The Worth of Their Work: The (In)visible Value of Refugee Volunteers in the Transnational Humanitarian Aid Sector

Introduction

This article focuses on the volunteer work that Syrian refugees perform in the humanitarian aid sector in Jordan. Specifically, this article investigates the peculiar position of refugee volunteers who explicitly identified and described what they did as labour, but hesitated to explain their relationships with aid organisations as ‘work’. What explains this ambiguity? What processes delineate refugee volunteers’ labour from work in certain instances but not others?

This article contributes to debates on invisible work (Daniels, 1987; Hatton, 2017, 2020) in terms of how labour is devalued and obfuscated in practice. Scholarship on invisible work is rich in thick description, but is less abundant in its development of this as an analytical category ‘to operationalise what “invisibility” in the world of work means’ and its effects (Hatton, 2017: 337). Hatton (2017) provides an important framework in this regard in outlining three sociological mechanisms that produce invisible work because they economically devalue labour through cultural, legal and spatial means. This framework suggests that volunteer work is rendered invisible through sociolegal mechanisms that frame it as noneconomic (e.g. acts of love and service, education and training). This article adds to this framework by showing how volunteer work is also rendered invisible when it is reframed as aid: labour is transformed into an object refugee volunteers ‘receive’ and ‘consume’ from aid organisations as a benefit because ‘work’ is understood as something they lack or need.

While scholars have significantly broadened conceptualisations of work to encompass paid and unpaid labour in both private and public realms over the past decades, scholarship on volunteering within and among populations and contexts located in the so-called Global South is less elaborate. This is surprising because underpinning the expansion of the concept of work to include multiple forms of labour is an ever-growing recognition of, and emphasis on, the notion of work as socially produced: that is, work as a relational and context-specific activity. How aid acts as a vehicle for redefining the boundaries of what work means and can entail in this present case therefore has critical implications, particularly in how it reconfigures and produces social inequalities through the global division of labour.

The layout of this article is as follows. I start with a brief review of the literature on invisible work, and describe how exploring aid sector labour is an opportunity to elaborate at least two dimensions of this scholarship: first, understanding processes that render volunteer work invisible; and secondly, the relationship between volunteering and invisible work in Global South contexts. The following section then provides details about the Jordan case, including the methodology used to conduct the research that informs this article. I then present the findings that show how refugee volunteers’ labour is rendered invisible when it is reframed as aid and its initial effects on labour relations in the aid sector as a result. The article then concludes with a discussion regarding the implications of this study for future research.

What is invisible work?

Invisible work is understood to constitute various forms of labour that are devalued because they are reframed as something other than work (Hatton, 2017, 2020) The concept was first used as a term to make sense of women’s unpaid labour performed in the private, domestic sphere in reaction to early sociology of work research that primarily conceptualised ‘work’ as paid labour (e.g. wage) in the public domain given ‘academic concerns with industrial capital. . .at the beginning of the twentieth century’ (Taylor, 2004: 30; see also Daniels, 1987). Labour that fell outside this definitional scope, like domestic work (e.g. unpaid, private), was overlooked in explanations of work and its effects on society. Invisible work is now used to describe multiple layers of additional labour performed in private and public spaces, with or without pay, that goes unrecognised and ‘unseen’ in various sectors such as retail, medical services, and digital labour (Allen, 2014; Gruys, 2012; Whiting and Symon, 2020). The concept of invisible work has therefore performed an important role in nuancing and expanding the conceptual boundaries of ‘work’ to include various forms of labour that produce value for others in the form of services or goods (for an overview and examples of these various forms of ‘work’, see Taylor, 2004: 41; also see Budd, 2011). Invisible work has also clarified that ‘work’ is not synonymous with pay or reducible to ‘employment’, and includes unpaid labour. In so doing, the concept highlights work as ‘embedded in and defined by the social relations within which it is located’ (Glucksmann, 2005; Taylor, 2004: 31).

Methods

I collected the data presented in this article as part of a larger research project on how localisation shifts in the sector were affecting the organisation of labour, and more specifically, the roles of local workers. With limited scholarship and data available on the labour process in the aid sector, particularly from the standpoint of local labour, this broader project was designed as a qualitative one to develop ‘[a] more comprehensive theoretical picture of social life’ (Tavory and Timmermans, 2009: 246). I adopted an abductive analysis approach (similar to abductive logic advocated by Dubois and Gadde, 2002), in which one ‘search[es] for surprises in empirical observations and in theory’ by moving back and forth between findings in the data and existing theory and analytical concepts (Mears, 2017: 143; Timmermans and Tavory, 2012). I drew upon studies about the labour process and localisation in other transnational industries to design the research and generate initial hypotheses. My previous research in Jordan, which included a 2016 pilot study for this project, also informed the scope of this research.

I collected the majority of my data between August 2017 and March 2018. I conducted interviews with aid workers and ethnographic observations of their daily routines in rural and urban areas of the country, as well as in one refugee camp setting (Zaatari). Since this broader project focused on the effects of localisation on local labour, I prioritised interviews with local workers (n=75). Local workers included anyone formally employed, interning, or volunteering for an international aid organisation based on their Jordanian residency or refugee status. Refugees often had to be officially registered with the United Nations High Commissioner for Refugees (UNHCR) to be eligible to volunteer with aid organisations. All refugees residing in camps are registered, but this is not necessarily the case for refugees residing in urban areas outside of the camps.

I used a combination of key informants, snowball sampling and cold-call outreach to connect with interviewees. My key informants included workers I met through previous research in Jordan, including the 2016 pilot study. Most interviewees were between 20 to 45 years of age, and 51 percent of the local workers interviewed identified as women. Interviews lasted anywhere from 30 minutes to two hours. Each participant provided either verbal or written consent prior to the interview, including consent to be recorded, and could refrain from answering questions or continuing with the interview at any time. Interviews were conducted in Arabic and English. I transcribed them myself. Conversations often switched back and forth between the two languages, with workers often speaking in English when discussing the specificities of their work roles. I speak and understand most Arabic dialects from Jordan and Southern Syria, but relied on translators on several occasions. My position as a young, white female in a context saturated with researchers like myself from Europe and North America, surely shaped these interactions as well, including when and why volunteers switched between English and Arabic when speaking with me (see Roth, 2019 on language in aid relations). Boston University’s IRB approved the research design of this project.

Interviewees worked in various capacities, as logistics officers, monitoring and evaluation (M&E) specialists, finance administrators, case managers, drivers, as just a few examples. Over half of the workers I interviewed worked with at least two employers in the past three years (36 had worked for more than two). Explanations as to why they switched jobs varied, but included everything from short-term contracts and boredom to workplace relations and professional development. Refugee volunteers, however, often switched jobs because their positions were designed to be short-term in the first place, as this article shows.

I became aware of the roles of refugee volunteers from my interviews and observations. During a visit to one aid organisation, for example, a Jordanian aid worker asked me if I wanted to interview ‘volunteers too or just staff’. I did not explicitly ask other workers about volunteers during interviews: whether or how these other workers identified them was telling data in itself related to aid labour relations. Given the repeated reference to ‘Syrian volunteers’ during interviews, I decided to start including more information about this group in my field notes. I also interviewed 10 volunteers directly towards the end of my data collection period. The findings presented in this particular article, while primarily based on these interviews and observations, are also based on questionnaire surveys (n=10), email correspondence, and Facebook exchanges with other volunteers who I was not able to meet with and interview in person. The questionnaire surveys were abbreviated versions of interview schedules. These were sent via WhatsApp or email. The other email correspondence and Facebook exchanges were longer interactions in their overall duration, because I would often ask a few questions and receive answers, and then would follow up with more questions based on the responses. The reasoning for using these alternative methods to communicate with volunteers varied. In some cases, a meeting was not possible because of transportation, distance and my permitted access (e.g. village five in Azraq refugee camp). In other cases, volunteers shared that they were worried about losing their volunteer positions if someone found out that they talked to me about their ‘employer’. I also maintained regular contact over the course of approximately two years with two volunteers, one of whom was working in an urban location in the north and another who worked in a camp, to see the ways in which their positions as volunteers changed (or not) during that period. This article only includes data from Syrian refugee volunteers, although refugees from other countries that live in Jordan participate in aid organisations’ volunteer programmes as well. The reasoning for the latter is partially because Syrian refugees represent the majority of refugees officially registered with UNHCR in Jordan. It is also reflective of aid distribution at the time of this research, which in many cases was exclusively marked for Syrian refugees and areas hosting high numbers of them. Including other volunteers with different national backgrounds is pertinent for future research because it may provide important implications for understanding the relationship between invisible and insecure work within groups that share certain statuses (e.g. legal), but not others (e.g. nationality, ethnicity).

Data were stored and codified for the overall project in NVivo. Sociological scholarship on transnational labour as well as interdisciplinary literature on humanitarian aid and development informed the initial codes that I used to analyse my data. I also used by-hand, line-by-line coding. The emerging themes from the latter informed how I revised these codes over time. I also triangulated my data with aid reports and meeting minutes from aid organisations operating in Jordan. I further tested and revised my interview schedule on multiple occasions to account for reliability and validity in my research design, and compared my interview findings to my field notes. I determined that I reached data saturation when I recognised similar and consistent code patterns in the interview data and limited to none ‘new’ or ‘emerging’ codes. However, and as previously noted, future research on this specific topic of refugee volunteers in Jordan (and elsewhere) could be further developed and expanded to include non-Syrians’ experiences.

Findings: ‘Work’ relations, but ‘aid’ benefits

Refugee volunteers’ day-to-day proximity structurally and relationally to aid employees, often positioned them as ‘workers’ for organisations. Like national staff employees (Jordanian aid workers), refugee volunteers were responsible for documenting the impact of organisations’ projects in camp and urban settings (e.g. data entry, project evaluations). They were also on the frontline, conducting home visits with beneficiaries. Many volunteers led training on topics about livelihoods, gender-based violence (GBV) and self-development as well. In many cases, employees and volunteers completed these tasks together, relationally situating volunteers’ labour as comparable to employees’ ‘work’.

How organisations recruited volunteers for these positions further suggested that they were ‘working’ for these entities. Like a typical job, volunteers applied to participate in these programmes based on their qualifications and skills. Aid organisations disaggregated refugee applicants into categories such as semi-skilled, skilled, highly-skilled, technical, or technical specialist to determine if and in what positions they would ‘hire’ them for accordingly during that cycle. ³ Workers also signed and agreed to some form of contract with aid organisations once they received a position. Some volunteers signed official documents, whereas others talked about verbal agreements (especially in the cases for short-term training). While these contract formalities varied, the fact that vetting procedures for volunteers were similar to employees’ (e.g. application, assessment of skill, contract agreement) situated them in what appeared to be ‘working’ relationships with aid organisations.

Volunteers’ ‘earnings’ in these positions also suggested, at least partially, that volunteers’ labour was understood as economic activity and thus ‘work’. First, like wages and salaries associated with blue-collar and white-collar work, organisations compensated workers with stipend payments scaled to volunteers’ positions and qualifications. Cleaners, for example, received smaller stipends than trainers and technical coordinators. Second, pay varied according to one’s living conditions (stipends were often reported as higher outside the camp). Third, organisations emphasised these stipend payments as ‘valuable boost[s] to [refugees’] income’, ⁴ and described volunteer schemes as ‘employment programmes. . . [that] offer refugees. . .an opportunity to earn some money’. ⁵

Yet, volunteers’ labour ‘at work’ was simultaneously framed as a form of aid; a benefit they received from these organisations. While ‘payments’ to volunteers were described as ‘boosting. . .income’, organisations simultaneously referred to them as stipends, rather than salaries or wages, that supplemented transportation costs associated with participating in the programme, ⁶ for example. This money was therefore distributed and sociolegally classified as a form of ‘support’ from aid organisations that would ensure volunteers’ ‘participation’ and engagement in this ‘experience’ (i.e. work; Hatton, 2017: 341). Stipend payments therefore marked volunteers’ labour as noneconomic, and framed it as something they were able to do and access because they ‘received’ it as aid from organisations.

Similarly, contract cycles delineated volunteers’ labour as a form of aid. The average contract length for volunteer positions was approximately three months. This short-term design was intentional to limit – quite literally – volunteers’ affiliations with organisations. In order to ‘reach’ a larger target number of beneficiaries and therefore distribute as much aid as possible, organisations rotated volunteers in and out of positions. In many cases, once a person volunteered for an organisation, the individual was not eligible to participate – work – again for the same organisation for a certain time in order to give other refugees like themselves opportunities to volunteer as well. As one volunteer in the camp explained: ‘[Organisation X] has three-month volunteer positions for Syrians. [We] are not allowed to [work] longer because they say they are trying to give as many Syrians as possible an opportunity “to have this experience. . ..and to have some income”’ (Interview Z6).

These rotations, the ways in which volunteers came and went from organisations, physically marked their labour economically and socially as a form of aid that they had an opportunity to receive. As multiple volunteers explained, they were ‘numbers’ for organisations to demonstrate to donors and other aid stakeholders that they were providing a benefit (e.g. work) to those in need (and to as many individuals as possible). This meant that, unlike aid employees, whose value often equated with the labour they performed, volunteers ‘counted’ and contributed to aid organisations through their presence (their labour during their contract), but also their absence (rotating out to allow space for other beneficiaries) because their labour was considered a noneconomic activity: devalued and delineated as aid.

Even in the workplace, volunteers performed additional tasks distinct from staff colleagues because they were understood to ‘receive’ their labour as a form of aid. For example, they participated in focus groups, and provided data for evaluations and surveys; documenting how their experiences as volunteers impacted them, their families and their communities. Volunteers also described how they served as the ‘face’ of the organisation when donors or researchers visited the project site in person. One interviewee was quite telling in this regard. Before even posing a question, the volunteer began [telling] his ‘refugee story’ as he called it, rattling off details of his experience in a manner that gave the impression that he had done this more than a few times (Interview Z4). These procedures evidently required volunteers’ hypervisible labour. Yet, it was ironically rendered invisible because sociocultural assumptions undergirding their refugee statuses intersected with sociolegal mechanisms that constructed their labour as noneconomic aid they ‘need’ (Hatton, 2017: 342). Specifically, their labour was devalued and framed as an important form of emotional aid as well: opportunities aid organisations provided for them to express their hardships to a global audience.

Discussion and conclusion

These findings highlight how volunteers’ labour was rendered invisible because it was reframed as aid. Work was the benefit that volunteers ‘received’ from aid organisations because it was perceived as something they lacked and needed given their legal status as refugees in Jordan, and how it conditioned their access to formal labour markets. Framing work-as-aid delineated their labour through sociolegal mechanisms as noneconomic. This subsequently transformed the labour they performed into objects they consumed, which devalued and rendered their work invisible as a result. However, the hypervisibility of framing work as an aid benefit had further consequences because it created labour processes that produced and positioned volunteers ambiguously as worker-beneficiaries in relation to aid organisations. This ambiguous positioning created confusing relations between volunteers and aid organisations that contributed to the burgeoning scope of their invisible work as well as to their feelings of precarity.

Drawing upon this case of refugee volunteers in Jordan’s aid sector therefore elaborates understandings of how volunteering is delineated as a form of invisible work in practice through labour processes that frame and produce their labour as noneconomic. This article specifically identifies aid as a vehicle that renders work invisible in this manner, and expands empirical knowledge on volunteering in the Global South. In so doing, these findings add to Hatton’s framework, because they contribute to clarifying and defining the scope of invisible work beyond a ‘vague and variable synonym for out-of-sight, unrecognised, marginalised, and/or degraded labour’ (Hatton, 2017: 345).

This article further leverages volunteers’ hypervisible roles to complicate understandings as to how volunteer work is rendered invisible in practice. As scholarship already predicts, the delineation of volunteers’ labour as aid versus work was constructed relationally and was context-specific: very much tied to their particular nationality and legal status as Syrian refugees in Jordan living and coming from Global South locations. However, volunteers’ statuses in relation to aid organisations were not as clear-cut as their legal ones. They were ambiguous because of aid’s hypervisible, performative role that made volunteers’ value to aid organisations confusingly tied to their simultaneous, fluctuating visibility as vulnerable ‘beneficiaries’ in need, but also resilient, empowered ‘workers’ in a protracted conflict setting. The latter matters because it points to the fact that aid, as a vehicle that renders work invisible, does so in a way that is by no means straightforward, unidirectional, or ‘out of sight’. Rather, the process is multi-directional and ambiguous – somehow strategically and explicitly articulated as such – so much so that volunteers seem to experience particular forms of precarity contingent on their confusing relationships with aid organisations as a result. The additional ‘hidden’ labour they performed to navigate their positions is indicative. Future studies could therefore further consider how the direction and fluctuation of labour delineated as work and not-work itself intersects with other mechanisms that render work invisible. The hypervisibility of these fluctuations might also provide an important empirical entry point to identify exactly where and when ‘mechanisms intersect to intensify disadvantage in the workplace’ and how these intersections ‘vary across cultural contexts and over time’ (Hatton, 2017: 345–346).

How refugees laboured as worker-beneficiaries, for example, points to the sociocultural mechanisms that rendered their work invisible and their intersection with sociolegal ones. Being ‘a refugee volunteer’ was not just a legal classification that conditioned the scope of where and what one could perform as ‘work’. It was also associated with particular expectations as to how one should ‘consume’ work as aid and why. These sociocultural assumptions were ‘enacted on workers’ bodies’ (e.g. expressing gratitude for work, serving as ‘the face’ of the organisation for donors) as Hatton’s framework predicts (2017: 339–340), but were further articulated temporally (e.g. expect temporary positions and rotating work opportunities, tailor career expectations). This is important because literature is prolific on refugee precarity stemming from their legal status as non-citizens in host contexts, especially in terms of their labour market opportunities. Yet, the ways in which ‘refugee’ status was articulated in the aid workplace points to other organisational and sociocultural mechanisms that shape the meaning of ‘the refugee’ in relation to work and labour markets as well.

While this article does not claim to systematically account for the relationship between invisible and insecure work, and is limited to the particular case of Syrian refugee volunteers in Jordan, its findings provide important insights for future research to explore in this regard. Elaborating accounts for the ways in which legal statuses and the sociocultural assumptions that undergird them relate to precarity is one important direction for future studies. This is critical at a time in which states’ reliance on non-state international and supranational organisations, including aid actors, to recruit and select workers to fill temporary work programmes (e.g. Canada’s partnership with IOM), implement migration management schemes (e.g. Frontex) and aid forcibly displaced groups like refugees (e.g. UNHCR) is rapidly growing. Normative meanings associated with aid as a remedy to social problems, particularly in the Global South, may therefore expand the ways in which state and non-state actors can and do define the boundaries of work. Considering the relationship between volunteering and invisible work and how it is articulated in different contexts that compare, but also diverge, from this present case may therefore provide important insights into the processes that create ‘the vulnerable and compliant workers on whom neoliberal precarity relies’ (Hatton, 2020: 20), and ‘provincialise universalising claims about precarity’ (Millar, 2017: 6).