Independence tends to be thought of as necessary for audit quality (Humphrey & Moizer 1990;Sikka & Willmott 1995; Jeppesen 1998; Power 2003, 2011; Kouakou et al. 2013; TammHallström & Gustafsson 2014). Auditors are experts with integrity – through their capacity tokeep a distance to the practice being audited and other commercial or political influences auditorsare expected to remain independent and thereby valuable. A central belief in auditing as well asother certification services is that users of these services should find value in receiving assurancefrom someone who has the incentive to provide good quality assurance, and this quality is partlydependent on independence (Jamal & Sunder, 2011).

Auditing can be understood as an assembly of techniques linked to ideas and rationales whosecharacter may change over time. For more than a century, the audit profession has played acrucial role in defining and promoting the independence of financial audits. Through standardsdeveloped by an independent, expert-based organization the politics involved in rule setting iskept at a distance from the auditing work that supposedly is turned into a transparent, neutral and“technical” matter of comparing rule against practice. Also, auditors are disciplined throughprofessional codes of conduct to support their independence towards the auditee. Withinauditing situated within a broader field of assurance, inspection, certification, accreditation andother oversight bodies, however, the role and organization of independence may differ bothacross time and space (Gendron et al., 2001; Power, 2011). The purpose of this paper is tocontribute to the emerging debate within critically oriented accounting research about therelationship between independence and audit and more specifically about ways of constructingindependence within diverse auditing contexts. This knowledge is specifically requested by bothaccounting and certification scholars, viewing independence as a social construct and thusdependent on context (cf. Jamal & Sunder, 2011; Power, 2011; Kouakou et al. 2013).

In order to advance the knowledge about the construction of independence within auditing, wehave studied an auditing practice in the “margins” (Miller, 1998), that is, within a new area ofauditing where independence norms and practices are not given or institutionalised. Morespecifically, we have studied the LGBTQ certification developed in 2008 and run by the SwedishFederation for Lesbian, Gay, Bisexual, Transgender and Queer Rights (RFSL). The RFSLcertification differs from many other audit operations in several ways: they decide themselveswhat standards should apply for the certification and adapt these standards to each operationbeing audited; they offer training to the operations that they are going to audit; and the auditorswho conduct the actual certification have no professional affiliation (like financial auditors) oraccreditation (like certification auditors) to guarantee their independence in the context of anaudit. From an independence perspective, this might seem like an “impossible” organization, andso to explain this we investigate how the RFSL works with and justifies the organization of itscertification, as well as how stakeholders around the RFSL perceive this certification.2 In thisway, we are able to analyse and explain what the RFSL bases its credibility and its potentialindependence on.