Network Connectedness and the Convergence of Audit Styles

Viewing audit teams as social groups of auditors, we utilize network analysis algorithm "n-clans" to identify clustering of individual auditors, that is, audit teams, based on individual collaboration history. Drawing upon social learning theory, we argue that team network connectedness, specifically network density and closeness, shapes the similarity in the audit styles among team members. We first validate the team construct by showing the existence of team styles in audit outcomes and audit pricing. Using the dispersion of audit outcomes and abnormal audit fees as inverse measures of au-dit-style similarity, we find that more closely connected teams exhibit lower within-team variation in these measures. Further, the role of network connectedness for the within-team convergence of audit styles is more pronounced when firm-level quality control is less robust, the teams include junior auditors, and the clients served by the teams are more divergent. Network density and closeness also strengthen the relation between audit-team style and that of team members. We extend prior literature by identifying the audit team as a relevant unit of the analysis for audit production and documenting the importance of network connectedness to more consistent audit outcomes.