Conclusion
Our review of the organizational memory literature and its application to the
ACUs of the largest five public accounting firms in Canada shows that the five
firms exhibit characteristics associated with differences in organizational memAccounting
Consultation Units: An Organizational Memory Analysis 687
ory. We classified two firms (3 and 4) as discovering organizations, two firms
(1 and 2) as conditioned viewing organizations, and one firm (5) exhibits characteristics
that are a mixture of both types. We argue that these different organizational
types, combined with how the ACUs are integrated into the broader
firm structure and organizational culture, result in differences in the ability of
the ACUs to act as a source of organizational memory.
There are, of course, limitations to our research. First, we rely on interview
data for our research. Although we corroborated tbis data to tbe extent possible
(e.g., interviews with various levels of personnel in the firm, walk-throughs of
actual consultations, examinations of archival data, physical inspection of
resources), there is the possibility that actual practice may vary from the interview
descriptions. Second, we employ organizational memory theory that was
developed as a firm level theory to examine one department of the audit firm.
We could, however, find no offsetting differences in the firms that would ameliorate
the observed differences in the ACU's capacity to serve as a source of
organizational memory on technical accounting issues.
Regulatory concern is growing about the amount and results of research
auditors perform on financial accounting issues (OSC 1996; Schuetze 1994).
The Ontario Securities Commission recently commented "It appears to [OSC]
staff that research is often not being done at an early enougb stage, is not well
documented and is not sufficiently thorough." It goes on to state "[OSC] Staff
will increasingly expect that individuals such as those in the national offices of
firms will have been consulted." (OSC 1996, 1-348) Clearly, the types of differences
in firms organizational memories that we have documented in this
paper are reflected by these comments. One approach for future research is to
relate our documented organizational memory differences with the OSC staff's
analysis of accounting research deficiencies. In one sense, this approach would
be an external measure of the "quality" of the research being done, and it may
suggest whether the organizational memory differences lead to differences in
quality of accounting consultations. Such researcb would require access to sensitive
OSC data — this may be difficult to obtain.
There is another related pair of questions about the organizational memory
effects of ACUs. First, do these differences in the capacity for organizational
memory result in different responses to consultations with the practice
offices for the same or similar issues when initiated by different individuals?
Second, do tbese differences result in the audit firm appearing to agree with different
accounting policies in seemingly similar situations? Both of these questions
may well be part of the OSC concerns.
Our analysis provides tentative answers to these questions. It suggests that
there is a greater potential for inconsistent guidance to be made by conditioned
viewing ACUs. This difference is due to their primary reliance on the personal
memories of the consultants and informal procedures for coordinating consultations.
Although the smaller size of the conditioned viewing ACUs may offset
the concern about inter-consultant coordination, it still does not resolve the
688 Contemporary Accounting Research
concern with consultant consistency. Prior research on memory reliance by
auditors extensively documents incorrect inferences being drawn from memory
(e.g., Moeckel and Plumlee 1989). Certainly, the anecdotal evidence in the
United States provided by Schuetze (1994) suggests that firms appear to agree
witb different accounting policies in seemingly similar situations. Studies done
in the context of other organizations (e.g., hospitals and colleges) would suggest
that performance differences might occur with organizational memory differences
(Thomas, Clark, and Gioia 1993; Milliken 1990). Future research
should consider performance/quality-related questions in more detail, although
access to data may be difficult to obtain.
Our research may also assist practitioners in developing consultation policies.
Practitioner interest is evidenced by a 1993 CICA-sponsored study group
that published a research report {not a standard) entitled "Guide for Developing
Quality Control Systems in Public Accounting" (MacLean 1993). As part of
this report the study group developed a statement of "best practices" on consultation
with peers (MacLean 1993, 90). The "best practices" consultation
statement is written at a high level of generality. Implementation of these "best
practices," however, would move ACUs that we classified as conditioned viewing
organizations towards the policies and procedures of ACUs we classified as
discovering organizations. The detailed ACU analysis provided in this paper
may aid in implementing these "best practices."
Finally, in this study we have described the interaction between an individual
audit partner who identifies an accounting issue for consultation and an
ACU partner. We have seen that in some firms this interaction results in an
entry into the firm's internal database, thereby potentially influencing future
decisions by other ACU personnel through on-going