service users—to mitigate these power hierarchies in decision-making. It is important that efforts to engage service users be meaningful and intentional, per Arnstein’s ladder of participation [27]. Embedded within the issues related to local-expatriate dynamics, our research also identified the power or “linguistic capital” [6] channelled through the use of the English language in meetings. We find Barnett and Duvall’s taxonomy of power helpful in framing the use of English as institutional power–part of how the humanitarian system operates [32]. Using the English language in humanitarian coordination meetings may exclude local actors [41,42]. Running Taskforce meetings in Arabic with English translation may represent an opportunity to shift power dynamics within the Taskforce, however this requires further discussion within the Taskforce. As these power dynamics stem from the overall structure of the humanitarian system, these findings have implications for other coordinating groups within the humanitarian system, specifically other clusters.

Our study also highlights opportunities to shift power through clearer communication about how decision-making occurs in the Taskforce. The perception that decisions are made by Taskforce leadership in a “closed” space through bilateral meetings and a “core group” may need to be corrected. However, drawing on Gaventa’s framework [23], “invited” spaces still exist to negotiate these decisions through information-sharing with members and member consultation on decisions and plans. Although some Taskforce members felt that final decisions occurred less transparently and that the use of feedback could be shared more clearly, the onus is not only on leadership but also on members to respond to requests for inputs and to contribute to tasks. More clarity may be needed for members on internal decision-making within the Taskforce.

In our study, the NMHP played an important role in influencing the Taskforce’s agenda, and as chairs, may hold what Gaventa’s power cube calls the most ‘visible’ power in public decision-making spaces, such as within Taskforce meetings [23]. This is not necessarily negative but has implications for who is perceived as holding authority. The discussion about the NMHP’s role included the perception among participants that the Taskforce emphasises clinical/medical aspects of mental health over more community-based psychosocial services, despite the emphasis on community-based services in Taskforce action plans. Psychosocial services were perceived as more of a focus for local and community actors, while mental health was seen as having a larger focus in international NGOs and UN organizations, perhaps reflecting the power dynamics discussed earlier. Psychosocial support services are also covered by the PSS Working Group, another body in the humanitarian coordination architecture in Lebanon, which may add to the confusion around the role of the Taskforce in this regard. This perception about the prioritisation of more medical or specialised services over psychosocial support in the Taskforce is different to the action plans and outputs of the Taskforce–indicating more work needs to be done with members to clarify the Taskforce position on this, given this has potential to create divisions in MHPSS coordinating groups as humanitarian actors may differ in their prioritisation of clinical or psychosocial services [43]. Despite MHPSS being intentionally integrated across multiple sectors globally in order to recognise the different levels and forms of support needed to adequately respond to the needs of the population, and despite the local mission of the Taskforce being to connect actors across sectors, the positioning of the coordination of the taskforce within the Ministry of Health may mean that extra efforts are needed to ensure that actors in non-health sectors feel equally engaged and supported. Miller and colleagues argue that clinical and psychosocial approaches are not inherently incompatible; instead, they address interconnected concerns and hold great potential for synergy when integrated into multilevel, multi-sectoral interventions [43]. We suggest that the Taskforce could engage in discussions with members to strengthen the harmonisation of these approaches.

Our participants emphasised how established relationships with Taskforce leaders may allow more confident engagement in the Taskforce. This aligns with the literature indicating that investing in high-quality relationships can contribute to better programme outcomes [44]. Such relationships with Taskforce leadership may make participation easier but can conversely make it challenging for newer actors. Staff turnover, a significant issue in humanitarian organisations [45,46], also poses challenges to efforts to influence and engage in bodies such as the MHPSS Taskforce. When staff holding strategic relationships move to other organisations or postings, the loss of institutional memory contained within these relationships may stall momentum.

Conclusion

This study highlights insights into how power dynamics influence decision-making in a humanitarian coordination body in Lebanon. Our findings suggest potential implications for how global humanitarian governance structures operate across multiple settings because the power associated with access to funding remains recurrent throughout the humanitarian sector and is not unique to the response in Lebanon. Despite efforts to localise humanitarian responses, funding remains disproportionately directed through international actors, which has ripple effects on power dynamics within the sector. Our study suggests a need to define how decision-making occurs within the Taskforce more clearly. As a nationally-led coordination body, the MHPSS Taskforce in Lebanon represents a new approach to humanitarian coordination. We also suggest that the MHPSS Taskforce should continue to be aware of power dynamics that endow senior or international actors with decision-making power and should intentionally seek to centre the perspective of junior and local actors who are often closest to the humanitarian response at the field level, for example through re-instating the regional Taskforce meetings. Balancing the need to hear from those most affected by issues–in this case, mental health service users—with the need to leverage decision-making authority is an ongoing challenge in the humanitarian sector. Engaging persons with lived experience in different modalities, including through the newly established Service User Association in Lebanon, may help ensure service users are consistently included. Many of the key issues identified in our study fundamentally stem from the structure of the humanitarian system, and further research on power dynamics (particularly within coordination groups) is required to set benchmarks for strengthening participation in decision-making and improving the effectiveness of MHPSS responses. Future research could also reflect on power dynamics more broadly in the humanitarian sector to contribute to ongoing momentum on the need to localize and decolonise aid.

Supporting information

Acknowledgments

We are grateful to Felicity Brown and Prudence Brown for their feedback on earlier drafts of this paper.