“…Specifically, SHOP(normal) is negative and significant for clients of all incumbent Big 4 auditors across all three competition proxies when competition is high, and is not significant 13 We use this method because interpreting the results is more intuitive than interpreting an interaction term, and the estimation of separate models for two groups is a common method used in cross-sectional tests [e.g., Jayaraman and Milbourn (2015), Chen, Gul, Veeraghavan, and Zolotoy (2015), Kirk, Reppenhagen, and Tucker (2014), Beck and Mauldin (2014)]. when competition is low.…”