After several audit failures, questions and concerns have risen about auditor independence. Independence is the cornerstone of the audit profession and a basic ingredient of users’ confidence in financial reports (Mautz & Sharaf, 1961). Auditor independence is needed for the long-term success of the audit profession, to sustain the profession’s credibility. Despite a number of different proposals and changes in auditor oversight and self-regulation, the recent financial scandals have reopened questions about auditor independence.

Auditors are supposed to be independent of the audit client and to act in the public interest. However, it is very difficult to guarantee absolute auditor independence as the auditor is normally hired and paid by management. The shareholders right to appoint the corporation’s auditors is of limited practical value as management effectively controls the appointment process.1 This question is important in the context of audit conflict between auditor and client-management, as a potential threat to independence may occur if management is able to increase the pressure on auditor reporting behavior. Threat of dismissal may reduce the auditor’s incentive to maintain an independent attitude, especially as empirical evidence shows a high probability of auditor change when disagreements arise between auditor and client, or when qualified audit reports are issued (Craswell, 1988; Gómez Aguilar & Ruiz Barbadillo, 2003; Lennox, 2000). If auditors are more likely to face dismissal after qualifying a report and if the economic losses associated with dismissal are potentially higher for incumbent auditors, then auditors may opportunistically reduce their independence to retain the engagement. This reasoning suggests that auditor reporting behavior may be influenced by their perception of how client-management will react to audit opinion, and therefore impairment of independence may result in inadequate audits.

The principal concern for regulators is whether auditors succumb to undue management pressures, and what factors may be associated with impairment of independence. Auditor independence has been the subject of decades of intense empirical research focused on different factors that might be threats to independence, such as provision of management advisory services, size of audit fees, financial dependence, fear of client loss and length of the auditor–client relationship.

This paper focuses on the relationship between auditor tenure and auditor independence, and more precisely on the impact of auditor tenure on the auditor’s reporting decision. Regulators and the public have long been concerned about the effect on auditor independence of the length of the auditor–client relationship, but in the literature, there is no consensus about that effect. An initial suggestion is that long tenure may create an incentive for auditors to compromise their independence because close identification of the auditors with the interests of their client-management reduces their necessary professional skepticism and capacity to analyze impartially the accounting information released (Petty & Cuganesan, 1996). In this sense, mandatory auditor rotation has been suggested as a measure to improve auditor independence, as auditor rotation may reduce the opportunity costs of losing a client and thereby reduce the auditor’s incentive to impair their independence.

A second suggestion is that the shorter the association with the same client, the less the capacity of the auditor to recoup its initial investment in the company, thus increasing the auditor’s interest in avoiding premature termination of the contract. From this point of view, auditors can be more easily influenced in the earlier years of the auditor–client relationship, as the stream of quasi-rents is needed to offset the initial start-up costs of the engagement (Dye, 1991 and Elitzur and Falk, 1996). After the first years, the incentives for auditors to report material errors will be seen to increase (Geiger & Raghunandan, 2002). In line with these analyses, some have argued that long auditor–client relationships do not impair independence and that mandatory auditor rotation is not necessary.

The preceding discussion suggests that the two conflicting visions predict that tenure will have opposing effects on auditor independence. Given that the evidence concerning the effects of long-term audit engagement is not conclusive, the aim of this paper, using a sample of Spanish companies, is to examine the influence of audit tenure on auditor independence. With respect to auditor independence, we have focused our study on the phenomenon known as opinion shopping, because it provides an explicit acknowledgement of material omissions or misstatements in financial statements. Therefore, it is not surprising that opinion shopping raises concerns about the quality of services rendered by auditors.

Opinion shopping is a phenomenon that has been subject to an intense debate within regulatory bodies and between accounting academics and professionals, and this debate is fully justified by the effects that opinion shopping has on the real value of auditing services. Opinion shopping takes place when a company obtains a more favorable audit opinion than the quality of its financial information justifies. As noted by previous studies, empirical observations of this conduct are highly complex. Consequently, instead of concentrating on opinion shopping, most empirical studies have examined a phenomenon that is indisputably closely related: auditor change undertaken in order to obtain a more favorable opinion from the new auditor. This type of study analyzes a chain of empirically observable events, such as the sequence of audit opinions before and after a change. They do not, however, analyze other possible situations, such as the possibility that the company might obtain a more favorable audit opinion than it deserves without changing auditor. We are specifically referring to the possibility that the company may pressure its auditor with the threat of dismissal, thus inducing him/her to issue a more favorable report.

In order to capture opinion shopping, we have developed an audit opinion model based only on variables that reflect the company’s financial situation. Using this model, we have been able to calculate the probability of a company’s receiving a qualified audit opinion according to its financial situation. Comparing the opinion that companies deserve according to this model with the opinion actually received has allowed us to identify those cases in which companies have received favorably biased audit opinions. We consider these companies to be opinion shoppers. Next, we analyze whether the length of the audit engagement affects the probability of opinion shopping. This has a double aim: first, the relationship between auditor tenure and opinion shopping has not been studied previously, and second, Spain offers a highly singular context for studying this kind of relationship because of the existence of great differences between this country and the well-studied audit markets of English speaking countries.