Journal of Small Business and Enterprise Development vol:21 issue:3 pages:470-488
◦Purpose: This study examines whether CEOs place their distinctive marks on the design of management control systems (MCS) in SMEs. ◦Design/methodology/approach: Based on upper echelons theory and using survey data from 189 Belgian SMEs, we investigate whether CEO demographics are a significant determinant in explaining the design of planning, control and evaluation systems (or MCS) in addition to the explanatory power of contextual influences. ◦Findings: CEO demographics are significantly associated with the design of the evaluation system in place. CEOs with longer tenure and higher education evaluate subordinates using another style of evaluation than their shorter tenured and lower educated counterparts. Planning and control systems are only significantly influenced by organizational and environmental characteristics. Our findings confirm that managerial discretion differs for different managerial activities (here MCS elements) (Finkelstein and Peteraf, 2007). Discretion is absent in relation to planning and control systems. This finding is probably due to the more external and observable character of these systems, which results in pressures to comply with institutional norms translated into so called “good practices”. Discretion in relation to the design of performance evaluation systems is high, which can be explained by the internal character of these systems. ◦Practical implications: Since evaluation systems are an important determinant of work-related attitudes and could lead to dysfunctional behavior, it is important for company owners and board members to consider the demographics of present or new CEOs, and to understand the associated inclinations reflected in evaluation systems. ◦Originality/value: In this study we apply a more comprehensive approach than (the few) extant SME studies by relating a larger number of CEO demographics to a more comprehensive set of MCS elements, simultaneously controlling for a larger group of contingent variables. Moreover, we respond to calls in the upper echelons and MCS literature (Carpenter et al., 2004; Chenhall, 2003, Naranjo-Gil and Hartman, 2007).