Audit credibility and LGBTQI rights: certification operation in the margins

Abstract

Purpose

The aim of this paper is to contribute to knowledge about the diversity of credibility arrangements in new audit spaces “in the margins” of auditing and the implications of such arrangements.

Design/methodology/approach

The paper is based on an in-depth qualitative study of the lesbian, gay, bisexual, transgender, queer and intersex (LGBTQI) rights certification run by the Swedish Federation for Lesbian, Gay, Bisexual, Transgender, Queer and Intersex Rights (RFSL) during its first decade of operation. We have interviewed employees and studied documents at the certification units within the RFSL. We have also interviewed certified organizations.

Findings

We highlight two features that explain the unusual credibility arrangements in this audit practice: the role of beneficiaries in the organizational arrangements chosen and the role of responsibility as an organizing value with consequences for responsibility allocation in this certification. These features make it possible for the RFSL to act as a credible auditor even though it deviates from common arrangements for credible audits.

Originality/value

The RFSL certification is different in several ways. First, the RFSL acts as both a trainer and an auditor. Second, the trainers/auditors at the RFSL have no accreditation to guarantee their credibility. Third, the RFSL decides for itself what standards should apply for the certification and adapts these standards to the operation being audited. Therefore, this case provides a good opportunity to study alternative credibility arrangements in the margins of auditing as well as their justifications.

Citation

Svärdsten, F. and Tamm Hallström, K. (2024), "Audit credibility and LGBTQI rights: certification operation in the margins", Accounting, Auditing & Accountability Journal, Vol. 37 No. 9, pp. 53-74. https://doi.org/10.1108/AAAJ-12-2022-6179

Introduction

Accounting scholars have long taken a keen interest in the growth and development of audit. Auditing can be understood as an assembly of techniques linked to ideas and rationales whose character may change over time (Andon et al., 2015; Humphrey et al., 2021). As Power (2011) notes, previously unexplored auditing practices “in the margin” provide an opportunity to understand how audit is developing and to question what have previously been taken as prerequisites for an audit. Against this backdrop, scholars have started to pay attention to the formation of different kinds of auditing practices in new audit spaces, including verification of baseball cards (Jamal and Sunder, 2011), online user reviews (Jeacle and Carter, 2011; Jeacle, 2017), sustainability assurance (Farooq and Villiers, 2017), auditing of advertising probity (Andon et al., 2022) and certification of eco-labels (Gustafsson and Tamm Hallström, 2018).

Some studies in this emerging body of research have shown that traditional and taken-for-granted features for trust and credibility in auditing, such as independence and objectivity, may not be as important as we might think when audit travels to new spaces (Gendron and Barrett, 2004; Mennicken, 2010; Andon et al., 2014). Instead, other features may stand out as salient, such as expertise and deep involvement in the audited issue at hand (Barrett and Gendron, 2006; Jamal and Sunder, 2011). When auditing enters new spaces, it might be exposed to new ideas, demands and expectations, which also means that the stakeholders that engage and succeed in articulating and promoting such ideas and expectations may change, with consequences for what is audited, how it is done and by whom (Andon et al., 2015).

In this paper, we employ the definition of audit space provided by Andon et al. (2015, p. 1400), which refers to “novel auditing and assurance services that have emerged at intersections between audit and other fields such as the environment, the public sector, sport and education.” Studies of such spaces are still relatively scarce, and calls have been made for more studies of new audit operations with specific attention to alternative claims for credibility (Andon et al., 2015, 2022). This paper responds to these calls with an in-depth study of the LGBTQI certification audit run by the Swedish Federation for Lesbian, Gay, Bisexual, Transgender, Queer and Intersex Rights (RFSL). The RFSL is a non-profit civil society organization that has performed LGBTQI certifications in Sweden since 2008. This certification has emerged at the intersection between certification audit and education, which makes it a relevant example of a new audit space.

We find the RFSL's LGBTQI certification interesting as it applies organizational arrangements that are far from what is associated with credibility in contemporary certification auditing. Such arrangements are typically based on a principle of separation – between the auditor, the auditee and the standards used for audits – which is assumed to secure independence, transparency, objectivity and thereby credibility for the auditing process and its outcome (Boiral and Gendron, 2011; Sandholtz, 2012; Brunsson et al., 2018). This way of organizing certification auditing has evolved during the past decades, with the European Union (EU) as a central actor in the establishment of a regulatory regime based on standards produced by private organizations such as the International Organization for Standardization (ISO), certification audits provided by certification corporations and accreditation as an audit of state authority (Higgins and Tamm Hallström, 2007; Loconto and Busch, 2010; Galland, 2017; Gustafsson, 2020).

Previous studies of certification audits in new audit spaces, such as those based on Fairtrade standards, Kontrollföreningen för alternativ odling (KRAV) standards for organic products and early work on ISO 9001 standards, have shown that even though these operations have diverged from commonly used organizational arrangements at the beginning of their operations, they tend to gravitate towards such arrangements over time as the audit operations matures and expands (Gustafsson and Tamm Hallström, 2018; Arnold, 2020; Arnberg et al., 2022). KRAV, for example, started as a labelling organization in 1985, both developing its own standards and performing certification based on these standards. As the distance between trading parties increased following expanding demand for organic products and intensified world trade, pressure to find methods to assure the credibility of trading parties also increased. KRAV contributed to the establishment of a new international accreditation program, run by the International Federation of Organic Agriculture Movements (IFOAM), targeting standard setters that were also performing certification within organic farming. Further on, however, KRAV decided to sell its certification operation to one of the multinational certification firms and to move to the arrangement of separate standard setting, certification and accreditation as an audit of state authority, which was launched by the EU through The New Approach in 1985 (Gustafsson and Tamm Hallström, 2018). The organization behind the Fairtrade label developed in a similar way, separating its certification operation from the standard setting (Arnold, 2020).

Such convergence tendencies have been justified as a means of increasing the overall credibility of certification audits that are considered necessary to facilitate global trade. Researchers have also highlighted that these organizational arrangements tend to lead to spirals of mistrust, with an escalating number of organizations operating to establish credibility in the auditing system (Hatanaka, 2014; Brunsson et al., 2018; Rasche and Seidl, 2019; Arnold, 2020) leading to unwanted effects such as diluted responsibility for the certification audits and their consequences (Brunsson et al., 2022).

The RFSL's LGBTQI certification, having expanded substantially since the organization's launch in 2008, is different in several ways. First, it offers training to all employees of the operations that they are going to audit, and the trainers act as auditors as well. Second, the trainers/auditors at the RFSL who conduct the certification have no accreditation to guarantee their credibility, and the LGBTQI certification is the only one of its kind. Third, the RFSL decides itself what standards should apply for the certification and adapts these standards to the operation being audited. Against this background, we consider this certification operation to be a case of a practice “in the margins” of certification auditing in the new audit space of LGBTQI rights.

Given that other certification operations have gravitated towards common organizational arrangements to assure their credibility as the audit operations have matured, it could be expected that the LGBTQI certification would follow a similar convergence pattern. Still, this seems not to have taken place in the case of the RFSL. Although it is impossible to foresee its future organization, we use this case to explore mechanisms that enable it to remain divergent in the margins of auditing by posing the following research question:

RQ1.

How and why is divergence from common organizational arrangements for audit credibility maintained in the margins of certification auditing?

We address this question with a study of how the Swedish LGBTQI certification was set up and has been reorganized during its first 15 years of continuous expansion. The aim of this paper is to contribute to the knowledge about the diversity of credibility arrangements in new audit spaces “in the margins” of auditing and the implications of such arrangements. We use insights from previous studies of escalating organizational arrangements in certification auditing (Boiral and Gendron, 2011; Kouakou et al., 2013; Gustavsson and Tamm Hallström, 2018; Brunsson et al., 2022) to create a contrasting image that we then use as a basis for our analysis of the Swedish LGBTQI certification during its first 15 years as a provider of certification audits. We specifically explore the stakes that motivate, and arguments used by, the RFSL to decide upon organizational arrangements that deviate from the dominant norms of credible certification audit.

Next, we present an overview of certification literature with a focus on three common arrangements for credible certification auditing, all based on the principle of separation: (1) the separation between the advisor and auditor role; (2) the separation between the certification and accreditation auditors; and (3) the separation between the actor setting standards and the one providing auditing. We then discuss recent developments within organization theory about escalating organizational arrangements surrounding certification auditing (Gustafsson and Tamm Hallström, 2018; Brunsson et al., 2022). We use this scholarly work as a framework to discuss the spirals of mistrust that tend to arise with organizational arrangements for credible audit and to understand the organizational decisions behind the RFSL's LGBTQI certification. After that we describe the methods we have used in this study, followed by an account of our findings, structured according to the three common arrangements for audit credibility. The paper ends with a concluding discussion of the implications of this study's findings for our knowledge about audit credibility in new audit spaces.

Literature review

It can be challenging for auditors to live up to demands for credible audits. One challenge for auditors is striking a balance between a reasonable certainty that the auditee fulfils the requirements of the audit while at the same time remaining at a distance to satisfy the ideal of the auditor's independence and objectivity. Distance and independence from the auditee as well as from political and commercial interests thus tend to be highly valued and prioritized (but difficult to achieve) and are commonly seen as crucial for audit credibility (Humphrey and Moizer, 1990; Sikka and Willmott, 1995; Jeppesen, 1998; Power, 2003; Barrett and Gendron, 2006; Kouakou et al., 2013; Tamm Hallström and Gustafsson, 2014; Andon et al., 2022).

However, studies of audit in new audit spaces have shown that in some contexts, such as sports settings, a deep involvement with the auditee is a prerequisite for a profound, meaningful and therefore also credible audit (Jamal and Sunder, 2011; Andon et al., 2014) and that the distance from the auditee, normally associated with audit credibility, is subordinate to having a close relationship with the auditee. Still, even in such contexts, there may be an inherent risk in an auditor getting too close. In Andon et al.'s (2014, p. 89) study of new assurance providers in the context of rugby and football for example, the authors show how the auditors had to “strike a delicate balance between ‘hard-nosed’ analysis/investigation and a more approachable style” towards the auditee. Even though one prerequisite for audit credibility in this context was the auditors' closeness to the auditees, they still had to go some way to remain distant to some extent.

This shows that some core elements of auditing, such as distance between auditor and auditee, may well be applied in new audit spaces even though such elements can appear in new forms and may be constructed in new ways (Andon et al., 2015). However, in certification auditing, the tendency to gravitate towards audit arrangements based on a principle of separation and distance has been strong (Brunsson et al., 2018; Gustafsson and Tamm-Hallström, 2018; Arnold, 2020), and many certification audits apply similar arrangements for audit credibility (Boiral and Gendron, 2011). Below, we elaborate on three organizational arrangements that are common in certification auditing.

Organizational arrangements for credible certification auditing

A common norm within certification auditing is that the role of the auditor should not be combined with an educational or advisory role, because this combination might compromise the independence of the audit: an organization that sells advisory services to an organization while serving as the auditor that evaluates the organization's results is not seen to conform to the requirements of objectivity and independence (Boiral and Gendron, 2011). In this context, certification firms that provide both audit and consulting services have taken organizational measures to keep these services apart, in accordance with an international standard specifying such requirements. Such measures may include preventing one person from acting as both consultant and auditor for the same organization (Kouakou et al., 2013; Tamm Hallström and Gustafsson, 2014). This makes a formal distinction between the audit, which requires distance from the subject of the audit, and the consulting services, which require close interaction with the client (Power, 1997).

Another arrangement for credible certification audit is accreditation. Accreditation is a “certification of the certifier” performed by an external party to assure the independence and objectivity of the certification audits (Hatanaka, 2014; Andon et al., 2015; Brunsson et al., 2018; Gustafsson, 2020; Arnberg et al., 2022). Accreditors are also monitored against an international standard, usually by meta-organizations of accreditors (Loconto and Busch, 2010; Brunsson et al., 2018).

Yet another action to demonstrate credibility in certification auditing is the establishment of standards that auditors can use for reference when the audit is carried out. One argument in favor of standards is that the audit becomes transparent and fair, because all audits follow the same standards and are also documented in relation to those standards (Boiral and Gendron, 2011; Silva-Castañeda, 2012; Thévenot, 2022). Another argument is that generic standards enable the certification audit to be seen as a purely technical and objective process that is independent of the auditor performing the audit (Boiral and Gendron, 2011; Jeacle and Carter, 2011). But, perhaps even more important, standards are established by organizations that specialize in setting standards, such as ISO, Fairtrade and the Forest Stewardship Council, which are separate from the organizations performing audits (Gustafsson and Tamm Hallström, 2018; Arnold, 2020; Gustafsson, 2020), allowing for distance between standards, auditors and auditees.

These organizational arrangements tend to give rise to a landscape in which an escalating number of organizations are taking on specialized roles and performing tasks such as training and consulting, standard setting, certification, accreditation and meta-accreditation. Together, such a collection of organizations contributes to establishing and maintaining credibility in certification auditing. This development is not unique for the field of certification audits but is salient to it and has been recognized in auditing literature (Boiral and Gendron, 2011; Kouakou et al., 2013; Andon et al., 2015) as well as in organization literature (Ahrne and Brunsson, 2011, 2019; Brunsson et al., 2018; Rasche and Seidl, 2019; Arnold, 2020).

What is highlighted, in particular in the literature using organization theory, is that in addition to the complexity generated by an escalating number of organizations doing different parts of the “credibility work,” these arrangements generate increasingly complex structures of decisions that connect all these organizations in various ways (Brunsson et al., 2018; Rasche and Seidl, 2019; Arnold, 2020). For an organization to decide to follow only a generic standard may be perceived as insufficient from a credibility standpoint, which motivates additional organizational measures, such as an external audit of compliance to that standard, to reinforce the credibility of the organization following the standard (Boiral and Gendron, 2011; Rasche and Seidl, 2019). But then comes the question of how we can trust the auditor – who is watching that watchdog (e.g. Shapiro, 1987; Gustafsson and Tamm Hallström, 2019) – spurring even more organizational measures that are connected to the previous ones, as described above.

One critique of this complex system of organizations and organizing between them is that it has resulted in a situation of “diluted responsibility” (Brunsson et al., 2022, p. 11) for organizational decisions. In a system of many interconnected decisions that serve as premises for each other, the attribution of responsibility becomes complicated or even impossible. Studies show that responsibility may be pushed between decision makers and thus impossible to attribute to a single one (Galland, 2017; Arnberg et al., 2022). A certifier being accused of having approved a certificate of an organization with bad business practices, for example, may claim that it is not responsible as it strictly followed the procedures stated in the standard for certifiers and it is accredited by an external accreditation auditor. That accreditation auditor, in turn, may claim that it also follows an international standard and is a member of a meta-organization for accreditors performing peer reviews of its members and thus is not responsible.

Another, related, critique of certification audits following generic standards and procedures is that the audits lose social and societal relevance as they tend to make the auditees accountable for having the “correct” abstract generic control system instead for their actual conduct and performance (Walgenbach, 2001; Boiral, 2007; Kouakou et al., 2013; Hatanaka, 2014; see also Power, 1997).

To conclude, the three organizational arrangements discussed in this section and summarized in Table 1 below tend to generate an organizational complexity leading to diluted responsibility (Brunsson et al., 2022) and dysfunctional accountability relationships (Silva-Castañeada, 2012; Thévenot, 2022). In comparison, the case of LGBTQI certification is different in the sense that the RFSL stands alone with its decisions for how its certification is organized and therefore also has the responsibility for these decisions. In this paper, we use the arrangements in Table 1 as a reference point for key features in the organization of credible certification operations.

Research context and method

The RFSL was founded in 1950 as a non-profit organization and currently consists of the central secretariat, 36 local branches and about 7,000 individual members. The RFSL's goal is for LGBTQI individuals to have the same rights, opportunities and obligations to live and work as everyone else in society. Its engagements to reach this goal include a crime victims' hotline, training activities, a newcomers' operation, a unit that works specifically with HIV and health and political advocacy and opinion forming. Since 2008, the RFSL has offered LGBTQI certification by which organizations can show that their employees have a basic level of training in LGBTQI issues to work systematically with an LGBTQI perspective. In 2023, more than 720 organizations in Sweden had been approved by the RFSL and received an LGBTQI certificate.

Our analysis is based on two sub-studies. The first is a study of the certification units within the RFSL – looking at decisions around its establishment in 2008, decisions made for a reorganization of the training and certification procedures in 2016–2017 in response to problems identified from the first years of fast expansion and organizational adjustments made after the pandemic. This substudy was based on 10 interviews (see Appendix 1), conducted jointly by the authors, with the founders, the management group, administrative staff and trainers/auditors, combined with analysis of material from the RFSL archive and website. We focused on debates about possible ways forward regarding how to divide and coordinate the work, how arguments and rationales motivating different paths were surfacing and where the RFSL decided to keep some of the original arrangements while changing others based on certain arguments. Most interviews were conducted in 2016–2017 as the RFSL reorganized its certification operation. One additional interview was done in 2023 with a trainer who had been interviewed in 2017, to follow up on how the organization and work with the LGBTQI certification had developed since 2017.

The second is a study of certified organizations, based on interviews with employees at nine certified worksites (4 preschools, 1 library, 1 youth center, 2 youth centers for job coaching, 1 youth center for domestic violence) to obtain an understanding of how this certification practice works and is perceived from a customer perspective (see Appendix 2). We focused specifically on questions about workings and perceived values of and challenges with, the organizational arrangements decided on by the RFSL, which diverge from conventional ways of organizing credible certification auditing identified in the literature overview.

The interviews generally lasted between 60 and 90 min and followed a semi-structured format. The interview questions were relatively open-ended to allow the interviewees to elaborate freely on issues that they saw as relevant for understanding the certification process. With the interviewees' permission, all interviews were digitally recorded. The interviews were transcribed verbatim.

The interviews were conducted in Swedish, and the quotations presented in the empirical section of this paper have been translated from Swedish to English by the authors. The text underwent linguistic scrutiny by a translator. Subsequently, we reexamined the quotations to ensure fidelity to the original statements made by the interviewees.

The initial analysis of the data took the form of ongoing, open-ended examination as the research unfolded and resulted in a relatively extensive text that accounted for key empirical findings. Our more theoretically informed analysis reported in the present paper then followed a largely abductive approach (Lukka and Modell, 2010; Timmermans and Tavory, 2012). This analysis started with the identification of a theoretical puzzle (Alvesson and Kärreman, 2007) rooted in the somewhat surprising observation that a certifying organization, situated in a context where credibility tends to be associated with common audit arrangements, was operating as a credible auditor even without applying any such arrangements. Following the procedure of abduction, we searched the empirical material for clues that might explain this observation while constantly comparing our emerging interpretations with audit and organization studies. Throughout this process we remained alert to alternative grounds for credibility that are not commonly found in auditing literature.

Audit credibility in the context of Swedish LGBTQI certification

In this section we account for our findings about the organization of credibility of the LGBTQI certification. The headings of these sections mirror the generic, fundamental aspects that an audit operation needs to consider and decide upon to demonstrate credibility: the organization of training and advice, the organization of audit integrity and the organization of standard setting. We account for decisions made by the RFSL about the organization of its certification practice and contrast them with decisions commonly made in accordance with the arrangements identified in the literature. This also means that we highlight the arguments used by the RFSL to motivate the decisions made about its organization, and we account for the reactions in terms of both praise and criticism expressed by various stakeholders towards the RFSL and how its certification has been organized and reorganized.

Advisor and auditor: combining the role of advisor and auditor instead of separating them

The RFSL started certification in response to requests for in-depth LGBTQI training from organizations that had been buying other training programs from the RFSL. The training unit then decided to offer LGBTQI certification as a pilot, combining its training offers with follow-up activities and a formal approval, which was an attempt to ensure that the training had a real effect in the customer organizations. The appropriateness of combining training and auditing within the same organization was debated within the RFSL; there were concerns about whether the RFSL would have the integrity to “fail” a customer after the customer had paid for the training and spent time educating themselves. It was discussed whether a solution would be to let an external actor perform the audit. The project group working with the certification initiative agreed that, in principle, it was a good idea to separate training and auditing into two different organizations but, in practice, they could not see any suitable actors with sufficient expertise to perform such audits. The discussion led to the final decision to continue letting the RFSL train as well as audit.

In the best of all worlds, it would be an external [auditor], but we ended up deciding in all modesty that no one had as much knowledge of these issues [as the RFSL], which at that time had 55 years of member-based experience and everything that comes with an activist organization. There's no one else who is qualified to judge whether this is good or not. (Interview RFSL employee B, May 2016)

It was concluded that the quality of the audit would be jeopardized if someone other than the RFSL performed it. The RFSL presents itself as the expert organization that also has the best potential to carry out a high-quality certification process, including a training program to prepare customers for the audit. This organizational arrangement differs from common arrangements for audit credibility in which training and auditing tend to be separated (Jeppesen, 1998; Kouakou et al., 2013; Tamm Hallström and Gustafsson, 2014).

Initially, the certification started with a number of training sessions to certify all employees in the organization. Following these, the RFSL administered a final test based on a number of checklists to verify that the employees had obtained sufficient LGBTQI knowledge. If the employees passed the test, a certificate was issued. Yet the RFSL described the certification process as essentially qualitative, meaning that the trainer throughout the process would estimate whether the participants had kept up during the course of the training and acquired the knowledge that had been conveyed. Thus, another reason for the difficulty of separating the roles of trainer and auditor was that the assessment process started when the training started. This was described by one respondent as follows:

[At times our] trainers have met with a lot of homophobia, transphobia, and offensive comments and negative treatment. [They] sense that there has not been any kind of progress. Then we note it, talk about it as soon as it happens [ …] So certification is not just sitting there and going over the final exams. (Interview RFSL employee A, April 2016)

The trainers/auditors were also careful to point out that they do not sell their label too easily, that is, the rainbow flag representing the entire membership of the RFSL. They demand a substantial commitment to LGBTQI issues in return.

There were also occasions when the RFSL denied granting a certificate in cases where the trainers/auditors noticed that the organizations did not live up to the RFSL's demands:

Interviewer: What makes you deny [granting a certificate]?

Respondent: Many things, there are clear signs. Well, not all resistance is clear, but sometimes there can be a lot of resistance in the group, and management can be a part of the resistance, or it could be that attendance [in training sessions] has varied a lot, or it could be that the working group does not do what it should or that the working group does not have the necessary authority to act, or that we see things in the follow-up where we think “ok, this person should not approach anyone at all.” Thus, it is an overall picture. (Interview RFSL employee I, April 2017)

Even though the RFSL trainers/auditors met resistance at times, they were greatly appreciated by the customer organizations most of the time. Several interviewees expressed admiration for the trainers'/auditors' extensive knowledge in LGBTQI issues.

In my opinion the RFSL is an organization that knows about this stuff, it really feels like that, they know what they are talking about. (Interview employee certified F, youth center for job coaching, January 2017)

One customer also explained that it was important for them to trust the certifiers' knowledge and competence as it was difficult for them as “outsiders” to evaluate the quality of the certification because of the lack of a generic standard set by a third party.

[It is important that the RFSL trainers] are competent because there is no [general knowledge]. It is not like KRAV, for example, where there is a standard […] Then you [as a customer] can say “this is fake, this is not for real,” but there is no standard for the LGBTQI certification. We have to be aware about the fact that the goal [of certification] is increased awareness, and that the RFSL sets the standard. (Interview employees certified G, library, January 2017)

Moreover, in addition to the RFSL trainers'/auditors' formal knowledge about LGBTQI, the customers also appreciated that the trainers/auditors themselves belonged to the LGBTQI community and had their own personal experiences of discrimination.

They're especially good […] a whole lot of them have personal experience with this issue […] They put their heart and soul into it. [With such a background] you understand in a completely different way what you yourself have experienced, anybody can see that […] if you've experienced it yourself, you know it in your bones. (Interview employee certified A, preschool, December 2016)

Thus, because LGBTQI knowledge was seen as experience-based, the fact that the RFSL both trains and audits the customers was not perceived as an issue. On the contrary, it was regarded by both the RFSL and its customers as a prerequisite for a high-quality certification process because the trainers/auditors “know it in their bones,” and therefore “put their heart and soul” into the certification process. This was another reason why it was difficult for the RFSL to outsource the auditing to an external party with potentially less experience in LGBTQI discrimination.

The customer organizations thus mainly described a positive view of the trainers also acting as auditors. Among the few criticisms that were raised by customers, none concerned the RFSL being a provider of both training and auditing; rather, they questioned the relevance of bringing LGBTQI issues into their organizations. Critique against the LGBTQI certification was also raised in media in a “hateful” tone by outsiders, from far-right movements for example.

The price that the RFSL charges for its certification, which according to critics can be linked to their monopoly position, is also a topic of debate. Often this criticism is framed within a discussion about public spending, as many of the customers for this certification are public sector organizations. The criticism surfaces through statements such as: “Is this knowledge really necessary and worth spending money on?” or “Why should a civil society organization with state support make public sector organizations pay for this?” (Interview RFSL employee G, April 2017).

Some of our interviewees at the RFSL explained the critique against costs by pointing out that the RFSL works with norms and values instead of providing a product that is tangible and thereby easier to charge money for. Some interviewees also explained that the RFSL, as an organization operating in the civil society, is not expected to charge money for their services:

[There is an expectation] that we should work for free and share our knowledge and that we should be nice and act in the best interest of the LGBTQI community. The fact that we charge money for our services is seen as bad. (Interview RFSL employee A, April 2016)

The RFSL confronts the critique by comparing LGBTQI certification with other certifications, courses and consulting services and thereby showing that the LGBTQI certification is relatively inexpensive. The RFSL also explains that quality costs money. If the customer wants high-quality certification, they need to pay for it.

Audit integrity: political engagement and closeness instead of distance through accreditation

At the time when the RFSL launched its certification there was no discussion about third-party accreditation or a change in the RFSL's political profile. On the contrary, the people from the RFSL, including its certification unit, radiated strong political engagement in LGBTQI issues and the trainers'/auditors' outspoken political agenda seemed to engender a certain respect in the customer organizations.

Customers moreover described the RFSL as a well-known organization, with its logotype appearing in various contexts and representatives appearing on TV talk shows and in other public settings, making it an organization that many people trust.

Interviewees at the RFSL also explained that it had strength in being a progressive organization constantly in touch with recent developments in the LGBTQI community – keeping abreast of news about concepts, parlance, statistics and so on – through involvement in social media and other fora. This progressiveness, however, needed to be balanced with the customers' ability to understand the LGBTQI community.

We may notice that there are emerging debates within the community, and that's really a strength with the RFSL. It can surface in social media, at a general assembly, or within a group of activists of some kind. We notice that an issue is being established, new concepts being used, simply new ways of expressing ourselves. The challenge is to notice such activities but still be clear and understandable in relation to the organization being trained and certified. (Interview RFSL employee E, March 2017)

The interviewees saw the LGBTQI community as something valuable for the credibility of the LGBTQI certification. First, the RFSL was seen as representing a significant part of the LGBTQI community, and this meant that the RFSL was likely to represent the customer organizations' own customers/clients. Second, because the RFSL is a member-based organization, the view was that it had to comply with the will and needs of the LGBTQI community. This meant, in turn, that the community was regarded as a control mechanism. An interviewee from one customer organization explained that he trusted the RFSL much more than other “specialists” such as management consultants.

I am somewhat skeptical towards all consulting firms that exist for all kinds of issues. I think it's better with organizations that are connected to a [social] movement and not only to specialists [consultants]. […] [I] don't trust specialists. Specialists have to be controlled, and I think that the [RFSL's] collective organization can be a sufficient control mechanism. In private consulting firms, that doesn't exist. (Interview employee certified H, youth center for job coaching, February 2017)

The fact that the RFSL had a clear political agenda and strived for change in society also made the customers trust the RFSL and its mission to do a good job – if the RFSL did not provide high-quality training/certification, it would not fulfil its mission. And in this case, high quality was associated with political engagement.

Another aspect that the customers associated with high quality was the close interaction with the RFSL in the certification process. One interviewee at a preschool made the following statement about the value of getting close to the trainer/auditor:

You could always bounce ideas off to the RFSL – “How should we word this? Does this sound good?” It wasn't that they gave you the “right” answer. They bounced right back a lot. But you still got that give and take that I think is really important. And that you could … you never felt stupid asking a question. Any question was, like, fine. (Interview employee certified B, preschool, December 2016)

One customer also explained that working with LGBTQI norms was a matter of practice and learning by doing. The work was not completed when the certification was over; that was just the beginning. To be able to work and act in an LGBTQI-friendly way, they needed to practice.

When acting against discrimination and prejudices, knowledge is of course very important. But it is relatively easy to gain knowledge. What is even more important is how we act and operate in relation to norms […] we have to practice, because that [norm-critical behavior/norm awareness] is something you have to practice. Knowledge is not enough. (Interview employee certified H, youth center for job coaching, February 2017)

The quote illustrates that customer organizations acknowledged that the LGBTQI certification was much more than just mediation of theoretical knowledge about LBGTQ issues; rather, it was something that needed to be practiced and reflected upon for some time, in dialogue with the RFSL experts. Because the LGBTQI work was an ongoing learning process that aimed to continue after the training and granting of a certificate, the customers wanted support from the RFSL even after the certification, and this continued “closeness” was appreciated. At the same time, the customers somehow took this support for granted because of the RFSL's political status and the assumed pressure from the RFSL members.

I hope the RFSL can be there in the background and assist us; another trainer would not do that. I think that [assisting] is in their [the RFSL's] interest […] and I don't mean financial interest; they have members that expect the RFSL to work with these issues […] that goes without saying somehow. (Interview employee certified C, youth center for job coaching, November 2016)

Still, not all customers agreed with the notion that employees in the customer organization entered a “close” relationship to the RFSL during the certification process. Some interviewees explained that they experienced a great distance because they were not familiar with LGBTQI issues and were afraid of doing or saying something wrong from an LGBTQI perspective. They also explained that the trainers/auditors were not open to problematizing the LGBTQI norm.

It [the training/certification process] was not neutral and they [the RFSL] did not say that it should be neutral, it was clear that they were acting in their own interest […] they were very competent [but] […] how can I put this […] in the way they talked about inclusiveness and broad-mindedness, they were quite narrow-minded […] many things I think are very good but […] I was not comfortable problematizing these issues because there was a risk of being called out somehow […] and being put in a box, so I didn't. (Interview employees certified J, youth center for job coaching, December 2016)

According to these interviewees, the LGBTQI norm was strong, politically correct and not possible to question without fear of losing the possibility to be certified. This suggests that even though the RFSL did get “close” to the customers in the certification process, they also represented a norm that was different from the knowledge and experience of the customer organizations, and in this sense, the LGBTQI norm created a distance between the RFSL and the customers. Thus, on the one hand, the RFSL trainers/auditors lacked accreditation (and never discussed it as an option), which could be seen as jeopardizing their independence given how closely they worked with the customers. On the other hand, they kept a distance from the customers through the knowledge base they were part of.

Level of standardization: customized certification process instead of a general certification process based on a generic standard

When the certification was set up in 2008, the RFSL had no intention of establishing any form of centrally determined generic system or standards for the audit to gain transparency, to depersonalize the audit or to create the distance from the auditee that supposedly assures the auditor's independence. When the certification operation started, the training consisted of LGBTQI knowledge that was sometimes research-based but also based on the trainers'/auditors' personal experience. However, as time went on, the customer organizations started to demand a certification process that was adapted to their particular operation and therefore the RFSL decided to hire people with specific competence in relevant areas, such as healthcare.

We noticed that we need to adjust our training to the operations we meet with. Offering general LGBTQI knowledge is not enough. We need to understand the operation in different ways […] It's not really about going there and talking about [the RFSL] or about LGBTQI people spreading the RFSL's agenda and information. It's much more about meeting the needs that each operation has. (Interview RFSL employee A, April 2016)

Because of the demand for tailor-made certification processes, the RFSL decided to organize the certification with two trainers/auditors working together on a certification project, one of them with specific knowledge of the organizational context. For example, if training in a healthcare organization, one trainer would not only have LGBTQI knowledge but would also have healthcare (e.g. nursing) experience. This meant that each certification engagement would be tailored to the specific needs of the customer.

It's our overall best selling point, the fact that our trainers have experiences from education, healthcare [etcetera]. They know the sectors we work with and, in various ways, they are capable of adjusting our competence and knowledge of LGBTQI issues to the daily work of the customers. (Interview RFSL employee E, March 2017)

One interviewee described how this way of working enabled the trainer to “check the temperature” during the training sessions and to be able to say “ok, this organization seems to perceive these questions as sensitive” and adapt the training accordingly. Another reason for avoiding generic standards had to do with the processual approach as a pedagogical principle, applied to increase the chances of the certification leading to the intended effects:

[Our work] is mainly about how people are treated, and this is an area where it's impossible to structure too much. The knowledge is really situated with such complexity, making it impossible to translate into a standard. One way of acting may be correct in one situation but completely inappropriate in another, making it impossible to specify a simple solution or treatment in a manual. (Interview RFSL employee I, September 2023)

The work with tailor-made rather than standardized certifications was also appreciated by the customer organizations:

In the training, they [the RFSL] brought in literature and a connection to the library, so we felt like the RFSL made an effort to take the perspective of a library […] It was a very good connection to literature. (Interview employees certified G, library, January 2017)

However, after a few years of operation, pressure for change started to grow internally because of a perceived need to become more transparent. The RFSL certification operation was not only growing in terms of numbers of certificates awarded but also in the breadth of sectors in which customer organizations were operating. At the same time, the responsibility that came with the LGBTQI certification started to show. The RFSL received emails from disappointed members who had been ill-treated by certified organizations. One interviewee who worked with the certification during its first years of operation described this in the following way:

When we put a stamp like this – “we have been here and checked” – what does it mean? It cannot mean the RFSL guarantees that everybody will get a good reception and treatment from an LGBTQI perspective […] although this is what we aim for. It's very difficult and we see examples of such difficulties. [Members emailing us] “I went to this LGBTQI certified healthcare center and they said quite astonishing things to me” […] So we have discussions about our target group. We don't want our members to encounter this type of situation […] because you may become very, very disappointed, mainly towards the RFSL as an organization […] We follow up on various Facebook groups where [these negative statements] may sometimes appear. (Interview RFSL employee D, March 2017)

This confirms the customers' assumption, as mentioned earlier, that the members had a controlling function. Thoughts about the members, described as “critical players” with a fundamental role for the RFSL's legitimacy, came up repeatedly in our interview material with the RFSL employees. Although critical messages through emails and social media come in only a few times a year, several interviewees referred to such emails from members. As expressed by one interviewee, “an angry and disappointed community and criticism that comes from there scares us more than a disgruntled customer” (Interview RFSL employee A, April 2016).

During the winter of 2015–2016, the RFSL launched an internal development process to create a more formalized structure for the fast-growing certification operation. There were four main problems motivating a change process. First, clients valued the tailor-made approach highly, but sometimes uncertainty was also expressed about the lack of a standard clarifying the certification requirements. Second, it was becoming difficult to handle the increasing number of both clients and trainers/auditors working with the clients. Third, occasionally members of the RFSL, being among the users of the services provided by LGBTQI-certified organizations, would contact the RFSL certification unit to report bad treatment experienced at a certified organization. Fourth, it was concluded that the work needed some kind of formalization to increase the efficiency of the certification operation without losing the RFSL's unique selling point: the tailor-made certifications. One employee working with the reorganization of the certification described the dilemma as follows:

To standardize and at the same time keep the tailor-made certification process was the big challenge. It was about creating a clear structure, clear boundaries, clear criteria, a standardized process so all trainers and customers know what it [the certification] is about. (Interview RFSL employee F, March 2017)

One major change was thus to standardize the structure of the work and to make that structure transparent to customers. According to the new structure, the RFSL would start each certification with an introductory meeting with the client to identify core characteristics of the customer organization and to adapt the training accordingly. The training workshops then started with general knowledge about LGBTQI rights and perspectives, going through fundamental concepts such as homo, bisexual, trans and queer and insights about living conditions and health situations, for example. After that, the workshops moved towards knowledge and exercises adapted to the type of organization being certified. One new exercise was about starting a change process in the organization that included a requirement to form an LGBTQI team. The team would be responsible for drafting an organizational action plan. The plan should include the formulation of a vision and purpose with the LGBTQI work based on discussions about the present situation in relation to the desired situation and strategies for how to reach set goals during the coming years within a chosen number of development areas.

The structure with workshops, formation of an LGBTQI team and development of an organizational action plan was in fact replacing the previous knowledge test that was in the form of a number of checklists. One RFSL employee made a comment about benefits of this change:

We tried to fit very much into them [the checklists], but we realized that many organizations work in ways that don't fit those checklists. If you don't have office space, for example, then the point regarding physical work environment doesn't apply. So, we have skipped the checklists and moved to a more flexible form. Now we encourage them [the customers] to ask themselves: “What do we need?” […] We think it's a better pedagogy. You make a decision about what to do, it becomes your own decision, not one that the RFSL has forced on you. But of course, we look at the action plan. (Interview RFSL employee I, September 2023)

Yet another part of the efforts to standardize the certification work regarded what happens after a certification. One year after the action plan has been approved, a follow-up meeting is held, and two years after that, a recertification would be needed, again adapted to the needs of the customer organization.

Although the structure of the certification was standardized, the content was still tailor-made. “Different operations can reach the learning objectives in different ways,” one interviewee concluded, making the following reflection about the changes:

Put bluntly, one could say that what we've done is to decrease the level of the criteria, that is, we've made them more realistic, as they were far too visionary and high level before, not in line with what we can provide to the customer organization. So, what we say now, what we think we can promise to provide as part of the certification, is that the organization will learn about the living conditions and health situation of LGBTQI people, and that the organization's employees will be coached in initiating a change process. (Interview RFSL employee F, March 2017)

The new message was that the RFSL could not guarantee that all individuals in the certified organizations would behave in accordance with LGBTQI norms, but it could guarantee that all employees had gone through training and that the customer organizations had installed the formal structure (action plan) to improve the conditions for actual change.

In addition to the decisions to standardize the work structure and the introduction of a few new requirements such as the LGBTQI team and the action plan, the RFSL decided upon a few more measures to make the training/auditing more efficient, such as going back to only one trainer responsible for each certification and decreasing the time devoted to each of the four workshops from 4 h to 2.5 h, while adding two digital sessions with prerecorded (standardized) training programs to be discussed at the following physical workshop. As illustrated in Figure 1, these changes implied that the total number of employed trainers/auditors at the RFSL certification unit actually decreased during the period 2015–2023, as did the coordination needed among them, although the certification operation in terms of approved and issued certificates continued to expand during the same period.

The changes were perceived to provide useful tools for trainers/auditors to deal with concerns about members' reactions when visiting a certified organization while also involving challenges associated with the well-known tension between structure and judgment in auditing:

There are also challenges linked to the new work structures with clear criteria […] You may have a certain feeling that the discussions in the customer organization are not reaching up to a sufficiently high level. Formally, they meet all criteria, but there is still something missing. Earlier, when we were more unclear about our criteria, I think it was easier to just make a decision based on the trainer's feeling – “No, this doesn't feel ok, the level of the discussions was not sufficiently high, they never landed,” whereas now we don't have this possibility. As a trainer, you don't have this mandate any longer and I think this is good. The personal judgment shouldn't be arbitrary, that a decision is based on the feeling of an individual trainer. (Interview RFSL employee F, March 2017)

To conclude, the gains of increased transparency through standardized and even slightly “blunt” criteria (workshops, LGBTQI team and action plan), towards customers, trainers/auditors and the LGBTQ community, thus seemed to be perceived as more realistic to work with. Moreover, these changes meant that the RFSL held on to its original way of organizing, thus continuing to diverge from the widely established way of organizing credible certification audits by separating the tasks of providing training, performing certification audits, writing standards and carrying out accreditation audits. We now turn to the discussion about audit credibility in the margins of certification auditing.

Concluding discussion: audit credibility through the lens of the margins

In this concluding discussion, we elaborate on why Swedish LGBTQI certification has not conformed to dominant norms for audit credibility as it could be expected to in view of previous studies of certification audit (Gustafsson and Tamm Hallström, 2018; Brunsson et al., 2018; Arnold, 2020). We highlight two features of particular relevance for our understanding of the case: the implications of responsibility for the effects of the certification and the role of beneficiaries in the organizational arrangement chosen.

Organization of audit credibility and responsibility

Comparing the RFSL certification to the Swedish eco-label KRAV (mentioned in the introduction), which was launched about 20 years earlier with a similar way of organizing, the normative contexts of these two periods clearly differ. In the mid-1980s, when the use of eco-labels and certifications had not yet blossomed, it was still quite open as to how to organize a credible audit operation within certification, although a trend towards separation and distancing between standard setting and auditing started to become discernible. In 2008, as the RFSL launched its certification, the way of organizing audit credibility based on an organizational separation of various roles in support of the certification was far more established and widely acknowledged. Nevertheless, the RFSL decided to perform all roles itself. By insisting on keeping training and certification within a single organization to make sure that the certification had the desired effects, the RFSL also decreased the risks of responsibility dilution associated with other forms of credibility arrangements (Brunsson et al., 2022). In this way, responsibility for LGBTQI certification was concentrated in the RFSL rather than diffused among several different actors. This, in turn, was deemed necessary for the RFSL to fulfil its political mission: to improve living conditions for LGBTQI individuals in Sweden.

This means that the RFSL's political mission comes with a certain kind of responsibility for the certifications' effects in the customer organizations. While other certification operations tend to become more concerned with abstract generic control systems rather than the actual conduct of the auditee (Walgenbach, 2001; Boiral, 2007; Kouakou et al., 2013), the present case is an example of a certification audit operation that has tried to take responsibility for the customer organizations' actual conduct, and this has had implications for how the LGBTQI certification has been organized.

For the RFSL, questions about a lack of credibility and independence did not arise as they were shown to do in previous studies of certification auditing (e.g. Gustafsson and Tamm Hallström, 2018). In line with previous studies of new audit spaces in the context of sport (Jamal and Sunder, 2011; Andon et al., 2014), expert knowledge about the issue at hand overrode concerns for audit independence. But in relation to previous studies of new audit spaces and the role of expertise in these spaces, the RFSL's experience-based expertise can be seen as unique. The fact that the RFSL trainers/auditors had their own experiences of discrimination and interactions with a heteronormative society gave them a special kind of personal legitimacy that would be difficult for someone outside the LGBTQI community to obtain. And it was perceived as impossible to standardize this particular kind of knowledge and to divide responsibility for it among different actors as is typically seen within certification audit (Loconto and Busch, 2010; Brunsson et al., 2022).

However, taking full responsibility for the certification and its effects came with challenges. Over time, it became impossible for the RFSL to ensure that all customer organizations complied with the RFSL's requirements. When the RFSL decided to standardize the certification process to some extent, it placed less emphasis on substantial requirements and outcomes and more on procedural ones, such as the establishment of an LGBTQI team and initiation of work on an action plan. This meant that the RFSL backed off from some of the responsibility for the consequences of the certification by recognizing that it was not realistic to expect full compliance with an LGBTQI perspective. However, even though the certification process was standardized to a certain extent, the RFSL never ceased to highlight the importance of being “close” to the auditee in the certification process, which included a mandatory training program for all employees of the certified organization. This part of the certification, emphasizing the importance of engaging with every individual employee of the certified organization and to do this during a training program based on several interactions spread out in time, indeed differs from the more common management system approach to certification. In such system-based certification audits the certification auditor, once a year, interacts with the certified organization's abstract management control system, used as a proxy for the organization's compliance with a certain standard, combined with interaction with a representative of the certified organization (e.g. the quality assurance manager) and perhaps a few more managers to hear how they work with the management control system. The motive behind the RFSL's choice to hold on to the resource-intensive arrangement of interacting with all employees and not reduce the certification to interaction with an abstract system and a few representatives of the organization to be certified was to take responsibility for the certification and its effects.

The RFSL's position as a mission-oriented actor with a clear political agenda moreover means that the organization is vulnerable in a different way compared to other audit operators. The RFSL's political mission seems to create expectations of a certain amount of idealism. Unlike Jamal and Sunder (2011), who study the rather “non-political” context of baseball cards and show that the auditors' expertise allows them to charge higher prices for their services, the example of the RFSL shows that mission-driven organizations face other types of expectations in the audit context – to be driven by an ideal and not by financial profit. The exposure to hate also makes it obvious that the LGBTQI certification is vulnerable in a very different way compared to audits in other new audit spaces such as sports (Andon et al., 2014), advertising probity (Andon et al., 2022) or online user reviews (Jeacle and Carter, 2011; Jeacle, 2017), which are seen as less controversial. It also differs from other “rational” and “objective” traditional certification audits (Boiral, 2007; Kouakou et al., 2013) that can always refer to generic standards and procedures that have been decided by someone else (Brunsson et al., 2022). Furthermore, the political engagement also comes with certain expectations, such as provision of constant feedback and assistance to the customers even after the certification is finished, which is not common in other certification audits.

Organization of audit credibility and the role of beneficiaries

Research about certification audits that are typically organized in complex systems of interdependent organizations to secure objectivity and independence reveals that beneficiaries (e.g. end consumers) are rarely involved in decision making about how certification auditing is organized and performed. Within standard setting, for example, where beneficiaries may have a legitimate seat at the table, their representation is usually low (Cochoy, 2005; Hauert, 2010; Tamm Hallström and Boström, 2010).

The predominant practice among actors working with certification audit and accreditation is to have a representational relationship with beneficiaries (Koenig-Archibugi and Macdonald, 2017). In such relationships, beneficiaries are not involved in making decisions about the certification audit as individuals. At best, as within standard setting, they are represented by a meta-organization for consumer organizations, or yet other organizations claiming to represent consumers' (or other beneficiaries') interests. Such representation is often restricted to the rhetorical level – to the interests of “made-up beneficiaries” – and based on one-way communication (e.g. Young, 2006; Koenig-Archibugi and Macdonald, 2017; Stenka and Jaworska, 2019). Certifiers working with ISO 9001 certification, or accreditors conducting accreditation of such certifiers, for example, highlight “consumers,” “customers” and “the general public” as crucial intended beneficiaries of their auditing services but they are mainly rhetorical, abstract, made-up beneficiaries ascribed imagined needs (Arnberg et al., 2022). When auditees make themselves auditable in relation to certification standards and procedures, they are also complying with the notion of the made-up beneficiaries and agreeing with the knowledge base put forward by the certification audit system (compare Power, 1997). In Figure 2 below, we illustrate the shared knowledge base between the auditor and the auditees with a representational relationship with beneficiaries.

In this context, the perceived needs of made-up beneficiaries can be far from the needs of actual beneficiaries (Walgenbach, 2001). Actual beneficiaries tend to become “smallholders” (Thévenot, 2022, p. 224) whose voices are seldom heard, and if someone wants to make a formal complaint of any kind, these complaints need to be formulated in ways that comply with the knowledge base of the certification audit and its associated procedures. This, in turn, makes it difficult for individual beneficiaries to make their voices heard (Bain and Hatanaka, 2010; Silva-Castañeada, 2012; Arnberg et al., 2022).

What we see in the case of the LGBTQI certification, however, is something else. Rather than a representational relationship with beneficiaries, the RFSL represents an example of a certification auditor with an identity relationship with the intended beneficiaries (Koenig-Archibugi and Macdonald, 2017). This means that, in addition to a rhetorical relationship between the RFSL as a certifier claiming to represent the interests of its beneficiaries as an abstract category, there is a more concrete relationship in the form of physical interactions between the RFSL certification auditors, themselves being beneficiaries of the certification and other beneficiaries that are part of the larger membership of the RFSL and the LGBTQI community.

Contrary to certification audits that have only a representational relationship with the beneficiaries, the beneficiaries of the LGBTQI certification at least to some extent share the same knowledge base as the certifying auditors, as they share the same experience-based knowledge of interactions with a heteronormative society and perhaps also of discrimination. This is knowledge that the auditees (at least in the present study) did not have. This means that we see a dynamic between the auditor, auditee and beneficiaries in the LGBTQI certification that is different from conventional certification audits with representational relationships to the beneficiaries (see Figure 3 below).

The beneficiaries have the possibility to assess the outcome of the certification every time they visit a certified organization, and they have the competence to do so as members of the LGBTQI community. In case of any complaints, and in contrast to conventional certification audit arrangements with diluted responsibility relationships (Brunsson et al., 2022), the beneficiaries of the LGBTQI certification can also relatively easily communicate their claims and make the RFSL responsible for the outcome of the certification as they are part of the same community and share similar experience-based knowledge.

The RFSL's identity relationship with beneficiaries has implications for how it conducts the certification work. An imagined fellow member seems to rest on the shoulders of the LGBTQI trainers/auditors as they work with the certification, reminding them about the purpose and desired effects of the certification as well as the risks involved in simplifying and decontextualizing LGBTQI knowledge. Thus, differently from other certification audit contexts, the identity relationship with the beneficiaries motivates the RFSL to go to considerable lengths to gain an in-depth understanding of the customers' operations and to adapt the certification process to the specific characteristics of the customers to ensure that the certification maximizes the intended effects for the beneficiaries. Our study also shows that the identity relationship with the beneficiaries has implications for the LGBTQI certification's credibility.

The identity relationship may, moreover, explain why the RFSL is careful to point out the limits of the certification. It would be impossible for the RFSL trainers/auditors to guarantee full compliance with an LGBTQI perspective in the certified organizations as they cannot keep a constant watch on these organizations to ensure this compliance. The imagined fellow member on the shoulders of the RFSL trainers/auditors seems to help them refrain from making unrealistic claims that the certification is a guarantee that the auditees have adopted a full LGBTQI perspective. Such a message would build on the common assumption underpinning certification auditing that an organization with a rational structure, clear goals and processes documented according to a recognized management system standard and checked by an external, trustworthy party would act in accordance with that structure. There is much research showing that this is not a very credible assumption.

Conclusion

With this study and analysis of a certification auditor in the margins of certification auditing, including its attempts to remain divergent from conventional organizational audit arrangements, we add new knowledge to the literature about the diversity of credibility arrangements in new audit spaces. An auditor advocating a political agenda and seeking to advance social rights for a minority group may find it difficult to embrace conventional audit arrangements, as these run the risk of distancing the auditor from both the auditee and the political issue at hand. We have shown that this does not necessarily compromise the credibility of the audit. On the contrary, dedication to the political mission carries greater weight than commitment to traditional audit arrangements in this context.

Building on organization theory studies about escalating organizing among organizations in relation to certification auditing, including theorizing about causes of diluted responsibility, we have specifically highlighted the role of responsibility for the effects of the audits, as well as the auditor's relationship with beneficiaries in terms of representational versus identity relationships. By doing this, our analysis also contributes with a contrasting perspective on conventional audit arrangements, which tend to become taken for granted and difficult to question. This contrast helps us expose the fragility of such arrangements, which may threaten their taken for grantedness, as the gap increases between the constructed needs of made-up, mute beneficiaries, on the one hand, and the beneficiaries as real persons with voices to raise, on the other. And in the context of social rights, these voices deserve to be heard.

As far as we are aware, this is the first study conducted in the new audit space of LGBTQI rights. We encourage further research on certification auditing in the area of social rights for minority groups to broaden our understanding of the relationship between certification and social rights. For such endeavor, questions about responsibility as well as the utility and influence opportunities for the beneficiaries of the certifications should be of high priority.

Figures

Figure 1

Numbers of certification staff and approved certificates

Figure 2

Representational relationships where the auditor and auditee share knowledge base

Figure 3

Identity relationships where auditor and beneficiaries share knowledge base

Table 1

Organizational arrangements for credible certification auditing and their intended functions

Organizational arrangement	Intended function (rationale)
Separation between the advisor and auditor role	The auditors avoid assessing their own advice
Separation between auditor and accreditor to “audit the auditor”	Audit integrity: loyalty to the accreditation certificate stronger than loyalty to the auditee
Separation between auditor and standard setter	Depersonalized audit processes; the auditors avoid assessing their own standards

Source(s): Table created by authors

Table A1

Interviews with RFSL employees

Employee	Length minutes	Month and year
A	55	April 2016
B	76	May 2016
C	77	September 2016
D	79	March 2017
E	57	March 2017
F	57	March 2017
G	58	April 2017
H	41	April 2017
I	57	April 2017
	51	September 2023

Note(s): All interviewees except one had academic training, several in political science, often combined with e.g. gender studies, pedagogy, leadership, project management or organizational development. Several interviewees had worked for the RFSL for a number of years

Table A2

Interviews with employees in certified organizations

Employee(s)	Length minutes	Type of organization	Month and year
A	47	Preschool	December 2016
B	55	Preschool	November 2016
C	55	Youth center for job coaching	November 2016
D	41	Youth center for job coaching	November 2016
E	58	Preschool	December 2016
F	50	Youth center for job coaching	December 2016
G*	49	Library	January 2017
	57		January 2017
H	59	Youth center for job coaching	February 2017
I	62	Preschool	April 2016
J**	67	Youth center and youth center for domestic violence	December 2016

Note(s): *Group interview with two employees

** Group interview with three employees from two certified organizations