This dissertation studies the interrelationship between management control systems (MCS), risk profile, and organizational performance by following a contingency approach. In doing so, an extension of Simons’
levers of control framework is put forward in order to integrate risk-based
dimensions of MCS and empirically test the extended framework. It is
hypothesized that the choice of MCS reflects the firm‘s risk profile, and
that firms that choose MCS design and use better suited to their risk profile perform better than others. Using data from a survey of 362 Chief Executive Officers of Austrian and German medium-sized companies in
the mechanical engineering industry, this study yields a model of fit that
relates types of risks to MCS design and use. I find that the choice
among specific control systems is associated with the types of risks an
organization faces. Specifically, it was found that preventable risks are
associated with an emphasis on formal controls, while strategy execution
risks and external risks are associated with an interactive use of control
systems. Furthermore, evidence on packages of MCS that are put in
place in practice is provided and configurations of rather loose and tight
MCS that differ in their risk profile derived. Finally, it is demonstrated that
a better fit between MCS and the organization’s risk profile is associated
with superior perceived and actual firm performance. This study contributes to an existing and growing body of literature in the field of management accounting and control concerned with implications from increased
uncertainty on MCS design and use. Besides contributing to academic
literature, the practical output of this dissertation is an integrated picture
of the interrelationship between specific control systems put in place and
types of risks an organization faces. These contingencies should be taken into account when setting up or developing MCS in organizations, as they enable the stimulation of selective improvements and help to
achieve a competitive advantage.