Date of Award

Fall 2002

Document Type


Degree Name

Doctor of Business Administration (DBA)


School of Accountancy

First Advisor

Michael S. Luehlfing


There were two main objectives of this study. The first objective was to develop a theoretical model that will explain auditor resignations. To meet this objective, a model was developed which was primarily grounded in prior research related to auditor resignations as well as prior research addressing auditor switching. The second objective was to estimate the model at four different quarterly dates immediately prior to an event date (i.e., the resignation date) so that the effects of time on the explanatory power of the model and each independent variable could be ascertained. This objective was met by identifying and collecting data at each quarterly date for each variable in the model and by using logit regression to analyze the data (i.e., to estimate the model at each quarterly date prior to the event date). The sample of companies included both a resignation group of companies and a non-resignation group of companies.

Overall results from estimating the model suggest that the model was significant for each of the four quarterly time periods and that using information closer to the event date increased the model's explanatory power. In regards to the specific independent variables, the results suggest that, for all quarterly time periods, firms are more likely to resign from (1) clients who are smaller in size; (2) clients who are in financial distress as modeled by a net loss; (3) engagements in which the tenure of the auditor is low; and (4) clients who are in industries in which the auditor has an increasing and/or a relatively high market share. Additionally, companies that had a greater likelihood of management misrepresentation were also associated with an increase in the likelihood of a resignation; however, this association was sensitive to time. Specifically, this association was only significant in the two quarterly time periods immediately prior to the event date. Overall, these results suggest that CPA firms do respond to risk factors associated with the client and (such firms) are not solely fixated on monetary rewards.

Included in

Accounting Commons