Management accounting systems and institutionalization in medium-sized and large family businesses--empirical evidence from Germany and Austria.

• Author: Neubauer, Herbert

1. INTRODUCTION

   Medium-sized enterprises and family businesses constitute a highly important economic factor in Germany and Austria. In both countries, more than 99% of all businesses can be classified as mediumsized enterprises. In Austria, 70 to 80 percent of all enterprises are family businesses (Frohlich, 1995; Hasch et al., 2000), and the corresponding figure for Germany is approximately 95 percent (Haunschild et al., 2007; Klein, 2000).

   These medium-sized enterprises (MEs) are characterized by limited resources in terms of capital, personnel (including management) and time. It is specifically those resource deficits which limit these companies in terms of management. While the term "business management system" refers to all tools, policies/rules, institutions and processes with which leadership tasks (e.g. definition of objectives, planning, decision-making, control and monitoring) are performed, management accounting is understood as a subsystem of business management which serves the purpose of coordinating information provision, planning and monitoring in an enterprise. Family businesses (FBs) are a very special type of enterprise in which the "family" system exercises a strong socioeconomic influence on business policy. Although a certain body of empirical research on business management in FBs has emerged at the international level (Dyer, 2003; Sharma, Chrisman and Chua, 2003; Sharma, 2004; Casillas, Acedo and Moreno, 2007), the topic of management accounting in FBs has only seen marginal attention in empirical research (Salvato and Moores, 2010). With regard to controlling (the term coined to denote the rough equivalent to management accounting in German-speaking countries), we have observed an increasing number of empirical studies on the topic in large enterprises (LEs) as well as medium-sized enterprises (e.g. Ossadnik, Barklage and Lengerich, 2004; Weber et al., 2006; Wagenhofer, 2006), but aspects related to FBs have seen only little attention in research (Schachner, Speckbacher and Wentges, 2006; FeldbauerDurstmuller, Wimmer and Duller, 2007). No multi-country empirical studies, e.g. on the manifestations of controlling, have been carried out to date.

   In this context, the question arises whether there are major differences between FBs and non-family businesses (NFBs) with regard to the formalization of controlling, which manifests itself in the establishment of a separate controlling unit. The relevant literature has proposed the hypothesis (which has not yet been verified empirically by a comprehensive multi-country study) that FBs have substantially less formalized controlling systems compared to NFBs ( Morris et al., 1997; Picot, 2008). In addition, it is not known whether the institutionalization of a controlling system is influenced more by business size or the differentiation between FBs and NFBs. The discussion above gives rise to the research questions addressed in this article:

1. Are there significant differences between FBs and NFBs with regard to the manifestation of controlling systems?

2. Are there significant differences between FBs and NFBs with regard to the use of controlling tools?

3. Which influence dominates in the formalization of controlling and in the use of controlling tools: business size or the differentiation between FBs and NFBs?

   In an attempt to find answers to the questions above, an empirical study was conducted among German and Austrian enterprises. The remainder of this article is structured as follows: Section 2 presents the reference theories used as a basis of the article, which are essentially systems and contingency theory as well as principal-agent and stewardship theory. In addition, Section 2 provides definitions of the terms "medium-sized enterprise" and "family business". The measurement and assessment of the question whether an enterprise is a family or non-family business is carried out on the basis of (and by applying) the theoretical concept of substantial family influence (SFI). Based on the theoretical background and previous empirical findings. Section 3 starts by discussing the state of the art in controlling research among medium-sized FBs. Section 4 then presents the development of hypotheses for the multi-country study. Section 5 describes the study methods used, the responses received, demographic data on the respondent businesses as well as the findings of the study, including an evaluation of the hypotheses. The final section summarizes the key results and provides an outlook with questions and avenues for further research.

2. THEORETICAL BACKGROUND

   2.1 Systems and contingency theory

   The first reference theories underlying the study are systems theory and contingency theory. As defined by Ulrich (1970), a system refers to the "ordered totality of all elements between which any relat ionships exist or may be established." The ideas underlying systems theory are closely interrelated with cybernetics, meaning that one can speak of systems theory-based, cybernetic systems of statements. Interpreted under systems theory, controlling mainly takes on a coupling (and thus coordinating) function for the individual subsystems within an enterprise (Horvath, 2009).

   Contingency theory, which examines the influence of environmental factors on the manifestation of a target variable was also used by Gordon and Miller (1976) and Khandwalla (1977) as an explanatory theory in management accounting and controlling research (see also Lawrence and Lorsch, 1967). Contingency theory attempts to generate empirically grounded statements about relationships between situational conditions, organizational structures, the behavior of organization members as well as the degree to which an individual organization attains its objectives (Kieser and Walgenbach, 2007), and it is therefore frequently used as a theoretical basis in accounting and controlling research (Luft and Shields, 2003; Chenhall, 2007; Byrne and Pierce, 2007). In this context, empirical studies on the link between various contextual variables and controlling as a dependent variable were also conducted in Germanspeaking countries at a very early stage, for example in the study by Uebele (1981). However, current international research on contingency theory, including Chenhall (2007), is based on the assumption that contingency theory holds only limited explanatory potential in questions related to management accounting and controlling, thus calling for an integration of theories. Taken individually, the two theoretical constructs - systems theory and contingency theory - each exhibit certain disadvantages, which is why it appears necessary to integrate the theories in empirical research on controlling.

   2.2 Principal-agent and stewardship theory

   Principal-agent theory has been used in both theoretical and empirical works to explain the differing manifestations of ownership, control and management structures as well as the differences in performance between FBs and NFBs (Hack, 2009). In general, principal-agent theory (Jensen and Meckling, 1976; Arrow, 1985; Eisenhardt, 1989) describes possible conflicts of interest between at least two parties to an agreement, specifically the principal and the agent. A typical example of the principalagent problem in FBs is the relationship between a shareholder and a manager who assumes responsibility for managing the business on behalf of the principal. Under this theory, the conflict of interest is analyzed within the framework of a contractual relationship which is characterized by asymmetric levels of information, bounded rationality, opportunistic behavior and incomplete contracts. The fundamental assumption in principal-agent theory is that the principal has an information disadvantage in relation to the agent and cannot monitor the agent's activities comprehensively and without incurring costs (Jacobides and Croson, 2001). This means that the provision of information throughout the performance process constitutes a crucial element in the principal's process of monitoring the agent. The installation of a separate controlling unit and the implementation of formal controlling tools can therefore represent a suitable approach for the establishment of formalized control and monitoring systems.

   Stewardship theory and the concept of altruism are especially important in the context of FBs (Dyer, 2003; Corbetta and Salvato, 2004; Songini, 2006; Witt, 2008; Le Breton-Miller and Miller, 2009; Vallejo, 2009). While principal-agent theory assumes opportunism on the part of the agent, stewardship theory postulates that agents will act in accordance with social principles and devote their efforts to the common good (Davis, Schoorman and Donaldson, 1997). Cooperative, selfless behavior is assumed among members of families in particular (cf. Schulze, Lubatkin and Dino, 2003). Altruism among family members and the reduction of transaction costs are supported by the case study in Karra, Tracey and Philips (2006) as well as additional empirical findings (Chrisman, Chua and Litz, 2004). Formalized corporate governance mechanisms are considered dysfunctional in this context because they precipitate a loss of trust and may demotivate stewards, who are primarily motivated intrinsically (Davis, Schoorman and Donaldson, 1997). These arguments warrant the conclusion that FBs do not require control and monitoring systems to the same extent as NFBs.

   2.3 Controlling: Conceptions and definition

   As a praxeological phenomenon, controlling did not arise from scientific theory, but from practice. While management accounting research in Anglo-American countries has advanced rapidly in both theoretical and empirical terms (e.g. Chenhall, 2003; Luft and Shields, 2003), controlling research in Germanspeaking countries is still in the midst of a debate on various possible conceptions of the discipline. Although controlling and management accounting are not exactly the same thing, it is...