Discussant's response to "An assertion-based approach to auditing";

              Discussant's Response to
"An Assertion-Based Approach to Auditing"
William L. Felix, Jr.
The University of Arizona
Critiquing an assertion-based approach to auditing is a bit like criticizing motherhood and apple pie, given its reliance on Mautz and Sharaf and an existing SAS, but my role here today is not to toss bouquets. My discussion of this paper will begin with a number of general issues where I think the paper misses its target or I have doubts about its content. I will conclude my discussion with a few points of lesser significance. These comments are intended to stimulate discussion.
Some Basic Issues
The authors of this paper present a wide-ranging analysis of their views on an assertion-based approach with some interesting insights into Clarkson Gordon's use of this method in their development of microcomputer tech-nology for audit practice. While very interesting and appealing ideas are presented, there are some major omissions that are critical to a careful evaluation of the ideas in this paper.
Beyond the author's assertions, there is no convincing argument in this paper as to why an academic or practitioner ought to view an assertion-based approach as either more effective or more efficient than some particular alternative or as a dominant strategy with respect to all available alternatives. An example of this lack of convincing argument is included in the last paragraph on the first page where it appears the authors suggest that since an assertion-based approach to auditing is in accordance with generally accepted auditing standards, that it should be used by practitioners and authors of auditing books. This is clearly an inadequate criteria. Our choices of both general audit philosophy as well as specific audit policies should be based on perceptions of improved efficiency and/or effectiveness. Ideally such perceptions would be based on some analysis or empirical data. Such evidence or other supporting analysis seems to have been omitted from the paper. If the authors or Clarkson Gordon have such data or analysis, it would be very worthwhile to present that information.
An assertion-based approach could be viewed as a planning framework to organize thinking about or planning for the types of errors that: 1) are possible or likely; 2) for which internal accounting controls may be considered; and 3) for which effective substantive tests (analytical review or substantive tests of details) need to be considered. Although the authors do not address the differences specifically, a useful focus for our following discussions might have been to identify key differences between the risk-oriented error-discovery
65