The Impact of departmental interdependencies and management accounting system use on subunit performance

Abstract

This study examines the effect of departmental interdependencies and the use of management accounting information for decision-making on subunit performance. It is proposed that higher levels of departmental interdependence are associated with (i) the use of greater amounts of MAS information during task execution and (ii) a greater frequency of MAS information use, which, in turn, are associated with higher subunit performance. The empirical analysis based on data collected from 132 production managers through a survey provides some support for the proposition of an indirect effect of interdependence on subunit performance, acting through a use of greater amounts of MAS information for decision-making, but not for the proposition of an indirect effect acting through a more frequent use of MAS information. Furthermore, additional analyses indicate that the impact of the use of greater amounts of MAS information on subunit performance is substantially more positive for departments experiencing higher levels of interdependence compared with those experiencing lower levels.

Acknowledgements

I gratefully acknowledge the useful comments made by the anonymous reviewers, Jan Greve and participants at workshops at CEROC (Centre for Empirical Research on Organizational Control) at Örebro University. Financial support for this project was provided by Örebro University, and the Jan Wallander and Tom Hedelius Foundation.

Notes

¹The author is indebted to one anonymous reviewer for pointing this out.

²The relevance of this assumption should be contrasted with that implied by a moderating model, namely that managers do not adjust their MAS use, or, that MAS use is exogenously given and uncertainty is a moderator, not an antecedent (cf. Hartmann and Moers, 2003).

³As suggested by one anonymous reviewer and others (Chenhall, 2003; Gerdin and Greve, 2004), the assumption that managers typically adapt their MAS use to match the demands posed by the context may imply that optimal solutions are always used because of survival of the fittest conditions. As a result, the anonymous reviewer held that it is inappropriate to include a performance variable in the research model as every organisation, by definition, has optimal performance given its situation (i.e. there is simply no reason to expect any relevant cross-sectional performance differences). In this study, however, it is argued that there are always organizations moving towards congruent (survival) positions and vice versa (cf. Donaldson, 2001). Accordingly, there is reason to believe that there, at any point of time, should be variations in performance as the result of varying degrees of fit between interdependence and MAS use also when the general expectation is that managers typically adapt their MAS use to the situation.

⁴PAR is a company providing comprehensive data about Swedish firms, including addresses, company size, industry affiliation and positions.

⁵Note that the scale range of the original ‘importance’ score in Question 5 (1–7) was changed to 0–6. The premise is that the performance criteria that were considered to be of no importance (scored 0), arguably, should have no impact on the overall subunit performance score. However, this change proved insignificant in so far as the two scale alternatives were perfectly correlated (r=0.99, p=0.000). That is, practically none of the respondents marked that a performance criterion was ‘of no importance’.

⁶Arguably, a SEM analysis could have been performed to conduct measurement and structural analysis simultaneously. However, SEM was not performed because the departmental interdependence instrument only consists of one indicator and the MAS instrument is a composite measure, not a multi-item construct (the author is indebted to one anonymous reviewer for pointing this out), which reduces the effectiveness of this form of analysis.

⁷A substantial literature indicates that unit size is associated with greater use of formalized management accounting arrangements (Bruns and Waterhouse, 1975; Merchant, 1981, 1984; Innes and Mitchell, 1995; Bjørnenak, 1997). To control for the possible effect of departmental size on the results from the test of Hypothesis 1, the sample was divided into two subgroups. The results of the subgroup path analyses are provided in Table B.1 in Appendix B. It can be concluded that only minor differences in path coefficients between the ‘small size’ and ‘large size’ group exist, which suggest that the proposed relationship between interdependence, the amount of financial MAS information used during task execution and subunit performance holds across different departmental sizes.

⁸The author is indebted to one anonymous reviewer for pointing this out.

⁹In fact, data on the principles of unit grouping were available, in terms of whether the organization structure was functionally oriented or product (flow) oriented (cf. Galbraith's notion of self-contained tasks). An analysis showed a positive and significant correlation coefficient (r=0.2; p=0.02) between the amount of MAS information used and the reliance on product-oriented structures. However, the inclusion of the unit grouping variable in the statistical model (not shown here) did not change the results reported above.