2 Theories of organisational change from below

In anthropological theory, organisations change because of multiple external influences and internal dynamics such as organisational learning (Czarniawska, Reference Czarniawska and Dierkes2007), bureaucrats’ motivation (Lentz, Reference Lentz, Bierschenk and Olivier de Sardan2014), cultures or politics (Hoag and Hull, Reference Hoag and Hull2017, p. 15) and power struggles (Heyman, Reference Heyman2004). Organisational change is the result of a complex and contingent process (Czarniawska and Joerges, Reference Czarniawska, Joerges, Czarniawska and Sevón2010). It happens when ‘people begin questioning things that were previously taken for granted’ (Czarniawska and Sevón, Reference Czarniawska, Sevón, Czarniawska and Sevón2010, p. 2). The acts of criticism I observed by junior bureaucrats thus indicate that they did not take the functioning of their organisation – and sometimes even the state – for granted and attempted to change it.

Only a few authors have interpreted bureaucrats’ actions as political practices of challenging or changing their organisation and more generally the state (Heyman, Reference Heyman1995; Reference Heyman2004; Andreetta and Kolloch, Reference Andreetta and Kolloch2018). In the case of Niger, Hamani (Reference Hamani, Bierschenk and Olivier de Sardan2014) departed from the frustration of bureaucrats and interpreted their practices as incremental reforms. Other studies have shown how civil servants went on strike to protest government spending cuts (Körling, Reference Körling2011, pp. 58–71) and how, outside the administration, social protests and associations frequently challenged the status quo (Sounaye, Reference Sounaye, Oinas, Onodera and Suurpää2018; Hagberg and Körling, Reference Hagberg and Körling2016; Schritt, Reference Schritt2015). While this research has carved out the relevance of everyday reform by street-level bureaucrats and of politics in Nigerien society at large, research on politics and conflicts inside West African bureaucracies still remains rare.

This gap is also reinforced by the recurring tendency to analyse West African bureaucracies through the lens of official and practical norms (cf. Bierschenk and Olivier de Sardan, Reference Bierschenk and Olivier de Sardan2019, p. 248). These concepts assume a shared organisational culture rather than highlighting conflicts and frictions between different bureaucrats, as the anthropology of the state approach has suggested (Heyman, Reference Heyman2004, p. 491; Bierschenk, Reference Bierschenk, Bierschenk and Olivier de Sardan2014, p. 242). In order to theorise organisational change from below, I instead propose to conceptualise the bureaucratic practices of voice, exit, compliance and everyday resistance by drawing on classics in the resistance literature (Hirschman, Reference Hirschman1970; Scott, Reference Scott1985).

Hirschman's (Reference Hirschman1970) influential work in economics offers a twofold conceptualisation of resistance from below in organisations. First, members can react to a deterioration of their organisation through ‘exit’, a private decision of abandonment. Second, they can express their dissatisfaction to the management through ‘voice’, ranging from a ‘faint grumbling to violent protest’ (ibid., pp. 15ff.). Hirschman's third option, ‘loyalty’ or members’ ‘special attachment to an organization’, does not constitute a strategy, but rather an identification with the office that reduces exit and activates voice (ibid., pp. 77ff.).

Mobilising Hirschman's framework for the anthropology of the state requires two conceptual adaptations. First, we need to replace Hirschman's primary interest in the causes, effects and interrelations of these strategies on an organisational level with the anthropological concern with bureaucrats’ emic perceptions and practices (cf. Hoag and Hull, Reference Hoag and Hull2017, p. 25). Focusing on the practices and discourses surrounding voice and exit allows us to centre the critical stances of these particular junior bureaucrats rather than formulating a critique from an external, normative standpoint – a tendency for which Blundo (Reference Blundo, Bierschenk and Olivier de Sardan2014, p. 83) has criticised much of the literature on West African bureaucracies.

Second, while Hirschman equates voice with politics because of its direct link to criticism, he locates exit in the economic realm. Avoiding direct confrontation, exit implies shifting to another organisation in the economic logic of competition (Hirschman, Reference Hirschman1970, pp. 15ff.). As Schaffer and Lamb (Reference Schaffer and Lamb1974, p. 88) convincingly argue, collective exit is political, too. Beyond these assessments, I see exiting the administration as a political act of demonstrating a refusal of exploitation. In his seminal work on the ‘everyday resistance’ of the peasantry in Malaysia, James Scott (Reference Scott1985) describes everyday practices of noncompliance, sabotage or slowdown that require little planning and, masked with ‘symbolic conformity’, no direct confrontation. Yet they uphold the class interests of the peasantry against more powerful groups. Similarly, as I argue in the following, bureaucratic practices of exit and slowdown can be interpreted as a refusal of exploitation because they limit the workload and convey dissatisfaction to supervisors while avoiding direct confrontation. In this sense, exit and everyday resistance are political strategies of mobilising for organisational change, too.

My analytical framework takes voice, exit and everyday resistance as political practices of challenging the organisational status quo and combines these practices with forms of compliance. In this way, it foregrounds mobilisations for change. Before applying the framework to the Nigerien Refugee Directorate, I briefly introduce the office and its junior bureaucrats to underline some structural constraints on their reform capacities.

3 The Refugee Directorate and its junior bureaucrats

The Refugee Directorate (Direction des Réfugiés), located in the Interior Ministry, is the office responsible for asylum and refugee protection in Niger. Its services for asylum seekers and refugees recognised under individual Refugee Status Determination are concentrated in the capital Niamey, but were extended to the Sahara transit town Agadez in 2017. In 2019, I conducted participant observation in Niamey both in its headquarters and in its outlet registration and protection offices inside the one-stop shop Guichet Unique, which apart from these government offices hosts NGO protection and assistance services in a building run jointly by the state and the UN Refugee Agency (UNHCR). Due to often unspoken concerns of the Refugee Directorate's decision-makers, my ‘internship’ was limited to a month and to certain activities. During this period, I stayed with the junior temporary staff while they registered asylum seekers, compiled asylum files, delivered asylum documents, produced statistics, conducted protection activities and organised reunions of the National Eligibility Commission and Administrative Appeal Committee. Senior staff members were absent more than half of the time as they attended missions in the country and abroad, remunerated with daily allowances (per diem). When they were present, they remained relatively inaccessible and some junior bureaucrats tended to avoid contact, suggesting differing trust relations with the ethnographer and loyalty to the office.

Due to these accessibility issues, the following analysis focalises those junior bureaucrats who voiced their criticism. Many office hours were filled with lively office talk, informal conversations, side activities and boredom, which provided ample insights into the criticism staff had. I extended these exchanges over ten months of authorisation runs and two months of courtesy visits and interviews with different actors after the internship, totalling thirteen months of fieldwork in 2018–2019. As the general research focus on asylum precludes an anonymisation of the office, care was taken to protect individual identities.

The bulk of the work observed was done by sixteen registration, eligibility and protection assistants on annual contracts. The majority had been employed some fortnights ago during the recent reinforcement of the office. Some had been working there for a few years.Footnote ¹ They were university students or recent graduates from law, sociology, administration and related disciplines. Although relieved to have a steady income in a context of structural underemployment, their career aspirations mostly went beyond the Refugee Directorate. Many dreamed of accessing UN organisations or international NGOs, establishing party careers or doing a PhD. Only some intended to enter the civil service, which they considered – similar to what Lentz (Reference Lentz, Bierschenk and Olivier de Sardan2014) described for Ghana – relatively secure, but low-paid and hardly accessible employment.

A motivation for university graduates to work in the office stemmed from its newly gained importance. Created during the handover of Refugee Status Determination from the UNHCR to the state in 2000, the Refugee Directorate remained a marginal state organisation until 2012. According to a civil servant, its head department, the Interior Ministry, only became important with the rise of jihadist activities in the Sahel since 2012, which also created refugee flows from Mali and Northern Nigeria that the Refugee Directorate became busy managing under refugee group status. Since 2015, Niger, an important transit state, has become a key partner for the EU in implementing migration control and refugee protection in an attempt to decrease refugee flows to Europe (Lambert, Reference Lambert2020). These latter developments pushed individual asylum applications in Niger from ninety-eight in 2017 to about 4,000 the following year (UNHCR, 2017; 2018a). According to a member of the National Eligibility Commission, the Refugee Directorate now ‘has become important …. Before, it was banal, it treated maybe a Congolese in need’.

For the office, this meant additional work, symbolic gains and EU resources for its capacity-building. Working for the asylum administration thus became an attractive career option in an ‘emerging field with opportunities’, as one intern quoted her father, who had urged her to apply. More than 800 people responded to a week-long job opening for seventy-six temporary positions in 2019. Staff nationwide nearly doubled since 2017 and eligibility and protection assistants at headquarters even quadrupled from three to twelve. Nationwide, eighty-six temporary staff stood against eighteen civil servants in 2019. Temporary staff clearly outnumbered the civil servants, which likely affected their supervision and the organisational continuity. In headquarters, they were supervised by two directors and four division leaders.

Although constituting the overwhelming workforce, junior bureaucrats had structurally limited influence on their employer. The office was organised as a ‘project’ – a common form in development co-operation in which a state structure is operated with donor funding, here the UNHCR. This structure granted its management more autonomy from departmental control and Nigerien labour regulations, allowing repeated annual contracts and exemptions from collective bargaining agreements. Consequently, the project structure of the Refugee Directorate constrained temporary staff's protection and reform capacities.

The new importance of the Refugee Directorate, its project structure, the relatively recent profession of eligibility, protection and registration assistants, their mostly short-term employment experience and alternative career orientations all structured junior bureaucrats’ perceptions of their organisation and structurally limited their reform capacity. In the following, I explore the acts of semi-private criticism that they nevertheless directed at the office.

4 Semi-private criticism of working conditions and the state

Sharing criticism with the ethnographer, as in the initial vignette on the lack of activity in the Refugee Directorate and more generally the state, was the recurrent and most frequent form of criticism I observed among temporary staff. A relevant number of temporary employees shared their dissatisfaction with me, although some remained silent. ‘Everyone is frustrated,’ a temporary employee claimed. Employees usually shared these views in quasi-private conversations with me and only one or two peers present to avoid potential sanctions. As I explore in this section, these semi-private acts of voice covered a broad range of work-related issues and were partly connected to a wider critical imaginary of the administration and the state.

As in the initial vignette, some junior bureaucrats regretted a lack of work. One recently employed staff member who lay with her head on the desk replied as follows to my question on whether she was tired: ‘No, I am bored. They do not give us work.’ In another situation, she explained her frustration:

The reasons for her inactivity were obscure to her: ‘I don't know why. The state agents are there, the files are there. But it does not advance.’ The bureaucrat quoted in the initial vignette connected the personal experience of boredom to a wider critical conception of the administration that ‘in the state you are not working’ and that service delivery was ‘slow’. This perceived inefficiency has been widely described in the literature on African street-level bureaucracies (Olivier de Sardan, Reference Olivier de Sardan, Bierschenk and Olivier de Sardan2014, p. 406) and on asylum bureaucracies worldwide (Gill and Good, Reference Gill, Good, Gill and Good2019, p. 6).

A related concern was the lack of learning and career advancement in the office. As in other West African bureaucracies (Bierschenk, Reference Bierschenk2008; Blundo, Reference Blundo, Bierschenk and Olivier de Sardan2014), there were ‘no ways to advance’ in the Refugee Directorate: ‘Here, you start as a [protection, registration or eligibility] agent and you end up as one.’ For a colleague, the initial learning on the job had made it fun, but was now missing. Those aiming to advance in their careers in the administration criticised the lack of possibilities to join the civil service, because the state did not organise the required competitive entrance examinations (concours) or cancelled them after clientelist ‘irregularities’ in the recruitment had been decried. The last completed concours for the Ministry of Interior was carried out in 2008, when the Refugee Directorate's current senior management was recruited. Junior staff criticised the missing recruitment as a lack of ‘political will’. For them, the older generation blocked the youth's access to tenured civil servant positions and instead filled vacancies in the department with retired state agents on contracts. In this way, junior bureaucrats linked their stalled career development to wider politics, such as the Structural Adjustment Programs targeting a reduced and competitive civil service recruitment, and social power relations like clientelism and gerontocracy in the administration.

As in other bureaucracies (Lipsky, Reference Lipsky1980/2010; Bierschenk, Reference Bierschenk2008), a shared concern across ranks was the lack of work material. Particularly remarkable was the lack of rooms for the confidential registration and eligibility interviews of asylum seekers. While a new building was under construction and more laptops had been ordered, staff had to work in ‘crammed conditions’ and regularly lacked computers for their work. In an attempt to ‘make do’ despite the limited means (Hamani, Reference Hamani, Bierschenk and Olivier de Sardan2014), they used private laptops and vacated offices for interviews. Staff members regularly pointed out this lack of means to me. For instance, in a wrap-up conversation at the end of my internship, one senior bureaucrat commented: ‘You have seen the working conditions here. It also affects our work. The lack of space …. When there are power cuts it breaks our material – and the heat.’

Similarly, low salaries and social security benefits were a shared experience among junior staff members, in resonance with other West African bureaucracies (Bierschenk, Reference Bierschenk, Bierschenk and Olivier de Sardan2014; Andreetta and Kolloch, Reference Andreetta and Kolloch2018). ‘We are just paid so we are not jobless, it is not a real salary,’ as one employee summarised in a small group discussion. Nevertheless, the low salaries paid to state agents were often naturalised as a feature of the administration. One staff member referred to the shared knowledge that salaries were comparatively low: ‘Everyone here knows that you touch maybe [300–500€] and this as a senior official.’ On another day, a sick contract-holder hesitated to seek medical advice given the consultation fees. My surprised question about their current medical scheme was put into perspective by his colleague, who contrasted it with their prior employment experience in an NGO: ‘Of course, the NGOs have more money. There you are better off, there you have certain advantages. And now quickly write it down!’ In her analysis, which she seemed unwilling to deepen with me in front of her colleagues, low social benefits figured as a given when working for the state, which was considered poorer than NGOs, although the directorate's funding likewise came from the UNHCR. Another staff member preferred to speak with the ethnographer, but not with her NGO colleagues:

The feeling of shame, or being exposed to the disapproving gaze of others (Khosravi, Reference Khosravi2007, p. 331), can be interpreted as accepting to work for a remuneration not considered adequate in comparison to their NGO colleagues. Here, the employee explained her absenteeism at work with her low salary – a relation Bierschenk (Reference Bierschenk, Bierschenk and Olivier de Sardan2014, pp. 239ff.) contextualised as a compensation strategy of pretending to work when African states only pretended to pay their officials. In the case of a donor-funded state project, close collaboration with NGOs allowed a direct comparison of working conditions between the state and NGOs, and the discrepancy to be challenged.

Beyond their NGO colleagues, some employees compared themselves unfavourably to their superiors in terms of salary. These staff members recalled UNHCR colleagues condemning these pay hierarchies: ‘Even the UNHCR people told us all the time: “But you have these small salaries and your patrons get all the money, but they do not even work!”’ Through exchange amongst peers, they also found out that the pay for most contract-holders was lowered by 30 per cent in 2018, while in parallel it was raised for those ‘up in the hierarchy’. These critics were aware that their superiors made salary proposals to the UNHCR in the annual budget negotiations and could have negotiated higher pay for temporary staff. The increasing pay hierarchies in the office in the context of rising EU funding were thus perceived as an illegitimate choice by their superiors. They were also an open secret between employees and shared with UNHCR staff and the ethnographer. Here, the critique of working conditions was also closely linked to a criticism of the directorate's project structure, which granted considerable autonomy to the senior management beyond collective bargaining agreements.

In addition to the low salaries, annual contracts created employment insecurity. They were also renewed with delays, leading to interim unemployment. One employee considered them ‘just a way to frustrate us’. Her colleague also criticised that being held more than twice on annual contracts contradicted Nigerien labour law. An employee recalled a situation in which a staff member was not reemployed the following year after objecting to corrupt orders of their superiors. Similar to Kolloch's examples (in this issue) of relocating judges to the countryside, staff also recalled how fellow employees were sent to work in the regions outside of the capital as a way to ‘get rid of people to then place their own people’. The work insecurity staff experienced was considered the direct responsibility of their superiors, chosen for reasons such as humiliation, staff control and clientelism. In these views, staff placement and contract renewal also required a protector in senior management or influence on it.

The need for protection was confirmed by a temporary employee: ‘If you do not have somebody there, you cannot survive.’ Apart from employment security, a protector also guaranteed a tolerable everyday atmosphere. Different staff members spoke about demeaning behaviour by their superiors or colleagues (coup-bas) and two saw this as a general feature of the administration. A senior staff member confirmed that the Interior Ministry was ‘a bit the jungle. You can receive coups-bas and you need to be very habile’. Additionally, two employees spoke about a perceived lack of recognition by superiors: ‘We are really demotivated. If the pay is bad, but there is recognition at work, this is one thing, you can enjoy your work. But here we are not recognized, we are looked down upon.’

Apart from these acts of humiliation, privileges in the administration were seen as hierarchised according to formal and informal power. As is generally the case in the ‘gift economy’ of external co-operation (Andersson, Reference Andersson2014), access to missions and allowances (per diem) was unequally distributed among bureaucrats. Employees recalled missions for which they had only received a share of the official allowance because they were told there was ‘no money’, and an instance in which a colleague was not paid at all. Their superiors, in contrast, had the reputation of always being on missions in search of a per diem and opportunities, causing frustrations about their absences in the office. Yet the resulting low supervision also granted a certain freedom: ‘It can be a week that I do not see [my boss] and I could stay home. [They] would not find out, but my religion tells me that this is not okay.’ Senior staff absences thus facilitated personal absenteeism, turning it into an issue of individual ethics rather than top-down bureaucratic control.

In parallel to the access to privileges, recruitment was often portrayed in semi-private discussions as a clientelist decision. ‘It is purely political now, it is purely political!’ one insider exclaimed. In these depictions, staff were hired because of their social networks, family background or political affiliations, leading potentially to the recruitment of unqualified staff (cf. Bierschenk, Reference Bierschenk, Bierschenk and Olivier de Sardan2014, p. 228; Bierschenk and Olivier de Sardan, Reference Bierschenk and Olivier de Sardan2019, p. 248). ‘Here it is not that you get a position because you merit it,’ an employee told me. As a part of this discourse of connections over merit, staff inscribed themselves in a nostalgic discourse of a formerly just, rigorous bureaucratic ethos. In this account, the ‘real administrators’ were replaced by people who were ‘la politique’ and went ‘from door to door’ to get a post. Being a real bureaucrat thus meant separating politics from bureaucracy:

This nostalgic discourse on Weberian ideal-type bureaucrats (Jaffré and Olivier de Sardan, Reference Jaffré, Olivier de Sardan, Jaffré and de Sardan JP2003, p. 44) suggested the rise of clientelist recruitment or the so-called ‘politicisation’ (Bierschenk, Reference Bierschenk, Bierschenk and Olivier de Sardan2014, p. 228) under the then Issoufou government. By criticising their superiors’ recruitment decisions, these staff members alluded to a closer imbrication of the administration and the government – a development they saw as threatening the quality of service delivery. Importantly, however, several of the temporary staff stressed merit as a reason for their recruitment. ‘I was the last one who was fairly recruited,’ one of them told me. As Lentz (Reference Lentz, Bierschenk and Olivier de Sardan2014) has described for Ghana, some staff members here pursued boundary work that ascribed clientelism to others and merit to oneself. Most temporary employees, however, kept silent on the issue.

As the aforementioned examples of voice shared with the ethnographer and colleagues suggest, the criticism uttered by temporary employees in the Nigerien Refugee Directorate covered a wide range of work-related issues ranging from boredom, lack of career advancement, office equipment, low salaries and contract insecurity to humiliation, unequal access to privileges and clientelism. They can be interpreted as political strategies of challenging the current work relations in their specific workplace. In some cases, staff also alluded to a wider critical imaginary of the state as inefficient, clientelist and authoritarian, widespread representations in Nigerien society (cf. Olivier de Sardan, Reference Olivier de Sardan, Bierschenk and Olivier de Sardan2014, p. 406) and amongst the office's clients, asylum seekers and refugees. In the following section, I explore the interrelatedness of semi-private criticism with symbolic conformity, exit and everyday resistance in mobilising for organisational change.

5 Symbolic conformity, external change and everyday resistance

As the agent in the introductory vignette explained, semi-private criticism was a prevalent strategy in a context in which bottom-up reform proposals remained effectless and could induce dismissal. However, causing change through semi-private criticism was more complex.

By keeping their frustration in restricted circles, junior bureaucrats made sure it would not reach their superiors. In the everyday, they performed compliance with the official and practical norms of the office. Staff meticulously filled in the time registration book and approximated their indicated arrival and departure times to the expected office hours. When colleagues or the ethnographer left work early, some staff commented with a serious expression: ‘It is not the time yet!’ Although these interjections could speak of a dry sense of humour, they still appealed to the official work hours and challenged the rule-breaker to come up with an explanation. When occasionally controlled by supervisors, other staff excused their absent colleagues. ‘It's due to fatigue’ was a widely shared explanation. Or else, ‘the roads were blocked’ and the rain were suitable external factors for arriving late. Staff also followed orders they considered unjust, such as showing up to work on the weekend or when all supervisors and most colleagues were sent on a mission. These examples of ‘symbolic conformity’ (Scott, Reference Scott1985, p. 33) suggested an agreement with the official rules of the office and partly concealed staff dissatisfaction.

Some staff also protected the office vis-à-vis criticism from outside. When an asylum-seeker picked up his three-monthly attestation and pointed out that it was only valid for another month, an employee replied that staff were ‘overburdened’ and that time remained. Demonstrating loyalty (Hirschman, Reference Hirschman1970) to the office was also a response to perceived criticism by the ethnographer. In an internship report I presented to a round of junior and senior staff, I mentioned the often-discussed lack of work material like computers and rooms and the creative solutions staff developed to provide service nevertheless (Hamani, Reference Hamani, Bierschenk and Olivier de Sardan2014). A junior bureaucrat refuted this analysis: ‘How is the material here important? … This is still the state of Niger. You cannot treat it as débrouillardise (make do).’ He thus dismissed academic narratives of lack and coping as carrying a disrespectful imaginary of the Nigerien state, while demonstrating his loyalty to the office and the state.

One effect of this symbolic conformity was passivity. In the introductory vignette, the threat of dismissal had incited the quoted agent to abandon his reform proposals. As already mentioned, temporary staff who resisted an unjust order were not rehired the following year. Non-compliance, or not following the rules of the administration, could be costly for civil servants (Lipsky, Reference Lipsky1980/2010, p. 24). Instead of bringing about reforms, the bureaucrat waited for the retirement or death of the older generation in power, whom he identified as the source of the problem. According to him, political change was due as long as some state departments placed youth in central positions. In the sarcastic discourse of a colleague, waiting could also directly target an unpopular superior: ‘You just wait. If it is not a sickness that brings him away it will be politics, because it was [a government member] who brought him.’ In this view, the politicisation of the administration also resulted in unpopular superiors changing posts without further ado. As a consequence, waiting for external change bore a potential for organisational change.

Junior bureaucrats’ acts of horizontal voice while waiting for external change is reminiscent of what Boyer (Reference Boyer, Fouquet and Goerg2018) describes for the social change envisioned by young Nigerien men. In order to become adult men, the youth have to become socially independent – a complicated undertaking in a situation of structural underemployment. In their popular nightly discussion circles (fada), they support each other and analyse the gerontocratic and clientelist relations that lead to their social domination. Their critical analysis, however, remains in the closed realm of their peer group. Boyer therefore suggests that the fada constitutes a space for ‘tactics of avoidance’ of domination by creating autonomy and operates inside these hierarchies rather than openly challenging them. Change comes from waiting for an improvement in their individual economic and marital status in order to acquire adulthood. As in the fada, when keeping their criticism amongst peers, staff in the Refugee Directorate created autonomy for themselves without openly challenging the power relations they saw at play.

Beyond horizontal voice in the peer group, criticism in the Refugee Directorate was shared with the ethnographer. Some encounters suggested that staff saw me, the European ethnographer, as an influential change agent with a ‘backpack of potential co-operation’ with Europe. One senior bureaucrat asked me: ‘Is your research purely academic or is it also political? [We] do not know whom you will present it to.’ Communicating criticism to researchers was therefore potentially also a strategy of effecting external pressure on the senior management through anonymous, time-delayed academic reports while limiting sanctions. Other situations suggested staff drew boundaries (Lentz, Reference Lentz, Bierschenk and Olivier de Sardan2014) vis-à-vis illicit practices or vented frustration based on a proximity developed on the grounds of shared friendships, age, education and interests. Inasmuch as the motivations differed, sharing criticism with the researcher meant entering the political realm of voice with open effects on organisational change.

Apart from mobilising for external change through academic reports, staff also shared their criticism with other external confidants. The concerned junior staff members were all members of the university-educated elite in Niger, and some of them were members of political parties, associations, NGOs or influential families. These elite sociability networks – as Lentz (ibid.) described them for Ghanaian bureaucrats – provided spaces in which junior bureaucrats could potentially mobilise the informal support of powerful allies for an improvement in their working conditions. The pervasiveness of these informal and multiple accountabilities requires an ethnographic methodology (Bierschenk and Olivier de Sardan, Reference Bierschenk and Olivier de Sardan2019, pp. 248ff.). Through my long-term immersion in society, I learnt for example that a senior contact of mine had been informed by a contact about the lack of work and supervision in the Refugee Directorate and was now trying to ‘see what [he] can do’. Just as these protectors supported my field access, they could choose to facilitate the improvement in an individual employee's position in the office. Semi-private criticism could thus also involve a tacit mobilising with influential actors for externally induced change through one's elite networks.

Furthermore, some observable staff behaviour can be analysed as everyday resistance. As has been described for the everyday in African street-level bureaucracies (Olivier de Sardan, Reference Olivier de Sardan, Bierschenk and Olivier de Sardan2014, p. 406), Refugee Directorate staff actively applied for other work during work hours, vividly exchanged information on founding or accessing NGOs, pursued higher education, did side businesses or cut short work hours with absenteeism. The literature on West African bureaucracies has interpreted these practices as ‘practical norms’ – widespread informal practices that although being sometimes outright illegal correspond to social logics and moral economies (Bierschenk and Olivier de Sardan, Reference Bierschenk and Olivier de Sardan2019, p. 248). They were rarely punished in West African bureaucracies (Jaffré and Olivier de Sardan, Reference Jaffré, Olivier de Sardan, Jaffré and de Sardan JP2003). The notion of practical norms thus assumes a shared normative understanding of different actors inside the administration. For instance, seeking exit options and absenteeism at work have been described as a compensation strategy of bureaucrats for their working conditions (Bierschenk, Reference Bierschenk, Bierschenk and Olivier de Sardan2014, pp. 239ff.). While the concept of practical norms usefully sensitises one to the reasons and societal origins of bureaucrats’ actual practices, I argue that it precludes a more power-analytical view on the negotiations and contestations between bureaucrats of different ranks. Rather than presupposing a shared organisational culture based on similar norms, Heyman (Reference Heyman2004, p. 494) challenges us to look at the stratifications and divisions inside organisations and their relation with societal power relations.

In my view, an interpretative angle on these practices of exit and slowdown as everyday resistance can account for contestations inside the office because they convey bureaucrats’ refusal of exploitation. A staff member herself interpreted colleagues’ practices of not doing what they promised as ‘a refusal’. Although these practices remained generally unsanctioned, as an insider confirmed, this does not imply that they were supported across hierarchies, as the notion of practical norms would suggest. At times, supervisors commented on this behaviour with questions like ‘Where is everyone?’, ‘Who is missing?’, ‘What is going on? Is everybody sleeping?’ or comments like ‘Today is not a holiday’. Through my personal networks, I learnt that a senior bureaucrat was frustrated by the ‘nonchalance’ of staff. My contention with the literature is therefore twofold: first, I challenge the notion of practical norms for presupposing a normative agreement between ranks and suggest reframing these practices of exit and slowdown as everyday resistance in a power play between ranks; second, in contrast to Hirschman's understanding of exit as following an economic logic, I argue that these practices were political because they communicated an opposition to their superiors. As such, individual exit and slowdown constitute a form of political contestation, even if they were only perceptible as ‘faint grumbling’ (cf. Hirschman, Reference Hirschman1970, p. 16) or nonchalant behaviour.

Additionally, many members’ practice of exit represents a strong form of voice (Schaffer and Lamb, Reference Schaffer and Lamb1974, p. 88). It communicates a clear opposition in a constellation of conflicting interests and interpretations inside the office and can induce organisational change. One employee told me: ‘Everyone is just waiting to jump off to an NGO or the UNHCR and those who are still here, they have just not made it yet.’ Later she added: ‘[The Refugee Directorate] can only be transitory, it is no place anyone would like to stay.’ A senior staff member regretted the ensuing ‘internal brain drain’ (Bierschenk, Reference Bierschenk, Bierschenk and Olivier de Sardan2014, p. 236) to the UNHCR: ‘These people are not here to stay …. How to find those who will stay is the question. Many people left to the UNHCR. It is us who trained them and then the UNHCR takes them over.’ On an organisational level, brain drain caused the office to lose trained staff after a short period. This issue also became a matter of concern for the donor UNHCR since it meant a loss of competences in the office. In a UNHCR study (2018b) on the office's capacity-building that UN staff shared with me, the UNHCR regretted ‘poor working conditions [and] no retention policy’ and proposed to ‘review [the] working conditions of eligibility and registration staff’. Therefore, the collective strategy of ‘exit’ in opposition to the working conditions also mobilised the donor UNHCR, contributing to the building of broader external actor coalitions often needed to challenge the bureaucratic status quo (Heyman, Reference Heyman2004, p. 494).

To summarise these points, the organisational change linked to acts of semi-private criticism was ambiguous. Officially, junior bureaucrats performed symbolic conformity with the norms of the administration. Practically, they engaged in organisational change in three ways. First, by keeping criticism to private conversations with their peers, the employees created autonomy for themselves without actually challenging power relations. Change-making here consisted of passively waiting for external change. Second, sharing criticism with the ethnographer and potentially powerful confidents from their elite networks mobilised external actors for their support and externally induced organisational change. Third, temporary employees’ dissatisfaction translated into subtle forms of everyday resistance and a collective opposition strategy of exit, which voiced a refusal of exploitation to their superiors and mobilised the donor UNHCR in their support. Therefore, external actors had an important role in bringing about organisational change in semi-private criticism.

6 Direct criticism and informal solidarities

Apart from semi-private criticism, staff occasionally also criticised their superiors upfront. This direct criticism was surprising given a general reluctance to criticise colleagues in bureaucratic organisations and the aforementioned potential sanctions this could entail (Lipsky, Reference Lipsky1980/2010, pp. 23ff., 203). In this section, I argue that direct criticism was facilitated by compliance through other forms of solidarity with their superiors in order to avoid falling out of favour with them, such as shared political and social affiliations or criticising the ethnographer. Early in my participant observation in the Refugee Directorate, I involuntarily facilitated an event of direct criticism and the ensuing apology of a supervisor.

In 2019, the UNHCR organised the annual World Refugees Day celebration in a refugee camp forty kilometres outside of the capital. The day before the event, most junior staff members were called into a supervisor's office. They returned cheerfully laughing with envelopes in their hands and their chatter about a two-day per diem (for what later came down to a five-hour trip) filled the room. In the office the following day, on World Refugees Day, I only found those five eligibility agents who had been recently transferred to the office from the UNHCR after an initial recruitment and training by the UNHCR. Although visibly frustrated, ‘the five from the UNHCR’, as they were sometimes called, only explained that they had been excluded from the mission without being given a reason, criticising the intransparency at play. They expressed surprise that I was not taken along. I had been told that no seat remained to take me on the mission.

The day after, I had an interview with a supervisor to ask some research questions. She spontaneously invited ‘the five from the UNHCR’ and two others to join us. After some initial questions by me, one of them used the occasion to ask three critical questions on my research, including his fear that, as a foreigner, I would expose the ‘secrets of the administration’. The supervisor cut short my explanation and the junior staff member then continued by directing criticism at his superior. He said they felt ‘like strangers in the house’ since their exclusion from World Refugees Day; they lacked computers for their work as well as information on their colleagues’ roles in the office; and he bemoaned a slow work rhythm in the office compared to their former employer UNHCR. The supervisor then justified the lack of computers with the office's comparably lower budget and the UNHCR's decision to send them to the Refugee Directorate. She confirmed that there was a higher workload at the UNHCR and challenged them to adapt (‘you will suffer a bit’). Only when a second employee reiterated that he felt like a stranger due to their exclusion from the festivities did the supervisor interrupt him and start an explanation, but did not find words. A third employee then heckled: ‘There is no reason!’ The supervisor replied: ‘I assume the responsibility for it. In any case, we are together (on est ensemble).’ She promised to give them the remaining public relations material – UNHCR T-shirts and scarves – and added: ‘I sacrificed you, but I am currently catching up.’ This exchange was similarly repeated once more and ended with the supervisor saying: ‘I accept. You are no foreigners, you are at home.’ She then ended this conversation by asking me whether I still had questions. Afterwards, they left for a collective prayer.

In this surprisingly confrontative form of group voice (Schaffer and Lamb, Reference Schaffer and Lamb1974), junior staff brought up similar issues that they conveyed in semi-private criticism, such as lacking means, boredom, unfair treatment and their superior's favouritism in granting access to a mission and its per diem. Although the critique targeted the actual behaviour of their superiors, junior and senior bureaucrats here settled on an imaginary of the state as resource-poor and slow, in contrast to the UNHCR. Apart from this very explicit negotiation, I observed a few more mundane instances of direct criticism in which junior bureaucrats hinted at the lack of work material and the respect for breaks and weekends. Nevertheless, these instances remained rare.

In the example above, junior bureaucrats’ confrontation led their supervisor to apologise, promise equal treatment in the future and perform conciliation. After the discussion, one of the junior bureaucrats was satisfied with the ‘good talk’ and that their superior had said something. On a social level, direct criticism could eventually improve working relations after an experience of blatant unequal treatment. Nevertheless, there was no guarantee that the apology would bear consequences. For the next mission, none of them was among the few selected staff members. No observable effects followed from contesting boredom and the lack of material. Staff also told me that in the past, they had ‘revolted’ for a salary raise, but finally their request was turned down. Another employee stressed that their wish for direct exchange with their superiors was responded to with threats of dismissal: ‘When we talk about these injustices and ask, because we want to talk, we are told that we just have to stay home and that there are 1000 people who would be happy to have our position.’ In all of these situations, staff directly expressed their dissatisfaction with specific working conditions, but without being able to commit them to noticeable organisational change. Instead, in one instance they were threatened with dismissal.

To avoid such sanctions, some junior bureaucrats relied on other forms of solidarity. The aforementioned vignette exemplifies the embeddedness of direct criticism in relations of proximity and power. The first, most daring speaker framed his criticism with a critique of the ethnographer as a critical outsider. That way, he demonstrated his sharing of the senior staff's mostly implicit concerns about ‘exposing the secrets of the administration’ and hence an interest in protecting the office from criticism. Some junior staff could also mobilise social networks and shared party affiliations. They did not mention them to me, but these connections became apparent over time in the everyday, in the way they attended the same social events and discussed politics. These political and social solidarities could potentially prevent sanctions after such direct criticism.

In other words, the demonstration of shared concerns about shielding the office from criticism and the protection deriving from social and political networks facilitated acts of direct criticism towards their superior. This is not to say that only those in these networks of informal accountabilities were able to upwardly voice their concerns, but this interlinkage of direct criticism with protection from sanctions hints to an ambiguous effect of direct criticism on organisational change. Although it represented a strong strategy of voicing dissatisfaction and mobilised an agreement on official norms across ranks, criticising favouritism and unfair treatment and more generally working conditions could not guarantee outcomes and partly relied on social and political solidarities that were similar to those the staff members vocally challenged.