© 2019, Emerald Publishing Limited. Purpose: This paper aims to investigate two issues. First, the authors test the effect of the Sarbanes–Oxley Act (SOX) on audit quality after 10 years. Second, the authors test whether it was necessary to close all of the Arthur Andersen offices due to the misbehavior of a few (e.g. the Houston and Atlanta offices). Design/methodology/approach: The authors have used conservatism (Basu) as a proxy for audit quality. Findings: The authors find that, over the long run (10 years) after SOX adoption, there is a significant positive change in conservatism as compared to during the previous similar period. In addition, the authors find that only 6 of the 20 city-level offices of Arthur Andersen were less conservative than were their other Big 6 competitors in the same city. Furthermore, the results also suggest that some city-level offices of Arthur Andersen were engaged in more conservative accounting practices than were their competitors and the Houston Andersen offices. Originality/value: This study documents, using empirical evidence, that the implementation of SOX is successful, and that one factor that helped lead to this success might be the harsh punishment on Arthur Andersen.