The Contingency Theory of Managerial Accounting: A Reply

Author(s): David C. Hayes

Source: The Accounting Review, Vol. 53, No. 2 (Apr., 1978), pp. 530-533
Published by: American Accounting Association
Stable URL: http://www.jstor.org/stable/245916 .
Accessed: 09/05/2014 15:16

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Accounting Association is collaborating with JSTOR to digitize, preserve and extend access to The
Accounting Review.

http://www.jstor.org

This content downloaded from 195.78.109.71 on Fri, 9 May 2014 15:16:39 PM

All use subject to JSTOR Terms and Conditions
THE ACCOUNTING REVIEW
Vol. LIII, No. 2
April 1978
The Contingency Theory
Accounting:
David C. Hayes
T IESSEN and Waterhouse (TW) raise
four questions concerning Hayes
[1977] which led them to conclude
that "' . . . the theoretical development
and empirical results on the causal links
between managerial behavior and sub-
unit effectiveness . .. were inadequate to
justify (the) criticisms of the accuracy
and importance of budgets and other
financial data as performance evaluation
devices."
This "conclusion" at the beginning of
their paper, differing as it does from that
at the end, is puzzling for two reasons.
First, no exploration was made or in-
tended between managerial behavior and
subunit effectiveness; in fact, individual
behavior items were completely excluded
from the questionnaire utilized, to which
TW acknowledge access. Second, the
implications drawn were, as is normal in
speculative extensions of results, in-
tended to explore plausible ramifications
of a study, couched in tentative language
and prefaced by the caveat " . . . (the)
exploration of some implications of this
finding is performed within the con-
straints on generalizability imposed by
the obvious limitations of the study"
[Hayes, 1977, p. 36] (emphasis added).
Apparently, they were not obvious to
TW. One of the difficulties in writing for
journals, particularly when the paper
involved is an attempt at a lucid and
cogent summary of a larger work such as
a dissertation, is meeting the space re-
quirements imposed. For this reason,
530
This content downloaded from 195.78.109.71 on Fri, 9 May 2014 15:16:39 PM
All use subject to JSTOR Terms and Conditions
of Managerial
A Reply
the limitations were not spelled out. To
indicate these briefly, which were eluci-
dated in Hayes [1975] and which TW
acknowledge perusing (though appar-
ently only as closely as they did the paper
on which they comment), I recognized
explicitly the difficulties with the opera-
tionalizations employed, the metrics, and
the small number of responses utilized.
Others can be added to that set which
could also be addressed to other ex-
ploratory studies of this type.
The language utilized and the warning
temper significantly the overall thrust of
the comments made by TW and the basis
upon which they drew their conclusions.
To some degree, they also answer some
of the specific points raised by TW; how-
ever, I shall respond to these more fully
in the order presented.
It is apparent with respect to the first
point raised, viz., the contingency dimen-
sions utilized in the study, that TW have
brought some of the confusion they
profess to find upon themselves. As they
acknowledge in footnote 3 of their com-
ment, the questionnaire did attempt to
measure types of contingency dimension.
In fact, considerable effort was expended
and several analyses were undertaken to
isolate and identify the types of con-
tingency faced by the various subunits
David C. Hayes is Assistant Professor,
Faculty of Commerce and Business Ad-
ministration, University of British Co-
lumbia.
Correspondence
[Hayes, 1975, pp. 70-98]. That these
were not separately identified in the
paper was a result of the space limitation
problem alluded to above. As they do
acknowledge, type and nature of con-
tingency were the central concern and,
thus, consistent with contingency theo-
ries of organization. Perhaps, I should
have revised the wording of the proposi-
tions in the paper to show that the specific
type analyses were not included, but I felt
that this was inappropriate. Simply, their
comment is not valid in the context of the
study as a whole-to which they con-
stantly refer. I confess there may be some
confusion for those without access to the
complete research report, but not, I think,
of a major degree.
With respect to the comments on ex-
planatory variables, TW make a number
of points, some of which I have difficulty
in following. As noted by TW, the "bud-
get" and "financial" variables are per-
ceptual in nature; their intent was to
elicit perceptual data on the ability of
such measures to reflect performance.
Interpreting them as inputs or determi-
nants of effectiveness may be stretching a
point, but it was difficult to obtain data
on relative explanatory abilities in an-
other way. Given the desire to compare
accounting type explanators of perfor-
mance with non-accounting measures,
this was the route chosen. Again, the
difficulty of operationalization is ac-
knowledged, though "illegitimacy" as a
description of the approach strongly
overstates the case.
Their second point with respect to the
classification of these variables, that there
is confusion between an implied surroga-
tion of performance explanators and
categorization only as internal variables,
indicates confusion on the part of TW
between normative and descriptive state-
ments. The surrogation implication
represents a normative ideal; the cate-
This content downloaded from 195.78.109.71 on Fri, 9 May 2014 15:16:39 PM
All use subject to JSTOR Terms and Conditions
531
gorization an attempt at description.
The third question they raise concern-
ing the inclusion of judgements from
other department and own department
managers as representatives of interde-
pendence and internal variables again
illustrates that TW did not, as they
profess, read Hayes [1975] for the more
complete description of why particular
research steps were undertaken. As was
indicated explicitly therein [Hayes, 1965,
pp. 98-100], the nature of the question-
naire hopefully led respondents sequen-
tially through the questions. The respon-
dents came upon the summaryjudgement
variables subsequent to a set of questions
dealing either with items internal to their
own department or specifically related to
their interactions with other departments.
Clearly, it is not and was not possible to
get into the heads of respondents to find
out exactly how they interpreted and
responded to these questions. A judge-
ment was made (and explicitly acknowl-
edged) that it was appropriate to treat
these variables in such a fashion; as with
all such judgements made in the course of
empirical research, it can be reasonably
debated. By implication, TW are sug-
gesting these variables should have been
treated as dependent variables which
would imply an arbitrary weighting
scheme and effective inclusion of the
same variable on both sides of the re-
gression equation. Is this what they would
term "legitimate?"
Finally, as indicated by the path
analytic representations in the paper
[Hayes, 1977, p. 32, 33, 35], the correla-
tions between variables are non-zero.
Further, there is no suggestion made of
orthogonality of categories anywhere in
the paper. To do so, as implied by TW,
indicates a lack of understanding of open-
systems theories of organizations. By
their very nature, there is a necessary
inter-relatedness between all variables
532
impacting on organizational outcomes.
TW would be well advised to read the
contingency theory and open-systems
literature (e.g., Thompson [1967]; Law-
ence and Lorsch [1969]) for reinforce-
ment of this point. The point TW
attempt to make concerning planning in
this context indicates that they fail to
recognize that planning activity is not
unidimensional. As indicated in my
research, planning is utilized with respect
to environmental contingencies which
may act on the firm and efforts to mediate
them, as well as with respect to the se-
quencing and coordinating of inter-
subunit activities, as discussed in Thomp-
son (1967). Clearly, these are different
planning activities and the measures used
reflected this.
Thus, the series of accusations made by
TWconcerning confusion of independent
variables lack foundation, both singly
and jointly.
With respect to the choice of controller
judgements as the dependent variable in
the analysis, the third point raised by
TW, I am not sure how to respond. The
rationale was indicated [Hayes, 1977, p.
33, note 6], though apparently ignored
by TW. I am sorry they find its choice
"highly suspect," but short of an en-
tirely different data collection procedure
which would have entailed significant
additional cost, there did not seem to be
any way around this problem. To reiter-
ate a point made before, there are many
judgements undertaken in such research;
I indicated where I made mine and why.
That TW did not read these reasons
would seem to be their problem.
Finally, with respect to the factor
analyses, the points made by TW are
well-taken. I recognize, as they do, the
problems in its utilization and interpre-
This content downloaded from 195.78.109.71 on Fri, 9 May 2014 15:16:39 PM
All use subject to JSTOR Terms and Conditions
The Accounting Review, April 1978
tation. Certainly, different procedures
produce different variables loading on
factors and the problems related to a
small n may be as valid as they suggest.
But it should be recognized that such
difficulties as are indicated are a frequent
problem in survey research not unique to
this study. Armchair speculation on other
likely results from different utilization of
the technique does not seem to me to be
greatly productive.
CONCLUSION
While proceeding through this exercise
of responding to my criticis, some of my
spleen has been vented at the style of
their remarks. Many of their points
could have been raised without the use of
emotive, value-laden terms such as
"valid," "illegitimate," and "incorrect"
which are rarely, given their absoluteness,
reasonable terms with which to describe
research. Shades of grey dominate the
area, not the extremes of right and wrong
or black and white. In fact, I would sug-
gest that the very overstatement they
have indulged in led them to have too
much faith in what they were doing,
failing to recognize the shortcomings
and contradictions within their own
comments.
Their parting shot, reference to the
work of Pennings [1975], is as strange as
some of their other comments, for there
are a number of theoretical and methodo-
logical difficulties with that work.
Finally, no claim was made or in-
tended by me that my research was
"ideal" or "perfect" in any sense. There
are a number of flaws, many of which
were indicated in its reportage, others
which I may keep to myself. TW cer-
tainly missed them.
Correspondence 533

REFERENCES
Hayes, David C., "The Contingency Theory of Managerial Accounting: An Empirical Test of an Assess-
ment Model" (Ph.D. diss., The Ohio State University, 1975).
, "The Contingency Theory of Managerial Accounting" THE ACCOUNTING REVIEW(January 1977),
pp. 22-39.
Lawrence, Paul R. and Jay W. Lorsch, Organization and Environment(Irwin, 1969).
Pennings, Johannes M., "The Relevance of the Structural-Contingency Model for Organizational Effec-
tiveness" AdministrativeScience Quarterly (September 1975), pp. 393-410.
Thompson, James D., Organizations in Action (McGraw-Hill, 1967).

This content downloaded from 195.78.109.71 on Fri, 9 May 2014 15:16:39 PM

All use subject to JSTOR Terms and Conditions