The problematics of accountability: Internal responses to external pressures in exposed organisations Critical Perspectives on Accounting

This paper seeks to contribute to the emerging stream of literature on NGO accountability by drawing on the theoretical assertions of the problematics of accountability. It discusses how accountability has become problematic for NGOs by examining how these problems are created and how NGOs subsequently respond. Drawing on in-depth interview data complemented by extensive documentary analysis, we explore the accountability experi- ences of Muslim NGOs operating in the UK. We highlight how speciﬁc accountability regimes are inﬂuenced by the prevailing political, social and economic context within which they operate. We suggest this has caused Muslim NGOs to become ‘exposed’ organisations where demands for accountability, framed by security and counter-terror con- cerns, are limiting their ability to be accountable. Utilising insight from Kearns (1994) we explain how Muslim NGOs respond to these problems. We ﬁnd engagement with dis- cretionary accountability tactics allow the NGOs to gain trust of stakeholders, whilst strategic change processes help the NGOs to anticipate and negotiate heightened accountability demands. We discuss the implications of our ﬁndings for both theory and practice. does not their in the (2005) of account-giving activities. The problematics of scepticism (Messner, (2005), the nature of organisations means scrutinised, their behaviour and activities are commented upon, and whatever they say and do is received with scepticism. In a our ﬁndings suggest that Muslim NGOs, operating within the current socio-political discourse, which views their Islamic identity with can become politically exposed to such limitations and therefore the normal routines of accountability no longer serve to satisfy or appease the regulator, or reassure the public. This suggests accountability demands are of a differing nature for Muslim NGOs than for similar non-Muslim NGOs. We now turn to analysing the effect these demands are having upon the organisational practice of Muslim NGOs.


Introduction
Increasing attention has been paid to the issue of accountability in non-governmental organisations [NGOs] over recent years (Agyemang, O'Dwyer, Unerman, & Awumbila, 2017;Ebrahim, 2003aEbrahim, , 2003bEbrahim, , 2005Edwards & Hulmes, 1996;Goddard & Assad, 2006;Hall & O'Dwyer, 2018;Lehman, 2007;O'Dwyer & Unerman, 2007. This research has suggested problematic implications of prevalent hierarchical forms of accountability, which prioritise funders at the expense of contextspecific and beneficiary-orientated accountability mechanisms. In a similar vein, studies within the broader accountability literature have also attempted to understand how accountability can become problematic and present other possibilities for accounting practice beyond its current narrow limitations (Roberts, 1991). For example, many commentators (e.g. Butler, 2005;Joannides, 2012;McKernan, 2012;Messner, 2009;Roberts, 2009Roberts, , 2018 argue for an intelligent accountability (Roberts, 2009(Roberts, , 2018 and suggest that more accountability is not always unambiguously desirable. It has been argued that in certain contexts, additional demands for accountability 'can become so great as to become ethically problematic' for the organisations expected to give an account (Messner, 2009, p. 918). This is an important insight for NGO accountability, because it acknowledges the ethical implications of greater accountability demands and the possible detrimental impact https://doi.org/10.1016/j.cpa.2019.01.002 1045-2354/Ó 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). and to seek reasons for actions taken (Goddard, 2004, p. 345). This implies certain expectations about what this person or organisation should be able and obliged to explain, justify and take responsibility for Cooper and Owen (2007).
In this paper we seek to add to the literature on NGO accountability. Research examining NGO accountability is rather embryonic. Existing research on NGO accountability theorises the extent to which, and to whom, NGOs should be held accountable (Ebrahim, 2003b(Ebrahim, , 2009Unerman & O'Dwyer, 2006) and examines the emergence and impact of various accountability mechanisms (Agyemang et al., 2017;Dixon, Ritchie, & Siwale, 2006;Ebrahim, 2003aEbrahim, , 2009Goddard & Assad, 2006;O'Dwyer & Unerman, 2007, 2010. This extant literature suggests that increasing demands for accountability are not necessarily beneficial for NGOs. There is evidence of inappropriate accountability mechanisms damaging rather than enhancing the social and environmental benefits that NGOs bring (Dixon et al., 2006;Martinez & Cooper, 2017;Najam, 1996;Unerman & O'Dwyer, 2006). The literature also suggests that a dominance of upward hierarchical accountability to donors, instead of wider holistic accountability incorporating beneficiaries, is resulting in NGO mission drift (Agyemang et al., 2017;Ebrahim, 2003aEbrahim, , 2005O'Dwyer & Unerman, 2007). For example, the emphasis on increased disclosure and oversight by regulators and government are often perceived by NGOs as coercive or punitive (Ebrahim, 2005(Ebrahim, , 2009Najam, 1996). Whilst Martinez and Cooper (2017), examining the role played by NGOs in social movements, suggest conventional accountability requirements hinder rather than advance the social change agenda of international NGOs.
Research also finds that accountees are only likely to react positively to upward accountability demands when the mechanisms enable them to perform their work better (Agyemang et al., 2017;Ebrahim, 2003a;Yasmin, Ghafran, & Haniffa, 2018). On the other hand, negative reactions are deemed likely when accountability demands are perceived as an attempt to control or curtail the activities of an NGO (Agyemang et al., 2017;O'Dwyer & Unerman, 2008). Thus, there have been calls for more situation and context-specific studies of NGO accountability to better understand the broader social and political context within which NGO accountability evolves (Martinez & Cooper, 2017) and in order to highlight how accountability can be made more sympathetic to the objectives and mission of NGOs (Ebrahim, 2003a(Ebrahim, , 2003b. We seek to add to this literature by exploring how and why accountability can become problematic for particular types of NGOs and subsequently how NGOs respond to these problems created by accountability. We do this by taking inspiration from the theoretical arguments surrounding the problematics of accountability and by employing an analytical frame, conceived by Kearns (1994), for considering the reactive and strategic responses of the NGOs.

Theorising the problematics of accountability
The research on problematics of accountability is largely conceptual and seeks to theorise the problems that can arise from greater or unambiguous demands for accountability. Although these problematics apply to all types of organisations and not just to NGOs, in this paper, we seek to specifically utilise insight from this literature to enrich our understanding of NGO accountability. The literature on the problematics of accountability broadly alludes to problematics as a social construction (Messner, 2009;Roberts, 2009Roberts, , 2018 and problematics as aporia (McKernan, 2012) 2 . McKernan (2012), taking inspiration from Derrida, suggests that accountability holds an unresolved contradiction, as the rendering of accountability undermines the very responsibility on which accountability is premised. For McKernan (2012), the very notions of 'giving of an account' undermine the morality of decisions implied by individual conduct. He suggests that one way of resolving the inherent contradictions could be to reconfigure accountability through the gifting of accounts. In this respect, the gift is a matter of responsibility and not a legal or ethical duty. The gift must be 'our responsibility to respond to the absolute demand of the other, without calculation (without calculating the return), without reason (without a why), without measure or reserve' (McKernan, 2012, p. 276).
On the other hand, informed by the work of Judith Butler (2005), Messner (2009) raises the question of whether in certain contexts more account-giving mechanisms are always and/or unambiguously desirable. This is especially important given his insight that the accountable self is an opaque, exposed and mediated self inherently limited in its ability to give an account for itself. Therefore, despite widely shared calls for increased public and managerial accountability, for Messner (2009), there are certain contexts in which demands for accountability can become so great as to be problematic for the organisation that is expected to give an account. This leads Messner (2009) to argue that the evaluation of accountability should also consider the wider context in which it is being delivered: 'Disembodied forms of accounting need to be complemented with a situation-specific sensitivity ' (2009, p919). This insight is highly relevant for NGOs where account-giving is usually based around impact stories, rather than on financial terms (O'Dwyer & Unerman, 2008).
In a similar vein, Roberts (2009Roberts ( , 2018 puts forward managerial and ethical concerns over demands for too much transparency. Also informed by the work of Butler (2005), Roberts (2009) calls for a more intelligent accountability while discussing the appropriateness of linking transparency and accountability. In doing so he does not suggest that transparency is not important, rather he suggests that relying only on transparency as a form of accountability results in a weakening of the effectiveness of accountability itself. Taking inspiration from O'Neill (2006), Roberts (2009) argues for an 'intelligent' accountability which acknowledges the impossibility of being fully accountable and also considers the ethical burden that an accountable entity has to bear when it is forced to account in a particular way. In more recent work, Roberts (2018) articulates the problems that can occur when organisations try to manage exclusively by meeting the demands of external trans-parency, with little or no regard to the complexities of operations. This is also an important consideration for NGO accountability, where calls for greater accountability usually equate to greater information disclosure based on the assumption that 'information is power', with greater transparency seeming to leverage greater accountability. Furthermore, NGOs also have diverse and often multiple stakeholder demands upon them, but are often limited in capacity to respond adequately to all of these demands.
To date, the extant NGO accountability research has predominantly focused on exploring how accountability to funders becomes problematic for NGOs and affects mission delivery. No specific empirical understanding has been developed of how other factors within the wider external environment of an NGO can also create problems with accountability. Whilst Joannides (2012) undertook an ethnographic study of the day-to-day practices of the Salvation Army. His study provided detailed insight into the notions of internal resistance and accounterability 3 as a response to the problematics of accountability, it did not specifically examine if and how external pressures shape problematic demands for accountability. We therefore believe our study is a timely endeavour as we seek to explore how the volatile environment in which Muslim NGOs operate (see Section 3) is responsible for the way demands for accountability are shaped and experienced within the NGOs.

NGO accountability: external demands and internal responses
In order to conceptualise how demands for accountability are affecting internal organisational activities and processes and subsequently how NGOs are navigating these demands, we utilise a framework proposed by Kearns (1994). Kearns (1994) developed a tentative tool for exploring non-profit accountability. This framework is particularly useful as it explores accountability from both external demands and internal responses in terms of tactical and strategic responses. Frameworks of accountability are particularly scarce in the non-profit/NGO accountability literature, with the focus of prior research being explicitly on conceptualising external accountability processes. There is less work analysing the internal organisational responses to external accountability pressures. Several studies, largely undertaken in the corporate or public sectors (Fry, 1995;Mulgan, 2000;Sinclair, 1995), explain the multiple and diverse ways in which accountability is construed and understood by managers and officers. These studies provide typologies, comprising structure and forms of accountability, to categorise accountability actions. They do not consider how organisations respond to accountability demands within their processes, nor do they specifically consider the external demand for accountability. 4 More recently, Agyemang et al. (2017) utilised the Kearns (1994) framework in the context of NGOs: they sought to ascertain whether and how upward accountability processes can be constraining or enabling in the effective deployment of development aid funding. The Kearns framework allowed Agyemang et al. (2017, p. 987) to map out how external control processes enacted by funders resulted in specific internal responses from aid workers. They argue that 'our mobilization of Kearns' (1994) conceptual insights enables us to conduct a more nuanced analysis of the range of possible responses'. Kearns (1994) provides an analytical frame for considering the reactive and strategic actions taken by non-profit organisations in response to governmental demands in the context of a developed economy. Kearns (1994) addresses controls and accountability from an organisational point of view, and extensively discusses what managers do in response to external controls and pressures. In this regard, we believe key aspects of the framework enable us to examine both the internal responses and the external demands of accountability associated with accountability of Muslim NGOs in the UK. This framework allows us to understand the demands upon the NGOs and subsequently, how they respond, either reactively or proactively, to standards of accountability set by institutions within their strategic environment. These standards can be explicit (i.e. regulatory or contractual) or implicit (behaviour defined by societal values, beliefs and assumptions).
The framework considers four possible internal responses to external demands, two of which Kearns (1994) considers tactical (compliance and negotiated) and two strategic (discretionary and anticipatory). Compliance accountability is reactionary accountability in the face of legislative and contractual obligations. Negotiated accountability refers to reactionary accountability due to changes in societal expectations (Ospina, Diaz, & O'Sullivan, 2002). The factors motivating negotiated accountability 'can arise from a gradual shifting of public sentiment and expectations, stimulating action and forcing the organisation into a mode of damage control' (Kearns,199,p. 189). Discretionary accountability is where an organisation may choose to internalise certain professional standards even when there is no requirement to do so and in the absence of threats or sanctions from its external environment. Finally, anticipatory accountability refers to organisations proactively formulating internal standards in anticipation of changing legislation. It also refers to attempts made by organisations to shape these standards that may eventually be imposed.
In sum, little is known about how accountability can become problematic and how NGOs seek to navigate these problems. In this paper, we answer calls for more situation-and context-specific studies of NGO accountability, to illuminate the continually shifting nature of accountability and its antecedents. We mobilise the conceptual notions introduced above to frame our efforts to understand how accountability demands are experienced within Muslim NGOs. In particular, we explore how 3 Kamuf (2007, p. 253) defines accounterability as ''a counterinstitution of resistance to the irresistible logic of accountability". Given the ambiguity of the notions of resistance, Joannides (2012) interprets accounterability as a reflection on the meaning of accountability and the subsequent working response to the problematics of its operationalisation. 4 For example, Fry (1995) considers accountability from the perspective of intrinsic 'felt' accountability and the external conversations for accountability that can arise from this. Mulgan (2000) considers accountability along the dimensions of responsibility, control and responsiveness. Whilst Sinclair (1995) draws on the structures (or forms) of accountability (i.e. the controls, audits) and the personal discourses of fear, anxiety and vulnerability associated with accountability.
accountability is becoming problematic for Muslim NGOs, how demands for accountability are shaped and subsequently how NGOs respond to these demands.

Context of study: Muslim NGOs operating in the UK
Muslim NGOs in the UK were predominantly established during the eighties and nineties as a direct response to the Bosnia crisis and the Ethiopian famine (Cordier, 2009;Benthall & Bellion-Jourdan, 2009). Many of the largest organisations have grown as a direct result of these and subsequent emergencies and are now involved in all aspects of international relief, local poverty alleviation schemes and advocacy. Today, there are currently over 1600 Muslim charitable organisations registered with the Charity Commission for England and Wales, with a combined annual income of more than £275 million (Charity Commission, 2009). These organisations operate freely in the UK, co-existing and often competing with other charities for funding and resources. Coupled with the findings that Muslim communities give more to charity than any other religious group in Britain (Ainsworth, 2013), the Muslim charitable sector is one of the fastest-growing areas in the charity sector. Of the top 50 Muslim NGOs in the sector, the largest has an income greater than £60 million and is also one of the founding members of the Disasters Emergency Committee [DEC]. 5 Muslim charities offer the same services as other NGOs and are likely to engage in 'activities to relieve suffering, promote the interests of the poor, protect the environment, provide basic social services, or undertake community development' (World Bank, 2001).
The war on terror 6 and subsequent discussions on terrorism, 7 anti-terror initiatives and counter-terror legislation has had a substantial effect on the accountability debate surrounding Muslim NGOs. These effects have been both intended and otherwise. The United Nations imposed unprecedented counter-terrorism obligations on its 189 member states, mandating that they freeze the assets and restrict the movement of designated terrorists and their supporters (Benthall, 2007;Benthall & Bellion-Jourdan, 2009). This resulted in the increased scrutiny by the banking sector of all funds coming into and out of Islamic organisations. This added scrutiny had a direct impact on international fund transfers and the ability to ensure that money is delivered for projects in host countries on time (Metcalfe-Hough, Keating, & Pantuliano, 2015). Muslim NGOs that were not able to address these pressures, closed or were forcibly shut down. Other NGOs realised the need to become more open and transparent in their activities and engaged various means and avenues to do this (Cordier, 2009).
In Sweden, the Montreux Initiative was launched in 2005, sponsored by the Swiss Federal Department of Foreign Affairs as a joint confidence building exercise between Muslims and non-Muslims to 'remove unjustified obstacles from Islamic charities through a system of self-regulation and monitoring' (Benthall & Bellion-Jourdan, 2009, p. xiii). In the UK, the Muslim Charities Forum [MCF] was launched in 2007 as an umbrella organisation of Muslim NGOs to engage in advocacy, improving governance, research and development and providing training services to other Muslim NGOs. As a measure of their success, a recent survey conducted by MCF found that 88% of Muslim NGOs had regulations in place to prevent the financing of terrorism and had adopted internal measures to protect against terrorist abuse (El-Banna, 2014). Despite these measures, 11% of all investigative enquiries undertaken by the Charities Commission between 2005 and 2012 involved Muslim NGOs (Charity Commission, 2012), although they make up only 5% of the sector. The National Council for Voluntary Organisations has suggested that the Commission is inappropriately entering the political arena, and is allowing the State's agenda on anti-terrorism to influence the type of statutory inquiries it opens on Muslim charities (NCVO, 2015). A report produced by a think tank, Claystone, raised concerns that Muslim NGOs were being 'disproportionately investigated and monitored' (Belaon, 2014;Ramesh, 2014).
In a similar vein, scholars in the fields of criminology and law have noted that the general feeling among Muslims in the West is one of a heightened sense of anxiety and belief that their communities are 'under siege' (Aly & Balnaves, 2008;Cainkar, 2009;Jamil & Rousseau, 2012). Cherney and Murphy (2015) note one feature of this state is the belief that majority groups (i.e. non-Muslims) and institutional authorities hold implicit biases about their group (i.e. Muslims) and attribute negative behavioural intentions to them. According to Breen-Smyth (2014), the discourse and practice surrounding these biases influences how Muslims see themselves and the attributions they attach to others. Such perceptions can also influence the way demands for accountability are received, and generate a number of responses, both individually and institutionally (Uz & Kemmelmeier, 2014).
It is in this environment that we attempt to understand how the unique context of Muslim NGOs shapes their accountability experience, and in particular how accountability demands are affecting their organisational practice. 5 The DEC was formed in 1963 as an umbrella group of UK charities to raise money at times of humanitarian crisis in poorer countries. It has 14 members from leading UK aid agencies and has raised over £1.4 billion since its inception. 6 The War on Terror, also known as the Global War on Terrorism, is an international military campaign that was launched by the United States government after the attacks on the New York World Trade Centre on 11th September 2001. 7 Terrorism is often used in the context of political, ideological or religious violence, however the debate surrounding terrorism and what constitutes a terrorist is complex and changing. For an excellent discussion on terrorism as a social construction please see Jenkins (2003). Moreover, Norris, Kern and Just (2003) also provide an extremely effective critique on how the Government and the media frame our understanding of terrorism.

Research approach
In this paper, we attempt to add to the empirical literature on NGO accountability and respond to calls for greater situation-specific studies of NGO accountability processes, mechanisms and responses. We choose to situate our research on Muslim NGOs operating in the UK as they act as ideal 'expressive' cases (Berry, 2005). 8 Muslim NGOs can be considered an expressive case for two main reasons. Firstly, Muslim NGOs are faith-based entities operating in a secular/different religious orientated society, suggesting that they are socially in some way different and thus 'set-apart' from their immediate environment. Furthermore, as NGOs, they are likely to be deeply entrenched in social concerns compared to their for-profit counterparts and thus any prevailing socio-political factors may have a greater effect on their operations. Secondly, the ongoing geo-political volatility in relation to their Islamic identity (as outlined in Section 3) and subsequent close links (in terms of funding, employees, donor base, and beneficiaries) to the Muslim world, has caused them to be more susceptible to calls for greater accountability (see for example, Belaon 2014; Uz & Kemmelmeier, 2014). Thus, we believe that the Muslim NGO sector provides a rich expressive case for us to understand how prevailing socio-political factors help shape accountability demands and in turn illustrate how accountability is becoming problematic for NGOs.
Following Gioia, Corley and Hamilton (2012, p. 17), we make 'extraordinary efforts to give voice to the informants in the early stages of the data gathering and also to represent their voices prominently in the reporting of the research'. This created an opportunity for us to discover new concepts rather than just affirm existing concepts. For example, we purposefully chose not to use the term 'accountability' within our data collection; we wanted to see what terms the interviewees themselves would use when talking of their reporting and governance practices. Yet, the term accountability was talked of more often than the terms we chose. Furthermore, we chose to keep the data collection process deliberately iterative and not guided by any specific theoretical framework in the initial stages because, consistent with Gioia et al. (2012) andO'Dwyer (2004), we believe that standardising the interview protocol would hinder the emergence of new themes and concepts from developing. For example, we chose to begin our research with the open idea of trying to understand accountability as experienced by Muslim NGOs operating in the UK. The idea that accountability was in some ways problematic for Muslim NGOs emerged from the initial interviews and subsequent interviews then built around this conceptual discovery, as explained below.

Data collection process
The analysis here draws upon empirical material from eighteen semi-structured interviews with individuals from three Muslim charities, based in London, Birmingham and Bradford, conducted in 2012 and 2013. The organisations chosen from which respondents were drawn include two of the largest and most proactive Muslim NGOs operating in the UK (detailed overview of these organisations can be found in Appendix A). Both these organisations were founder over 25 years ago and have established formal accountability and governance systems. They also have greater exposure and are therefore likely to have more specific accountability experiences than other smaller organisations. The third organisation is a smaller NGO which is undergoing a process of strengthening its governance systems. These organisations were chosen as it was envisaged the research interest would be transparently observable (Pettigrew, 1988) and would provide a range of views on how accountability was experienced in Muslim NGOs. Access was negotiated through informal contacts in the first instance. This was done through email and telephone contact to introduce the researcher and ascertain potential interest of the organisations in the research. This process took a few months and a total of eight of the top Muslim charities in the UK (based on income) were contacted.
The interviewees consisted mainly of trustees and senior management; in addition, four external stakeholders were also interviewed. These included an accountant working for organisation A, two external consultants to the sector and a research advisor working for the MCF. The interviews were conducted face-to-face with interviewees at their place of work and lasted at the most 100 minutes. Interviewees were promised confidentiality and signed a consent form agreeing to be recorded. More details about the interviewees are provided in Table 1. To preserve confidentiality, the interviewees are not matched to the NGOs.
The interview questions were formulated from prior literature examining NGO accountability mechanisms. 9 Therefore, questions were asked around reporting, auditing and governance mechanisms, and about where the organisational members perceived demands for these accountabilities were coming from. The specific purpose was to get their experiences and thoughts about how they viewed and responded to accountability. The questions were revised during interviews as new themes became apparent (Huberman & Miles, 1994;O'Dwyer, 2004). The interviewing process was conducted as a conversation to make interviewees feel at ease with the interviewer (Shah, 2004).
Findings from interviews are supplemented by in-depth documentary analysis of publicly available documents, including news articles from mainstream news outlets in the UK, annual reports of the NGOs and parliamentary evidence documents given in relation to the Draft Protection of Charities Bill 2014 [DPCB]. Information from these documents was used to both 8 An expressive case is described as a case neutralising some functional issues in order to bring to light phenomenon that otherwise would not have been readily observable. 9 The interview schedule is available from authors upon request.
identify themes to be explored in interviews and to confirm and enrich understanding of interview findings. More details are provided in Table 2. The transcriptions generated 210 pages of data (single spaced, font 12). The field notes and interview transcripts were initially analysed to ascertain any common themes in relation to how accountability demands were being experienced and organisational responses thereto. As the data was being collected it was analysed inductively, first by the initial interviewer and later by the second researcher. A set of codes was developed based on the main questions and sub-questions contained in the interview guide (Huberman & Miles, 1994;Ryan & Bernard, 2003). What quickly emerged was a perception that some accountability processes were in some way problematic to the mission and day-to-day activities of the organisations. These codes were subsequently re-analysed and eventually collapsed into the following broad themes: importance of compliance, suspicion from media and regulator, working under constraints, curtailment of activities, developing positive image, engaging in strategic change. These themes were both perceived and real and formed the basis of the preliminary descriptive analysis of the findings. Using these themes, we then sought to gain additional insight from documentary evidence. Specifically, we cross-checked claims made by interviewees in newspaper archives and vice-versa. We undertook a detailed analysis of annual reports of the NGOs to explore specific reporting and accounting practice. Further, postinterview we analysed parliamentary debates from the draft protection of the DPCB to ascertain the role being played by Muslim NGOs in lobbying activities.  These insights were fed into the empirical analysis in order to refine aspects of the preliminary analysis. Subsequent analysis therefore drew on both the interview findings and analysis of documentary evidence to allow a more focussed theoretical framing to emerge around the concept of problematic of accountability. The iterative process of listening to the audio of the interviews, re-reading the transcriptions and referring back to the core literature driving the study (latterly focussed especially on Kearns, 1994), allowed an understanding to emerge of how accountability demands were being experienced and how organisations responded to them (Locke, 1996). Continual redrafting, reanalysis and interaction between the data and core informing literature was undertaken to craft the case analysis presented in the next section (Appendix B provides an overview of the final coding matrix). The empirical findings and discussion section that follows is organised around these patterns. Illustrative quotations from transcripts have been selected to exemplify themes and to reflect a range of opinions.
The researchers' own experiences, perceptions, values and prejudices are likely to impact both the research process and the interpretation of the data, and this is largely unavoidable in qualitative research (Seale, 1999). In addition, there is a possibility that in some cases the interviewees may have provided an 'official line' to their answers, and therefore may not be fully reflective of reality. This possibility has been considered when analysing the interviews, with the analysis also critically considering the context in which answers are given, rather than just what is being said (Sinclair, 1995). Reliability and validity were ensured by making certain that each step of the data collection was documented by the authors and all transcripts were codified and the codes were used in the final analysis (Bartlett & Payne, 1997). Interviews were transcribed as soon as possible and then re-checked for accuracy by the interviewers, and interviewees were sent a draft of the final report for verification.
The following section presents and discusses the empirical findings.

Findings
The narrative in this section is structured according to the objectives of the study and the themes emerging from the data analysis process described earlier. We describe the specific 'problematics of accountability' that Muslim NGOs are facing and argue that formal compliance with Charity Commission requirements is no longer enough in terms of establishing the legitimacy of their activities. Drawing on Butler (2005) and Messner (2009), we suggest that Muslim NGOs have become 'exposed organisations', where their formal accountability is no longer effective and is met with suspicion, partly in response to the climate created by the media. We then go on to explore how the NGOs seek to respond to these pressures through proactive attempts to counter negative public perceptions and shape future policy concerns. We begin by exploring the importance of accountability for Muslim NGOs. We then go on to analyse how accountability is becoming problematic for them, which then leads to a discussion of the effect these problems are having on the activities and operations of the NGOs. Finally, we explore the tactical and strategic actions the NGOs are undertaking in order to navigate these problematics.

Accountability is vital for organisational survival
Muslim NGOs have spent the last few decades on developing their charitable identity, and providing a similar façade to that of Western NGOs (Cordier, 2009). The Muslim NGOs that took part in our research also had similar structures and processes to similar sized non-Muslim NGOs operating in the area of international aid (see Appendix A). We find that these NGOs use the same technical-logistical jargon, mobilise large-scale fundraising and undertake required communication, both legal and otherwise. For example, all three NGOs were registered with the Charity Commission for England and Wales and annual reports show that they all follow the requisite reporting and accounting regulations. All of the NGOs viewed reporting as an integral organisational process, especially in anchoring them within their wider accountability responsibilities. The accountees seemed to view the process and mechanisms of reporting as a legitimising device. Both finance directors agreed: 'Apart from it being a statutory requirement, it ensures effectiveness, it ensures proper procedures are maintained and stops any chance of corruption. This helps us in case things go wrong'.
[8] 'I think it is absolutely key. . . not only to the organisation but to the sector as a whole. These financial statements get audited so this gives assurance to your donors. . . to the institutions. . . that the organisation is being run effectively. . . and that the funding they are providing is being utilised as correctly as required. ' [3] The two directors were very much aware of the need to publish accounts and audited reports on the Charity Commission website, seeing the associated transparency as an important aspect of their duties. One senior manager noted: 'You do need to have this bureaucracy [of annual reporting and auditing] for the sake of accountability and transparency.' [6] Furthermore, all the interviewees were very keen to point out their detailed professional websites, user-friendly public annual reports and various fundraising initiatives and amounts raised, stating: 'We were able to raise a quarter of a million pounds. The fundraising has been a very sophisticated marketing technique that we have employed... very cost effectively and very efficiently. '[8] 'I always kept records, even when we were on a voluntary basis, things were happening and I was the one who was keeping the record of exactly what was happening, where the money was coming from, where was it going to go, etc. So in our second year, we raised almost a million pounds. '[4] Two of the three NGOs also had a governance and compliance function. This is expected as many of the governance structures for UK-based Muslim NGOs are modelled on organisations such as CAFOD (the Catholic Agency for Overseas Development) and Catholic Aid (Cordier, 2009). Therefore, Muslim NGOs were engaging in all the ''correct" accountability 'talk and action' that was minimally required of them (Dhanani & Connolly, 2014;Yasmin, Haniffa, & Hudaib, 2014).
All organisations were also voluntary members of various accountability bodies, as is the case with all major NGOs (Ebrahim, 2003b). Adherence to these bodies allowed the NGOs to engage in discretionary accountability processes. Kearns (1994) suggested that discretionary accountability arises when organisations themselves take responsibility for identifying and codifying acceptable standards of practice. This was viewed as helping the organisations focus their activities and strengthen their internal governance procedures. As one senior manager suggested: 'they provide a form of external monitoring currently lacking in the sector'. These voluntary membership bodies include the Department for International Development, European Commission for Humanitarian Organisations, and the International Organisation for Migration. All the organisations were also signatories of various standards/benchmark bodies such as the International Commission of the Red Cross, The Sphere Standards, The Fundraising Standards Board, Humanitarian Accountability Partnership, Bond and National Council for Voluntary Organisations. Membership of these bodies was a step towards professionalisation for these organisations (Cordier, 2009) and reflects a need to gain recognition and legitimacy in the eyes of institutional donors and the international aid sector. As accountees suggest: 'Our financial audits are carried out by PWC, our program evaluation is carried out by KPMG, so big brands of evaluation that carry a huge amount of responsibility. We go with them which means tougher evaluation, tougher scrutiny and the end if it comes out good it improves our reputation.'[7] 'Externally we are a candidate for DEC membership so we have adopted the DEC accountability framework and we are implementing it as if we are now a DEC member so we have that accountability framework. ' [2] Additionally, to be seen to be compliant the accountees revealed how they took actions to prioritise the outputs and processes required by their upward accountors -that is, the Charity Commission and institutional funders. Thus, oversight in the form of legislation and funder contractual obligations provided the impetus for improving governance in all organisations, with senior managers viewing the 'consequential bureaucracy' and 'added cost' as necessities. For example, all accountees highlighted how they used sophisticated technology, in terms of software programmes, the internet and the electronic banking system, to implement and monitor controls over their resources. This allowed the organisations to provide detailed, technical reports demonstrating compliance. Kearns (1994) identifies compliance as a tactical internal response to accountability demands that operate as external control systems. Our analysis suggests that accountees were continually anxious to demonstrate compliance, but also keen to strengthen compliance by engaging in discretionary strategic actions. For some of those interviewed, the impetus for this technological advancement and voluntary membership of oversight bodies was a direct result of feeling that they were not trusted by their upward accountors and therefore must engage in discretionary accountability activities in order to demonstrate their compliance. A consultant and senior manager stated: '. . . nowadays we have to be very careful. Muslim charities especially need to show where the funds are going. This has increased costs, because there are more procedures, more rules, more oversight now.' [16] 'We don't deal with cash payments and we don't transfer the money through any source other than direct bank transfers. If we cannot do that we do not work in that country as we need to leave an audit trail.' [3] Our findings therefore suggest that tactical compliance accountability in respect of legislative and contractual obligations was seen as highly important. Yet, it was being proactively bolstered by strategic discretionary accountability processes, where the NGOs chose to internalise voluntary professional standards even when there was no requirement to do so.

The creation of a problematic accountability
This section explores how and why demands for accountability have become problematic for Muslim NGOs. While undertaking the interviews and analysing the information collected, it became apparent that accountees felt the counter-terror legislation, and narrative surrounding this, was influencing the attitudes of the Charity Commission towards them. Furthermore, there was a sense that the counter-terror narrative in the media was shaping political decision-making in relation to Muslim NGOs and consequently this limited or undermined the effectiveness of their formal transparency and accountability. In other words, the NGOs felt their tactical reactive responses to compliance were not being viewed as providing enough accountability.
Our findings suggest that Muslim NGOs have a complicated relationship with the media. On one hand, the media has always been vitally important for them for fundraising purposes, something both senior management and junior staff agreed: 'Usually the general public is moved more by emergencies in the media. If they are in the media then they will come forward and give. . . ' [12] 'The only time these countries get any media attention is if there's a severe drought or a famine. Prior to that they would get very little attention, so we are working in some of these countries to raise their profile. ' [9] 'Some countries are in desperate need but because they're not in the media and they are not being commonly talked about they are not getting much funding. ' [10] On the other hand, analysis of news archives suggests that right-wing media outlets have been heavily critical of Muslim NGOs, 10 linking Muslim NGOs to terrorist funding (Benthall, 2016;Cordier, 2009). The international relations literature suggests that this has increased in recent time, as the Western world has seen a rise in Islamist-related terrorist activity (Heath-Keely, 2013), as well as an increase in right-wing populist politics, causing fear of the 'other'. 11 This has created a scenario where, as Said (1997) notes, reference to Islam and Muslims is often news of an unpleasant sort. Some media outlets have run specific anti-Muslim NGO stories in the past which were subsequently found to have been either erroneous or false. For example, the Daily Telegraph published a story which accused the MCF, of having alleged extremism links (Turner, 2014). This was despite no such claims being proven and the MCF itself working with the government on a broader strategy to counter extremism. This led Ben Jackson, the chief executive of Bond (the NGO advocacy body) to assert this was symptomatic of a wider pattern of un-evidenced assertions being made in relation to Muslim NGOs (Delmar-Morgan, 2015). Subsequently, a MCFsponsored event at the Conservative party's annual conference in 2015 was cancelled after the Sunday Telegraph questioned why the MCF was allowed to take part. This decision, according to Sir Stephen Bubb, chief executive of ACEVO, was an attempt to 'demonise' the Muslim community (ACEVO, 2015).
There was a feeling among accountees that hostile media reports were affecting their organisational relationships with their regulator, the Charity Commission. All three organisations have had some form of investigation conducted into their activities on the basis of complaints. Although these investigations ultimately led to improvements in organisational practice: 'We have had the Charity Commission come and do this because we were investigated for an allegation in 2010. In the process they did tell us a lot of things that we were not doing. We were not maintaining electronic records of our funding agreements. We were filing reports without reviewing them.' [2] 'The Charity Commission said this is not enough, we've got to see the [funding] report, if it's not good then it will need to go back'.
[15] 'One question was whether we got the Charity Commission's approval on our Constitution. They also wanted to know whether our governance includes women. It doesn't at the moment but the trustees are very much conscious about this and they want to bring in more women trustees. ' [4] A worry amongst accountees stemmed from what they viewed as politically motivated rhetoric by the Charity Commission and a perception that they were not fully trusted by the regulator. Some interviewees pointed to statements made by the chairman of the Charity Commission, who said that Islamist extremism was the 'most-deadly' problem the charity sector faced, despite Muslim charities making up less than 5% of the total sector. This subsequently led to various independent bodies (e.g. ACEVO, 2015;Belaon, 2014;Third Sector, 2014) stating that the Charity Commission was unduly exaggerating the number of charities at risk of Islamic extremism. This was an important concern for the NGOs as they felt that overstating the risk was undermining the credibility of the sector as a whole (Home Office & HM Treasury, 2007). A lack of a transparent and open dialogue between the NGOs and the Commission was exacerbating the problems further. Thus, it was important for the NGOs to demonstrate their commitment to transparency and compliance, a desire derived from a long-standing fear of being viewed as suspect (Belaon, 2014;Cordier, 2009) as well as from evidence suggesting that Muslim NGOs are not as transparent in their reporting as they could be (Yasmin et al., 2014).
The direct cost an innocent charity faces in responding to a mistaken Charity Commission investigation is also substantial (Third Sector, 2014). As the Commission tends to report their inquiries into named charities before the findings are published, this can cause adverse repercussions, in the form of drop in donations and hostile media reporting, even for those charities later found innocent. As noted by the CEO of ACEVO giving evidence at the parliamentary DPCB hearing: 'To give you an example of the problem that was created by the Charity Commission, [NGO X] found two problems in their international development work in Sudan and the Gambia in terms of fraud. They reported those problems to the Charity Commission. As a result of that, a statutory inquiry was launched and it was publicly announced. Immediately that raises issues for the charity in terms of perception. They have had significant problems. That was done during the main Muslim charity-giving period, which is during Ramadan.' 10 'Charities linked to terrorism at record high: extremists pose deadly threat by abusing fundraising groups, warns commission chief' (The Telegraph, 17 Jan 2017); 'Government donation to Muslim Charities Forum denounced as "madness"' (The Telegraph, 23 Sep 2014); 'Islamic charities in UK fear they are being unfairly targeted over extremism' (The Guardian, 22 July 2015); 'Charity leaders appalled by Muslim ban at Conservative Party Fringe' (ACEVO, 2015). 11 It is important to mention here that the "othering" of Muslims in the West is not only due to increased violent incidents in the west. The work of Orientalism by Said (1997) shows the context of media/public /popular/intellectual and political hostile discourses about Islam in the West goes back in history to before the events of 9/11. Space in the paper does not allow for this historical evaluation of populist hostile discourses and suspicion of Islam to be fully explored yet, it does help better explain the "limits of accountability" faced by Muslim NGOs.
A senior manager at the NGO concerned also stated: 'This definitely affected our funds and our reputation. It had a knee-jerk effect not only on our Muslim donors but on other donors. ' [5] Furthermore, there was also a perceived lack of guidance from the Charity Commission surrounding the applicability of legislation causing the most harm. The CEO of MCF giving evidence at the parliamentary DPCB hearing noted: 'How can you have access through those people without being listed? You need to give us guidance, because we are on the ground and it is difficult to see people dying every day when we can't have access because this individual or this group are preventing us. We do not have any regulation or any guidance to deal with them.' The accountees accepted that it was important for the Commission to take issues of corruption seriously, yet they also felt it was vital that this does not hinder their main activity to help those in need. This is especially important given that local or international Muslim NGOs are sometimes the only groups that can get aid through to isolated areas (Benthall & Bellion-Jourdan, 2009;Benthall, 2016;Cordier, 2009). All accountees highlighted how they already had procedures in place, in line with UK fundraising requirements, to stop unknown persons raising funds for them. These procedures included vetting procedures for trustees and volunteers and specific guidance on expenditure, which had been approved and is audited.
Consequently, we find upward accountability pressures (Edwards & Hulmes, 1996) on Muslim NGOs materialising from unexpected quarters. In particular, we find the right-wing media and Muslim related terrorist activity to be influencing the accountability demands upon Muslim NGOs via the counter-terror legislation and the actions of the Charity Commission. According to O'Dwyer and Unerman (2008), hierarchical accountability favours accountability to those stakeholders who control access to key resources for both resource use and immediate impact (Ebrahim, 2003a(Ebrahim, , 2003bEdwards & Hulme, 1996). In our case, the right-wing media although not a direct accountor, did not directly control access to resources but attempted to manipulate those that did. Roberts (2009Roberts ( , p. 1549 conceives hierarchical accountability as a form of external oversight and control where organisational managers must continually strive to demonstrate 'performance'. Yet, in our case we found the accountability demands were not perceived by accountees to be linked to their individual performance, but linked to organisational identity. Thus, apprehension regarding the power of external others to deny possible mission accomplishment, a key characteristic of hierarchical forms of accountability (Najam, 1996) was strongly evident.

Muslim NGOs as politically exposed organisations
The main accountors, in a normal hierarchical accountability relationship for any NGO, are the funders who provide resources and the regulator who ensures correct applicability of procedures and processes. For Muslim NGOs key accountability relationships include the government/regulator and the Muslim community [who are both their beneficiaries (abroad) and donors (in the UK)]. Our findings suggest the organisations believe they are being perceived, because of their Islamic identity, as lacking accountability and being non-transparent. This leads us to suggest that Muslim NGOs view themselves as politically vulnerable organisations. The accountees in Muslim NGOs believe the accounts they provide, either to the regulator or the media, are viewed as incomplete and unconvincing. This is particularly important given that Pantazis and Pemberton (2009) note the Muslim community largely views government rhetoric with suspicion due to everincreasing legislation perceived to be targeting the community. Accountees also expressed these sentiments and suggested that Muslim NGOs have to be increasingly careful who they interact with, especially within the Muslim community, and this is having an impact on their recruitment and donations: 'We are now moving towards a system where we are doing criminal record checks on people: For all volunteers and wherever possible and wherever needed.' [3] 'We even have a system where any donor that gives more than £5000, straightaway we have to do a CTF check. That's counterterrorism and fraud, as we have to check to see the money is legitimate. ' [13] We therefore present the notion that Muslim NGOs are becoming exposed organisations. We suggest that demands for accountability, coming from counter-terror legislation and a politically motivated regulator are influenced by an anti-Muslim rhetoric. This rhetoric is shaped by wider Muslim-related terror concerns and the right-wing media. This is creating a problematic accountability for Muslim NGOs. Thus, while they meet their formal accountability requirements, this does not satisfy or reassure their critics in the press. Butler (2005) suggests an exposed self is likely to engage in an abundance of account-giving activities. The problematics of accountability arise because the account is received with scepticism and viewed as being incomplete (Messner, 2009). According to Butler (2005), the exposed nature of these organisations means they are closely scrutinised, their behaviour and activities are commented upon, and whatever they say and do is received with scepticism. In a similar vein, our findings suggest that Muslim NGOs, operating within the current socio-political discourse, which views their Islamic identity with suspicion, can become politically exposed to such limitations and therefore the normal routines of accountability no longer serve to satisfy or appease the regulator, or reassure the public. This suggests accountability demands are of a differing nature for Muslim NGOs than for similar non-Muslim NGOs. We now turn to analysing the effect these demands are having upon the organisational practice of Muslim NGOs.

Problematics of accountability in action: effect on activities and operations
The findings of this study suggest a number of areas where the demands for greater accountability, brought on by counter-terror legislation, were having a detrimental effect on the activities and internal operations of Muslim NGOs. Muslim NGOs by their very nature tend to operate in Muslim majority areas, and arguably areas which also have the greatest risk (Benthall, 2016). The wider climate of suspicion and ongoing legislative changes, as discussed earlier, are also having a detrimental impact on what they can and cannot do. For example, global and national policy-makers in the United Nations, national Treasury Departments, and in particular the Financial Action Task Force, suggest the not-for-profit sector is particularly vulnerable to abuse by terrorists (FATF, 2014).
The organisations interviewed expressed huge frustration with the various anti-terror legislative regulations, with managers and trustees stating it is a: 'logistical, legal and financial nightmare dealing with those measures' [1] and 'I think counter-terrorism legislation is preventing us from having access to the neediest people ' [4].
This suggests that the legislation was not only affecting their internal operation but also their humanitarian activities. Our findings suggest that the main concerns are linked to the uncertainty in the wording of the legislation and its subsequent applicability to NGO operations.
The Terrorism Act 2000 and its subsequent amendments sets out offences which could be deemed to be committed when meeting representatives from organisations designated under the Act. According to those giving evidence at the parliamentary DPCB hearing, the territorial applications of the Act are unclear and could be interpreted very broadly. 12 In particular, the Chief Executive of Bond suggests that the uncertainties can have a 'chilling effect' on the choices made by NGOs who might come into contact with such groups. The UK Independent Reviewer of Terrorism Legislation (2014) recognises that charities are likely to come into contact with terrorist groups even while acting legitimately. For example, interviewees suggested this could happen when they are delivering aid to areas where certain groups may have political and military control or, in situations when they are delivering life-saving aid, there is a risk that some of those receiving aid includes individuals who have been designated as terrorists, or who have links to designated individuals or groups. As the CEO of the MCF and Chief Executive of Bond both note in the parliamentary DPCB hearing: 'There are proscribed groups in those areas, and we know them. They are the gatekeepers. How can we go through the gatekeepers to reach the neediest people in Syria, Somalia or different parts of the world? This is very important.' And an external stakeholder also suggested: 'If you are working in Gaza and you are taking anything in through the crossings, you have to take account of Israeli authorities and the authorities in Gaza. The authorities in Gaza are Hamas; that is just a fact of life. ' [17] Furthermore, our analysis suggests that other operational limitations specific to Muslim NGOs have come in the form of restricted access to financial services. To comply with counter-terrorism regulations, UK banks have engaged in risk management strategies and de-risking activities. De-risking refers to the practice of financial institutions exiting relationships with and closing the accounts of clients perceived to be 'high risk.' Rather than manage these risky clients, financial institutions opt to end the relationship altogether, consequently minimising their own risk exposure while leaving clients bankless (Durner & Shetret, 2015). Analysis of news archives suggests this practice has had a tremendous effect on Muslim specific individuals and organisations, 13 and particularly on the NGO sector as many Muslim NGOs operate in high risk regions. It has therefore become increasingly challenging for Muslim NGOs to transfer funds for humanitarian work through formal banking channels to support operations abroad, with accountees noting: 'We truly understand the importance of this but the nature of our work sometimes needs us to respond quickly to the events to save lives. . .the funds need to reach them during the most crucial time when it is needed. ' [11] 'For the past 10 years this has had a significant impact in terms of finance for Islamic Muslim charities more than it has on other charities. Being Islamic or Muslim in orientation means funds coming in and going out of the organisation would face more scrutiny from the banking sector than others. ' [18] NGO A, specifically, was affected by this issue when their accounts were closed by HSBC without warning in early 2016. This resulted in them stating on their website: 'UK payments to others who bank with HSBC were held up for weeks or even months -including a payment to procure tents urgently for earthquake victims in Nepal. And a number of donors who bank with HSBC had donations to [Org A] blocked.' As they had already been made aware of HSBC's reluctance to provide service to them, NGO A had already moved the majority of its banking to another financial services provider in the preceding year. These developments are particularly detrimental, because they not only damage the activities and operations of those organisations affected but they also expose donor assets to greater risk, as highlighted by Bond to the parliamentary DPCB hearing: 'If NGOs cannot transfer funds to 'high risk' destinations then informal mechanisms -use of money service bureaux (MSBs), or carrying cash across borders -will replace the safer and more transparent transfer of funds through the banking system. . ..' In some cases, it is alleged this had also led to a direct loss of life: 'In the Somali famine in 2011, we heard . . .. that they lost 260,000 people. This was because of the bank restrictions. It didn't happen in the famine before that.' [CEO of MCF giving evidence to the parliamentary DPCB hearing] Representatives at the parliamentary DPCB hearing also expressed unease at the way in which organisations were singled out for de-risking checks, as both individuals and organisations are usually checked against: 'formal and informal counter-terrorism lists, media reports etc in the UK and overseas. A charity will be treated as suspect if mentioned in an unproved media report, a Charity Commission warning, an overseas government's listing of a UK Government-funded charity etc. There is no effective appeal against this.' [Bond representative] A recent report by the BBC (2015) found a standard software 'World-check' (owned by Thomson) is used by 49 out of the 50 British Banks to screen their customers. However, serious concerns were raised in this report over the reliability of the sources used by the software. This has been given more precedence in recent years as the financial and reputational costs of being complicit in such cases are huge for the banks concerned. Kearns (1994) discusses anticipatory/positioning accountability as involving agents anticipating the formulation of new standards in order to better position themselves for compliance. In our case, the accountees suggest that they are not able to comply even if they were able to anticipate a change, due to them not being trusted (Messner, 2009).
Given these concerns, we also find there is still much work to be done on behalf of the NGOs themselves to reduce their internal risk and thereby anticipate accountability pressures. Although all accountees discussed risk management, and have risk management policies in place, we found poor understanding of the risk of reducing imitation and fake volunteer cash collections. This could be because such issues are difficult to fully manage and control. For example, accountees in NGO B highlighted how it had recently had its name cleared in a terrorist funding related enquiry, where individuals had used their name to raise funds for suspected terrorist activity. The organisation itself notified the police after it was made aware of this. The organisation was subsequently named in the media as 'supporting terrorist causes' before the enquiry had concluded, which resulted in a substantial loss of income for them. Senior management stated: 'It was very difficult because we lost a lot of income and funding.' [2] 'It was guilty until proven innocent. ' [6] The organisation was later fully cleared of any wrongdoing. The accountees were keen to highlight that they have since tightened their volunteer registration rules and do not engage in street cash collections any more. Risks associated with fraudulent fundraising and imitation can be difficult to mitigate for any NGO which deals with cash fundraising; this is despite adhering to the specific rules set by the fundraising standard body to mitigate common risks associated with such collections (Benthall, 2016). One thing that perhaps could be done by Muslim NGOs is to have a publicised blanket ban on allowing fundraising by individuals in the name of the organisation. A specific focus on only large-scale fundraising initiatives may help to counter this risk. However, the charities may then be in danger of losing out on substantial donor income that comes from impromptu donations.
Our analysis suggests that the issues highlighted here seem to be particular to Muslim NGOs compared to other faithbased NGOs. As research in this area is particularly scarce, it is difficult to state this with certainty. Muslim NGOs themselves are aware of the unique position they are in and there was a realisation among the accountees for a need to undertake topdown strategic change in their mission and governance boards to ensure they were able to handle external pressures and overcome these problems specific to the Muslim NGO sector. Continuing this discussion, the following section outlines how the NGOs have engaged with various strategic and tactical measures in an attempt to overcome these difficulties and move to building trust and influence.

Negotiating and anticipating accountability: steps towards building trust and influence
The organisations were found to be active in trying to manage the external perceptions on the Muslim NGO sector through various negotiatory and anticipatory processes. For example, to counter their negative image, reputational concerns were found to be prominent among accountees, with all organisations having a PR/Media relations person/department to specifically deal with media-related queries as an attempt to influence the ways in which they were being represented. Kearns (1994) suggests negotiated accountability processes arise when implicit accountability standards emerge from shifting societal expectations, thereby allowing for some reactive, tactical actions through negotiation regarding the standards to which organisations will be held to account (Agyemang et al., 2017). Consequently, organisational members stated how they were much more aware of the need to prove their transparency and accountability: 'We are currently doing a mini campaign, trying to address all the negative stories through our publications, etc.' [10] 'Because we're Islamic. . . it's a double edged sword. Being Islamic means you get greater scrutiny but it also means you can work in areas where other charities can't work. ' [11] Thus, the organisations were choosing to become more proactive in formulating their public image. Processes linked to negotiating accountability were therefore linked to gaining political and media approval.
The organisations were also found to be building closer relationships with the government and regulator. Enacting funding partnerships with large funder bodies, including the UN and the DEC, and maintaining close relations with the Department for International Development was an important factor in this process, as it allowed them to anticipate accountability demands to a degree (Kearns, 1994). This was seen to be very important for the organisations as it allowed them a measure of stability in the politically volatile environment. For example, one senior manager noted: 'We have been the government's favourite both Labour and Conservative which means we've transcended past the politics of it. It means that institutionally we believe we have an operation that can withstand environment and political changes.' [3] To do this, the organisation was regularly represented at (and very often invited to) functions and other networking events where key people in power would also be in attendance. Reciprocally, the organisations also invited local MPs and people who they felt portrayed a positive image of the organisation, to their headquarters. These visits were then highlighted on their website and in the local media, as even a junior staff members noted: 'We have relations here with Downing Street, a lot of the key MPs, Prince Charles etc. ' [12] This points to a deliberate strategy on part of both parties to engage each other, with the government in exchange being viewed as not hostile to the Muslim community (Pantazis & Pemberton, 2009). Having said this, accountees were at pains to point out that they did not engage in politically motivated activities at the operational level. In-fact all NGOs had refused to take government funding as part of the 'Prevent' initiative. This funding was provided to organisations to stem the threat of extremism but has been viewed with great suspicion since its inception. Thus, both trustees and senior managers note: 'When the Prevent strategy came, not only did we not get the funding, we refused to apply for it. We did not even engage with it because of its political nature and security agenda. ' [5] 'A lot of government funding is more of a political nature -such as Prevent. This is not something we would even look to apply for. ' [11] This seems partly to appease the wider Muslim community and is evidence of tension between the government and the wider Muslim community, some of whom view the government with extreme suspicion (Pantazis & Pemberton, 2009). Thus, while the organisations were willing to engage with various strategic and tactical responses, they would not do so if it felt it would harm their mission or their relationship with the Muslim community.
The extent to which Muslim NGOs organisations effectively managed their relationships with political players was found to be dependent on their strategic focus. All of the NGOs taking part in the study were in different stages of organisational restructure as a direct response to the changing political and societal environment. This involved the changing of strategy as well as organisational roles and responsibilities: 'Our strategy has shifted from a voluntary based to a proper organisational charity with a proper infrastructure.' [4] 'Our strategy change process was a two-year process which involved outside stakeholder engagement; what others thought about us, from the government, to DFID [the Department for International Development], peers, Christian Aid, Oxfam. . . '. [3] Accountees noted a big push to move into advocacy-related areas. This would not only enable them to maintain closer links with political stakeholders but also allow them a voice on issues that affected the Muslim NGOs directly. Building trust is an important precursor to gaining influence and creating these links was seen as imperative before offering suggestions for change (Agyemang et al., 2017). While Kearns (1994) discusses anticipatory/positioning accountability processes as helping organisations position themselves for compliance, we find the accountees were engaging in anticipatory accountability processes to also give them greater influence: 'The whole sector has changed and we need to think at a different level. It's no longer 'here's my money, and help the needy'. . . People have to start thinking beyond this.' [3] 'Previously, prior to the current strategy, advocacy was not a big player but now advocacy and campaigning are key to moving the whole sector. . . to making real change.' [10] 'What we realised was for us to be received, be recognised in the media, we had to say something unique. ' [5] Thus, strategic change supported by stakeholders was seen as necessary to maintain closer relationship with DFID and the government. This campaigning is itself a mechanism for NGO accountability, providing a new means for being heard, and improving their situation (Bendell, 2000). For example, all accountees stated they had initiated dialogue with government contacts and engaged in lobbying to see if the impact of the counter-terror legislation could be reduced for NGOs. Although this raises policy issues related to the rights and abilities of NGOs to monitor emerging legislation and lobby for favourable outcomes (Hayman et al., 2013), lobbying and activism were deemed important for the NGOs taking part in the study. Some accountees suggested that undertaking joint activist activities through umbrella organisations, such as the MCF, could help the Muslim NGO sector collectively engage with negotiated and anticipatory tactics. The MCF is an umbrella organisation for Muslim NGOs/charities. They meet every few months to initiate various training and development programmes for their members, as well as working closely with the Charity Commission (MCF, 2015). A transparency and governance subcommittee was also recently set up to help members strengthen their reporting and accountability procedures. Although this forum is still at an early stage of development, it could play an instrumental role in helping Muslim NGOs engage collectively with joint issues of concern for the sector. Simons (2005) suggests that, if responsibility for decisionmaking is shared among several people, not only is this likely to increase the information content of the decision made, but it also strengthens the commitment of each person to the decided course of action. If they are able to use advocacy to shape emergent external accountability demands, this can be extremely important in helping Muslim NGOs mediate accountability (Messner, 2009).

Discussion and conclusion
NGOs provide crucial aid delivery to areas of greatest humanitarian need. In the present age this need includes both areas of natural crisis and man-made crisis. Due to their religious and/or cultural affiliations, Muslim NGOs are often better placed to access areas of the world where other NGOs have difficulty in operating. Muslim NGOs also have a comparative advantage because of their long-term presence in areas where, at least until recently, few organisations worked. However, due to the rise in terrorist-related activity originating from these areas, and the wider volatile socio-political environment, these NGOs have become the focus of extra scrutiny and suspicion. The specific aim of this paper was therefore to understand how demands for accountability were shaping organisational practice within Muslim NGOs. Our findings suggest Muslim NGOs place a strong emphasis on accountability and transparency, where compliance accountability processes in the form of Charity Commission reporting requirements and funder specific formal and informal processes, are being supplemented with strategic discretionary processes of accountability. However, our findings also provide a number of insights into how accountability is becoming problematic for Muslim NGOs and how demands for accountability are shaping organisational practice.
First, we find demands for accountability, whilst still emanating from upward accountors, going beyond accounting and reporting concerns. Compared to prior studies (Agyemang et al., 2017;Ebrahim, 2003aEbrahim, , 2003bO'Dwyer and Unerman, 2008) we find Muslim NGOs have a much more complex set of external pressures to manage than other similar NGOs. Our findings suggest the Charity Commission has been adversely influenced by the right wing media and Muslim related terrorist concerns and as a result has been accused of giving undue attention to Muslim NGOs. In this regard the media serves to influence the accountability debate. Consequently, our case analysis also suggests problems with accountability do not necessarily need to be caused by direct demands for accountability. For example, requirements from the counter terrorism legislation, although not directly aimed at Muslim NGOs have had an adverse effect on NGO financial and operational activities, due to their affiliation with the Muslim community and activities in high-risk geographical areas. This insight develops our understanding of how specific NGO accountability regimes are influenced by the prevailing political, social and regulatory context. Second, we suggest for Muslim NGOs, the normal routines of accountability no longer serve to satisfy or appease the regulator, or reassure the public. In this respect, we argue that Muslim NGOs have become politically exposed organisations, such that they are perceived negatively with suspicion and distrust due to their Muslim identity 14 . Exposed organisations are likely to face increasingly onerous accountability demands, which nevertheless will fail to satisfy, or resolve the suspicion. This means their accounts will always be received with scepticism and viewed as being incomplete (Messner, 2009). For Butler (2005, p. 91), one cannot really account for this exposure; it becomes a 'burden' that the self cannot negotiate away. This insight adds an extra dimension to our current understanding of how upward accountability demands can become overly burdensome. In addition, this insight questions the purpose of routine account-giving activities if they cannot satisfy the accountor. Subsequently, in trying to meet this impossible condition, Muslim NGOs were moving beyond compliance to the proactive management of accountability relationships through engagement with various discretionary, negotiated and anticipatory accountability processes. This included enacting strategic change, becoming more proactive in formulating their public image, attempting to build closer relationships with the government and regulator, and actively engaging in lobbying and advocacy campaigns. Thus, whilst on one hand our findings suggest meeting conditions of accountability can be impossible for exposed organizations, we also highlight the proactive steps being taken by such organisations to mitigate or limit the effect of their exposure.
This paper contributes to both the theoretical debate surrounding the problematics of accountability and to the NGO accountability literature. Little is known about how accountability can become problematic and how NGOs in particular seek to navigate these problems. The paper adds to the theoretical debate on the problematics of accountability by articulating how socio-political narratives and populist politics/rhetoric can influence the conversations surrounding accountability 14 This is a reflection of the current political climate both in the UK, Europe and US, where being Muslim itself is cause for suspicion. For example, media coverage surrounding the refugee crisis has been unduly focused on the religion of the migrants rather than their humanitarian needs. Said (1997) argues that this type of coverage of Islam is not actually interpretation but rather an assertion of power. As knowledge and coverage of the Islamic world and Muslims is defined by geopolitics and economic interests, this corresponds to a general understanding of Islam, Muslims and by extension Muslim specific organisations, being not what they actually are, but what ''prominent sectors of a prevalent society take it [them] to be" (Said, 1997, p144). According to Pantazis and Pemberton (2009) the current political discourse has designated Muslims as the new 'enemy within' thereby facilitating the construction of Muslims as a suspect community.
for particular types of organisations. We present the notion of an exposed organisation (Messner, 2009) as one which is perceived negatively with suspicion and distrust. In doing so, we bring new insight into the debate on NGO accountability by arguing that accountability can become problematic for exposed NGOs due to the prevailing socio-political environment in which they operate. We thus build on prior research on NGO accountability by responding to calls to extend examination of accountability within different socio-cultural organisational contexts (see Ebrahim, 2003aEbrahim, , 2003bEbrahim, , 2009Hall & O'Dwyer, 2018;Hopwood, 1983;Martinez & Cooper, 2017). Our findings serve to strengthen the ongoing argument that accountability systems cannot be viewed in isolation from the wider societal, cultural, religious and institutional contexts in which they are delivered. Our findings imply that these issues are even more important, and perhaps much more so, for organisations who are considered 'exposed' because the accountability requirements upon them are disproportionately damaging their mission and activities.
The paper further adds to the understanding of the continually shifting nature of NGO accountability (O'Dwyer & Boomsma, 2015) and the detrimental nature of upward accountability demands (Ebrahim, 2003a(Ebrahim, , 2003b(Ebrahim, , 2009O'Dwyer & Unerman, 2007, 2010 by revealing the extent to which exposed NGOs must act to prove their accountability. We consider the wider social and political context within which demands for NGO accountability evolve (Martinez & Cooper, 2017). We suggest there is a need for more intelligent accountability, which recognises the pursuit of justice and fairness in accounting and accounting regulation (McKernan & Kosmala, 2007). In this regard, we argue the Charity Commission, as the regulator of NGOs in the UK, needs to engage more closely with the moral and ethical dimensions of the account it demands (McKernan, 2012;Messner, 2009). Consequently, as both Messner (2009) and Roberts (2009) posit, the need for an ethical and intelligent accountability that acknowledges the context in which an organisation must give an account, is especially important for exposed organisations.
Finally, we note the extent to which accountability becomes problematic cannot be determined theoretically or a priori. Messner (2009, p. 934), suggests that while 'we can argue that responsibility with respect to the limits of accountability is necessary, we cannot specify in advance what this responsibility entails and what form it takes. For what appears problematic, disturbing, or ambiguous in the particular situation at hand can only be grasped in that very situation.' In this sense, we suggest future studies should attempt to understand how different situations can bring about a problematic accountability and consider the particular nuances and effects this has on organisational practice. There is a rich agenda of research to be continued in this area.

Acknowledgments
We are grateful to Christine Cooper (the editor) and two anonymous reviewers for their constructive comments that have greatly helped us improve our paper. We would like to thank Christopher Humphrey and Brendan O'Dwyer for their helpful comments and suggestions on drafts of the paper. We also acknowledge the comments from conference participants at the 2016 Asia-Pacific Interdisciplinary Perspectives on Accounting [APIRA] Conference (Melbourne).

Declarations of interest
None.

Appendix A. Overview of organisations
NGOs A and B were established in the 1980's as a direct response to the African famine and Bosnian crisis and currently operate worldwide. NGO C was established in early 2000. Table a provides an overview of the income and expenditure of the three organisations from 2010 to 2014. All organisations have headquarters in the UK, with their humanitarian efforts almost exclusively concentrated in the rest of the world. All organisations also have a strong Islamic ethos which shapes their activities and mission, outlined in their official core objectives. NGO A has close to 300 paid employees and over 700 volunteers worldwide. The board of trustees of NGO A has 5 members who do not take part in the day to day running of the organisation. They have an appointed global CEO, country-specific CEOs as well as global operational directors (i.e. Communications, Finance, HR etc.). This organisation has over 25 field offices operating in Europe, Africa, Asia and the Middle East and operates in vital areas such as education, disaster emergency relief, sustainable livelihoods, water solutions, orphan and child welfare etc. Within the UK they have two main offices, one located in central London and one outside of the Capital. Table b provides a more detailed comparison between organisation A and organisation B. NGO B has 19 trustees, of whom a large majority are high-profile persons, able to actively promote the organisation in their respective arenas. These trustees also take an active part in the organisation, with some trustees overseeing key internal committees. Many of their trustees are also trustees of numerous other organisations. They have approximately 60 paid employees in the UK and over 1600 volunteers. This organisation has over 15 field offices operating in Asia, the Middle East and Africa and provides much the same coverage as NGO A. Both organisations rely heavily on institutional funding, with NGO A also having a large percentage of its income coming from its partner organisations overseas via voluntary donations. Of which 60% come from international partners. NGO A has a much larger income than NGOs B and C and also has a more diverse strategic focus. It is heavily focused on advocacy issues and has made a concerted effort to develop itself in this role to make an impact on the wider international development scene. As such the strategic policies of organisation A are more formalised and explicit compared to organisation B. Organisation B has not yet got to the stage of actively re-defining its strategic position. In contrast, NGO C is a smaller organisation with four trustees and only 4 paid staff. Two trustees of the organisation take an active part in the running of the organisation as CEO and accounts manager. The NGO operates primarily in South Asia and the Middle East, but also engaged in emergency relief worldwide.