| Auditor's client acceptance and continuance decisions (CACD) are considered by professionals and academics as the most important decisions in audit practice (Gendron, 2001). The CACD have significant economic implications for audit partners, audit firms, and public companies (Johnstone and Bedard, 2004). Moreover, CACD's importance has risen sharply in the current audit market due to expanding and stricter regulation, and rising litigation risk (CCRC, 2005; GAO, 2008).;Despite the importance of the CACD for the audit services industry and the critical issues they raise, no literature review exists on the topic. The first paper of this dissertation attempts to fill this gap. The paper initially develops a conceptual framework to organize and review prior research on the CACD. Subsequently, the paper presents the main findings and limitations of existing studies thereby identifying valuable directions for future research. Finally, the paper sets the foundations for a general theory of the CACD. In the second paper, two major limitations in the extant literature on the auditor's client acceptance decision (CAD) are addressed. First, we respond to the lack of explicit theoretical frameworks in prior studies, and to Penno's (2005) and Wilks and Zimbelman's (2004) recommendations on promoting analytical modeling, especially game theory, in auditing research. Second, we investigate the roles of both the new client and regulation in the CAD process. Despite the importance of the new client and regulation in the CAD setting (Gendron, 2001; Johnstone and Bedard, 2004), these two factors have not been examined in previous CAD studies.;The second paper develops a signaling game theoretical model with two active players: a new client and the auditor. Each player can reduce the prevailing information asymmetry in the CAD setting using separate mechanisms. Costly signaling is the mechanism used by the prospective client. This mechanism is derived from signaling games and consists in the client providing voluntarily costly information about his business and riskiness to the auditor. The second mechanism, costly testing, is used by the auditor. This mechanism is common in the CAD literature and refers to the auditor conducting a costly investigation of the client's riskiness.;Equilibrium analyses show that the client's and the auditor's actions depend mainly on three factors: 1) the auditor's a priori opinion about the client's riskiness, 2) the auditor's testing technology attributes (cost and reliability), and 3) the client's signaling costs. In addition, the findings of the equilibrium analyses refine our understanding of the two players' behaviours in the CAD context. For instance, the model shows that under some circumstances, a low-risk or good client provides (no) private information to the auditor, even though the auditor believes he is good (bad).;Furthermore, the results show that in the CAD setting, signaling and testing mechanisms are complementary rather than substitutes. Mixing the two mechanisms may increase their usefulness while decreasing the testing error risk. Additionally, the two mechanisms together reduce the prevailing information asymmetry and prevent the auditor from making erroneous CADs more effectively.;Finally, comparative statics bring forth the critical role of regulation in the CAD context. The findings suggest that the stricter the legal liability regime, the more sceptical is the auditor. Accordingly, the auditor is less (more) inclined to test an a priori bad (good) client and more (less) inclined to reject (accept) him right away. In the same vein, a Big 4 auditor is more sceptical than a non-Big 4 auditor, while a specialist auditor is less sceptical than a non-specialist auditor. |
