Alexandra Dawson Private Equity Investment Decisions in Family Firms: The Role of Human Resources and Agency Costs Journal of Business Venturing Final version: http://dx.doi.org/10.1016/j.jbusvent.2009.05.004 Citation for published version: Dawson, A. (2011). Private equity investment decisions in family firms: The role of human resources and agency costs. Journal of Business Venturing, 26(2), 189-199. 1 Introduction Family firms play a significant role in the global economy (Anderson and Reeb, 2003; Chrisman et al., 2007) and have been recognized as “a major source of oxygen for the combustion of entrepreneurship” (Rogoff and Heck, 2003: 559). However, many of them are unable to access the resources and capabilities needed to sustain competitive advantage and to grow, while several others are undergoing succession without any, or any suitable or interested, family successors (Howorth et al., 2004; Shanker and Astrachan 1996; Sirmon and Hitt, 2003; Upton and Petty, 2000). Opening up the family firm’s capital to private equity (PE) investors, through management buyouts and buyins (investments made together with existing management or with a new management team, respectively), is a viable solution to both of these problems. It can actually be the preferred route for the family, because PE is an alternative to selling out to another company or to going public. This has two advantages: continuity of the firm and, in some cases, sustained family presence in the business (Dreux, 1992; Howorth et al., 2004). From a PE firm’s perspective, family firms represent an important investment opportunity. First, family firms offer a vast potential deal pool, because they are the dominant form of economic enterprise throughout the world (Chrisman et al., 2003; La Porta et al., 1999) and often do not have the necessary resources to survive or grow. Second, family business owners are moved not only by economic but also by noneconomic objectives, such as creating jobs for family members (Chrisman et al., 2004; Corbetta and Salvato, 2004; Sharma et al., 1997). This gives PE investors an opportunity to increase the firm’s value by cutting agency costs through stricter governance systems (Jensen, 1993), and by promoting strategic managerial innovation through changes in organizational structure and managerial practices (Markides, 1998; Reid, 1996). Given that PE firms seek to take an equity stake in potentially high growth unquoted companies and to obtain high returns through an IPO or a sale to other investors or industry players (Mason and Harrison, 1999), they need to select their investments carefully. This type of investor is considered to be successful at predicting performance potential (Zacharakis and Meyer, 2000) and 2 bases the investment assessment on criteria that, in the strategy literature, are associated with superior firm performance (Shepherd, 1999). The aim of this paper is to shift the research focus from startups to family firms as investment targets and to assess whether the decision making criteria that PE investors use in selecting the latter are consistent with antecedents of firm performance and the presence of agency costs, as identified by family business scholars. Existing theory shows that performance outcomes in family firms result from the interaction of the business and the family (Habbershon et al., 2003). Traditionally, family firms have been represented through two (family and business) or three (family, ownership and managers) circle models (Gersick et al., 1997; Habbershon et al., 2003; Tagiuri and Davis, 1996). These overlapping subsystems have been helpful to describe individual and organizational behaviors, roles and perspectives and to explain how family relationships influence firm objectives and strategies (Sharma et al., 1997). Some more recent contributions to understanding family firm performance have come from mainstream theories. Applications of the resource based view (RBV) have identified sources of family firm capital leading to unique strategy making and competitive advantages over nonfamily firms (Habbershon et al., 2003; Sirmon and Hitt, 2003). Applications of agency theory have highlighted how asymmetric altruism can be associated with free riding and management entrenchment, leading to extraction of private benefits (Gomez-Mejia et al., 2001; Schulze et al., 2001, 2003). Thus, greater focus on economic objectives and better monitoring can lead to performance improvements. This research is based on a conjoint experiment made up of 1312 assessments by 41 PE professionals working in 35 PE firms, who were asked to evaluate their likelihood of investment in family firms. Given that decisions are nested within individuals and that individuals are, in turn, nested in organizations, data were analyzed using hierarchical linear modeling (HLM), which accounts for possible autocorrelation among observations. The paper makes three main contributions. First, this study contributes to the family business literature by addressing a gap in the literature regarding nonfamily succession routes, which have 3 not been fully investigated (Howorth et al., 2004). From a family firm’s perspective, an outside investor can help solve succession problems when there are no (or no suitable and/or interested) family heirs or when some family members wish to sell their shares and exit the firm. At the same time, this route offers other advantages including access to funds for the firm’s growth aspirations and acquisition plans (in addition to, or instead of, internally generated funds), preservation of independent ownership of the firm (as opposed to a sale to a competitor or another industry player), and a chance for family members to continue being involved in the firm (Corbetta, 1995; Wright and Coyne, 1985). Second, there have been several calls for more rigorous methodological research in family firm studies, which still rely largely on survey methods and case studies (Chrisman et al., 2005, 2007; Chua et al., 2003). This paper uses a conjoint experiment, capturing real time decisions and avoiding the pitfalls of questionnaires, such as recall bias and post hoc rationalization (Sandberg et al., 1988; Shepherd and Zacharakis, 1999). Conjoint analysis has been used in numerous studies on decision making, including consumer purchase choices, managers’ strategic decisions and venture capitalists’ selection of startups (Muzyka et al., 1996; Riquelme and Rickards, 1992; Shepherd and Zacharakis, 1999; Shepherd et al., 2000). Furthermore, this is one of the first studies to apply a multilevel model to the family firm context, by analyzing data (decisions, nested within individuals, nested within firms) with HLM (Eddleston et al., 2008). This technique has previously been employed by scholars investigating investor decision making (Choi and Shepherd, 2004), group behavior (Barsade, 2002), individual performance over time (Deadrick et al., 1997), and organizational performance (Chaganti and Damanpour, 1991; Hofmann, et al., 2000). Third, the paper contributes to investor decision making literature which, to date, has focused on startup selection (Franke et al., 2006; MacMillan et al., 1985; Meyer et al., 1993; Muzyka et al., 1996; Riquelme and Rickards, 1992; Shepherd, 1999; Shepherd et al., 2003; Tyebjee and Bruno, 1981, 1984; Zacharakis and Meyer, 2000; Zacharakis and Shepherd, 1999, 2001). Investment decisions in family firms have received little attention (Birley et al., 1999; Carter and Van Auken, 4 1994; Elango et al., 1995), despite the fact that family businesses account for a large proportion of investments, particularly in Europe, and are the single largest receiver of PE in some major European economies, including Italy, France, and the UK (CMBOR, 2005). The paper proceeds as follows: first, relevant family firm literature is reviewed and hypotheses are generated. Second, the research design is explained, including method used, sample selection, and data collection. Third, empirical results are presented. Finally, there is a discussion of key findings, followed by limitations of the study and implications for further research. 1. Theory and hypotheses 1.1. Distinctive resources in family firms In general, RBV scholars have highlighted the importance of resources and capabilities as key antecedents of a firm’s strategy and profitability. Competitive advantage has been linked to certain resources i.e., those that are valuable, rare, difficult to imitate and non-substitutable (Barney, 1991). Scholars applying RBV to family firms have stressed the importance of human resources in creating competitive advantage. Habbershon and Williams (1999) found that family firms have a distinctive bundle of resources and capabilities, which they termed familiness, resulting from the interaction between business and family, and leading to competitive advantage over nonfamily firms. Sirmon and Hitt (2003) identified the sources of family firm capital as being human capital (knowledge, skills and capabilities of individuals), social capital (relationships between the family on the one hand and stakeholders, such as suppliers and customers, on the other), and governance structure and costs (involvement of family owners/managers in the firm). Family ties contribute to building a family’s values and norms, and firm specific tacit knowledge (Lee et al., 2003; Sirmon and Hitt, 2003). Family members generally develop close relationships with the firm’s employees (Horton, 1986) and are able to gain in-depth understanding of their local environment, allowing them to identify emerging entrepreneurial opportunities more easily (Randøy and Goel, 2003). Research has 5 also found that family members are more productive than nonfamily members (Kirchhoff and Kirchhoff, 1987). However, not all family members working in a firm are valuable resources. Sometimes, they are chosen as a result of nepotism, birth order, or gender, rather than merit (Dyer, 1986; 2003). It is not easy for an outside investor to assess human capital in a target firm, since it is an intangible resource and there are no accepted models or methods for carrying out human capital evaluations in the context of PE investment (Smart, 1999). Therefore, one way of assessing family members’ quality is to consider whether they have previously worked outside the family firm. Past experience allows individuals to develop their cognitive models and helps them make more successful decisions (Hambrick and Mason, 1984). Work experience acquired outside the family firm increases heterogeneity of perspectives and beliefs and improves strategic decisions (Sirmon and Hitt, 2003). Furthermore, it makes it easier for family members to meet the high expectations of professionalism they need in order to be recognized as able managers by nonfamily employees (Aronoff, 1998; Salvato, 2004). Thus, H1. The likelihood of PE professionals investing in a family firm is higher if family members working in the firm have gained outside work experience. Family firm owners have a tendency to rely exclusively on family members because they often find it difficult to delegate to outsiders, have insufficient knowledge of formal management techniques, fear losing control, or believe that professionalization is an unnecessary cost (Dyer, 1989; Sharma et al., 1997). In turn, nonfamily managers frequently decide to stay away from family firms because they are likely to offer outsiders limited potential for professional growth, restrict their role to that of tutor, counselor or confidant, and exclude them from succession (Chua et al., 2003; Covin, 1994; Gallo and Vilaseca, 1998; Klein, 2000). However, nonfamily managers can play a critical role, as CEOs or executives, and have a positive impact on firm performance if they are included in strategic decision making (Chua et al., 6 2003; Gallo and Vilaseca, 1998). Many of them have formal business training and experience (Dyer, 1989), and may possess cultural competence, which helps them understand and be receptive to the socio-cultural configuration deriving from the family/firm interaction (Hall and Nordqvist, 2008). Similarly to family members, nonfamily managers tend to have idiosyncratic knowledge of the firm (Lee et al., 2003). Additionally, they enhance heterogeneity and add new perspectives that are not based on the family’s experiences. Another advantage over family members is that professional managers are generally less invested and not tied by emotional connections to the family and the firm, making certain sensitive decisions less difficult (Schein, 1995; Sirmon and Hitt, 2003). Thus, H2. The likelihood of PE professionals investing in a family firm is higher if there are nonfamily managers. 1.2. Agency costs in family firms According to traditional agency theory, agency costs in family firms are minimized or reduced to zero because principal (owner) and agent (manager) coincide (Jensen and Meckling, 1976). Familiarity between principal and agent, ease of communication, and cooperation among family members create a convergence of interests, thus avoiding the need for formal controls and incentive systems. This view is reinforced by the stewardship perspective, according to which there are altruistic behaviors among family members deriving from an alignment of objectives between family members and the organization (Davis et al., 1997). This type of altruism can be an important resource for establishing competitive advantage, if family members are highly dedicated to the success of the family firm and put its interests before their own (Corbetta and Salvato, 2004; Eddleston and Kellermans, 2007). Similar positive outcomes derive from psychosocial altruism, entailing the transfer of socially embedded values and norms from parents to offspring and leading to reciprocity, and simplified communication and decision making (Lubatkin et al., 2007). 7 Because family relationships are driven not only by economic but also by noneconomic objectives (Chrisman et al. 2004, Corbetta and Salvato 2004, Sharma et al. 1997), parents tend to be generous to their children, and family members help each other in times of need (Schulze et al., 2003). These behaviors, however, can also lead to negative consequences and to the emergence of agency costs (Burkart et al., 2003; Chrisman et al., 2004; Gomez-Mejia et al., 2001; Morck and Yeung, 2003; Schulze et al., 2001, 2003). First, they can be associated with unreciprocated generosity and manipulation on the part of some family members, leading to free riding and shirking. Second, if altruism is based on paternalism and if children perceive their parents as being coercive, offspring may rebel against their parents’ wishes, wearing down family bonds (Lubatkin et al., 2007). It is difficult to know whether stewardship or agency type relationships prevail in family firms (Chrisman et al., 2007). The incidence of agency costs varies from one family firm to another and may occur unevenly during the lifecycle of the same family firm (Sirmon and Hitt, 2003). However, we do know that PE investors face problems of asymmetric information when they are evaluating a potential target firm, because these are companies that are not quoted on the stock market and for which little information is available (Wright and Robbie, 1998). This issue is exacerbated when the target is a family firm, because these firms are characterized, more than others, by high levels of tacit knowledge possessed by family members (Howorth et al., 2004). Given that family firms generally pursue both economic and noneconomic goals and that family involvement can potentially lead to agency costs (Chrisman et al., 2004), in order to minimize their investment risk (Zutshi et al., 1999), PE professionals should assume that there are agency costs in the firm they are assessing. Furthermore, some actions are not considered to lead to agency costs in family firms (although they may reduce economic performance), such as an owner employing firm resources to pursue noneconomic goals by providing jobs for unqualified family members. However, these are considered to be agency problems in nonfamily firms (Chrisman et al., 2004) and are likely to be viewed as such by an outside PE investor. 8 Given that agency costs in family firms are associated with the family, a reduced family presence may therefore be an incentive to invest. First, if individuals who were previously receiving benefits, shirking or free riding exit the firm, the source of agency costs is removed. Second, if there are fewer (or no more) family members after the deal, it is easier for the PE firm to implement changes such as tighter monitoring, governance structures, and performance incentives (Robbie et al., 1999; Wright et al., 1994, 2001). Monitoring and incentive compensation have indeed been found to improve family firm performance (Chrisman et al., 2007). Thus, H3. The likelihood of PE professionals investing in a family firm is higher if there are family members wishing to sell their shares and exit the firm. Conflict can be another source of agency costs in family firms. Emotions are hard to avoid in family firms because family and business are so entangled (Boles, 1996; Harvey and Evans, 1994; Miller and Rice, 1998) and potential for conflict is greater than in nonfamily firms (Lee and Rogoff, 1996). Ownership fragmentation often enhances such tension, by causing sibling rivalry and disagreement between old and new generations. This leads to personal conflict, goal misalignment, diminished loyalty, and weaker commitment to the firm (Eddleston et al., 2008; Schulze et al., 2003). Some forms of conflict can be beneficial because they promote creativity and innovation and increase environment understanding and opportunity recognition. These include task conflict i.e., disagreement over what tasks should be pursued, and process conflict i.e., disagreement over how tasks should be carried out and how strategy should be implemented (Cosier and Harvey, 1998; Kellermans and Eddleston, 2004; 2007). However, other types of conflict, such as cognitive conflict concerning goals and strategies, are harmful for individual and group performance, reduce morale and productivity (Jehn, 1995), and are associated with declining performance in family firms or even family firm failure (Eddleston and Kellermans, 2007; Harvey and Evans, 1994; Olson et al., 2003). 9 A PE investor evaluating a family firm with fragmented ownership can face two types of dilemmas relating to conflict. First, there can be an immediate setback, if disagreement among a high number of family owners, and between PE firm and family members, leads to a breakdown in negotiations. Second, if the deal goes through and ownership fragmentation persists, there can be conflict between the PE firm and remaining family owners, as well as among remaining family owners. Ownership fragmentation can exacerbate the information asymmetry problem among vendors, purchasers, and financiers, which is typical of PE deals (Howorth et al., 2004). This can lead to agency costs deriving from differing interests, difficulty in observing behavior and asymmetric information between old/new and majority/minority owners (Chrisman et al., 2004; Jensen & Meckling, 1976). There can also be difficulties in implementing the PE firm’s chosen strategy, mismanagement, leadership inability, and difficulties with exit (MacMillan et al., 1985). These problems are tricky to deal with, especially for an outsider, because sources of conflict in family firms are complex and rooted in history (Haynes and Usdin, 1997; Kaye, 1991). Thus, H4. The likelihood of PE professionals investing in a family firm is higher if there is limited ownership dispersion. 2. Research design 2.1. Conjoint analysis Social judgment literature has highlighted the difficulties in identifying individuals’ decision making models. It is hard for decision makers to isolate the variables they use, identify the links between variables, and express the process by which they combine information into a decision heuristic. Furthermore, when they are asked how they have arrived at a decision, individuals are often inaccurate when they describe the heuristics they have used (Keats, 1991). Thus, in this study, it was decided to identify “theories in use” rather than to focus on “espoused theories”. Espoused theories are based on asking individuals to recall criteria used in their decision making or to assign relative importance to a list of predefined criteria. This type of retrospective reporting often causes 10