Governance Systems in Family SMEs: The Substitution ... · governance mechanisms and family gover-nance mechanisms. Particularly, the paper’s main objective is to explore the roles

Governance Systems in Family SMEs: The SubstitutionEffects between Family Councils and CorporateGovernance MechanismsBy Luca Gnan, Daniela Montemerlo, and Morten Huse

The main objective of this paper is to explore the role of family councils vis-à-vis corporategovernance mechanisms. Particularly, the paper explores whether family councils perform onlytheir distinctive family governance role or if they also substitute for the roles performed bycorporate governance control mechanisms. Based on a sample of 243 Italian family SMEs, ourresearch findings show that the family council partially substitutes the shareholders’ meeting andthe board of directors in playing their respective corporate governance roles of ownership andmonitoring. These findings are interpreted in the light of both agency and relational perspectives.

IntroductionThis paper considers governance systems in

family small and medium-sized enterprises(SMEs), governance system meaning the set ofgovernance mechanisms—both individual andcollective—in charge of directing and control-ling an organization (Charkham 1994).

Corporate governance systems are typicallyhierarchical (Johnson and Scholes 2002), andthey are based on agency theory assumptionsabout the need for contractually assigning dis-tinct governance roles to governance mecha-nisms at different levels (Jensen and Meckling1976). There are three basic corporate gover-nance roles and mechanisms from this perspec-tive: the ownership role, the monitoring role,and the leading role. The shareholders’ meeting

plays the ownership role of control over theboard of directors (Charkham and Simpson1999); the shareholders’ meeting also delegatesto the board a monitoring role over the CEO andhis/her team (Charkham and Simpson 1999;McNulty and Pettigrew 1999); and the CEO isassigned the role of leading the company (Fayol1949; Finkelstein and Hambrick 1996; Minichilli,Corbetta, and MacMillan 2010). Every roleimplies different tasks to be performed by eachmechanism (Huse 2003; Zahra and Pearce1989).

Corporate governance systems can be quitearticulated, and not only in large firms, but inSMEs as well (Brunninge, Nordqvist, andWiklund 2007; Gnan and Montemerlo2006; Gubitta and Gianecchini 2002; Zahra,Neubaum, and Huse 2000). Still, and despite

Luca Gnan is Professor of Organizational Behaviour at Department of Economics and Finance, Tor VergataUniversity of Rome.

*Daniela Montemerlo is Professor of General Management and Management and Governance of FamilyFirms at the University of Insubria, and Professor of Strategy and Family Business at SDA Bocconi School ofManagement.

*Morten Huse is Reinhard-Mohn-Chair of Management and Governance at the University of Witten/Herdecke, and Professor of Organisation and Management at BI Norwegian Business School.

[*Corrections added on 15 November 2013, after first online publication on 25 October 2013: Theaffiliations of Daniela Montemerlo and Morten Huse have been updated.]

Address correspondence to: Luca Gnan, Tor Vergata University of Rome, Economics and Finance, ViaColumbia 2, Rome 00133, Italy. E-mail: [email protected].

Journal of Small Business Management 2015 53(2), pp. 355–381

doi: 10.1111/jsbm.12070

GNAN, MONTEMERLO, AND HUSE 355

mailto:[email protected]

most firms being small and medium-sized andfamily-owned, the topic of governance infamily SMEs is relatively recent, and it mostlyfocuses on boards of directors (BODs). In par-ticular, studies show that the BODs of thesefirms have peculiar characteristics (Corbettaand Tomaselli 1996; Fiegener 2005; Forbes andMilliken 1999; Pieper, Klein, and Jaskiewicz2008) and they may contribute strongly to valuecreation (for example Castaldi and Wortman1984; Huse 2000; Van den Heuvel, Van Gils,and Voordeckers 2006).

Our study takes a broader perspective, as itinvestigates family SMEs’ governance systemsand their various mechanisms at both the cor-porate level—that is, looking at the agency-based hierarchy—and the family level.Governance systems, mechanisms, and rolesare relevant concepts at the family level as wellas in firms, as the family has to be directed andcontrolled to protect both its intangible assets(unity, trust, values, and so on) and its tangibleones; when the tangible assets includecompany ownership, the focus of family gov-ernance is the relationship between family andbusiness. The most important family gover-nance mechanism is the family council, that is,regular gatherings of adult family members(sometimes all the family members; at othertimes a group of family representatives)entitled to design and manage the family–firmrelationship in a certain generation and to planit for the next generation (Carlock and Ward2001; Gallo and Kenyon 2004).

The family council is generally analyzed froma relational perspective. Particularly, steward-ship studies highlight that family councilsfacilitate social interaction, which develops acommon sense of stewardship and consequentlyunity in family ownership and the commitmentof each individual member to protect the busi-ness continuity and pursue the family owners’common interests (Mustakallio and Autio 2001;Salvato 2002). This is one of the few studies onfamily councils in SMEs, as they are generallyanalyzed in the context of large families owninglarge groups (Gnan and Montemerlo 2006; Suàreand Santana-Martìn 2004).

This is also one of the few studies thatexamine the interaction between corporategovernance mechanisms and family gover-nance mechanisms. Particularly, the paper’smain objective is to explore the roles and tasksof family governance mechanisms like familycouncils vis-à-vis corporate governance mecha-

nisms such as shareholders’ meetings, BODs,and chief executive officers (CEOs). Based onan exploratory survey of 243 Italian familySMEs, we found that a substitution effect oper-ates and makes the family council exceed itsfamily governance role by taking on some rolesof the corporate governance hierarchy, notablysome ownership roles of the shareholders’meeting and some monitoring roles of theboard of directors. The findings are interpretedin the light of both relational theories andagency theory, joining the theoretical conversa-tion on the importance—to which scholars aremore and more sensitive—of matching the rela-tional perspective of stewardship and the con-tractual one of agency to go deeper intoshedding light on family business governance(Braun and Sharma 2007; Corbetta and Salvato2004; Dyer 2006; Miller and Le Breton-Miller2006). The exploratory nature of the study ismainly due to the survey’s low response rate(and also to some missing information aboutthe overlapping of governance roles playedby individual family members that is typicalof SMEs and that will be addressed in thediscussion and limitations sections). Also, inItaly BODs are not mandatory, as they are inother countries, thus limiting the findings’generalizability. However, as not much empiri-cal evidence is available on family and corpo-rate governance in SMEs, we found that evenan exploration could make a contribution, andso it was worth pursuing with the availabledata.

As regards contributions, the paper makesempirical as well as theoretical and practicalones.

Empirically, the paper contributes to adeeper understanding of governance in familySMEs, as it explores governance systems withregard to their family and corporate articula-tion, the roles that various mechanisms play,and the extent to which they play them.

Theoretically, the paper contributes to theconversation on agency and relational perspec-tives, showing that they complement eachother. In fact, on one side, corporate gover-nance systems are agency-based in terms of thehierarchical mechanisms they feature and theroles these mechanisms are assigned legallyand contractually. On the other side, wesuggest that the family council might intervenein the corporate governance hierarchy andreplace part of its control roles for relational,stewardship-based reasons.

JOURNAL OF SMALL BUSINESS MANAGEMENT356

A practical contribution of the study is that itreflects on the possible pros and cons of thesubstitution of corporate governance mecha-nisms by the family council and on the impli-cations for the family and the firm; thereby, itmay help SME-owning families design andimplement their governance systems.

The paper proceeds in four sections. First,the reference theories and concepts are pre-sented and the hypotheses are derived. In thesecond section we describe the researchmethods and the sample used to test thehypotheses. The main findings are shown inthe third section. In the fourth section wediscuss these findings, reflecting on the limita-tions, implications for family firms, and direc-tions for future research.

Agency and RelationalPerspectives: Literature Reviewand Hypotheses

In this section we present our theoreticalframework and derive three hypotheses on therelations between the family council and somecorporate governance mechanisms.

The Agency-Based ContractualPerspective

Agency theory is one of the literature main-streams that looks at corporations from a con-tractual perspective. The key issue is how toalign the interests of owner–principals withthose of manager–agents, overcoming theagency threat that brings about various agencycosts for performing ad hoc monitoring activi-ties and operating systems (Eisenhardt 1989;Fama 1980; Fama and Jensen 1983; Jensen andMeckling 1976). This is achieved through aclearly defined principal–agent hierarchy ofcorporate governance mechanisms with dis-tinct roles. The shareholders’ meeting, theBOD, and the CEO are the key corporate gov-ernance mechanisms in this hierarchy, andthey take on three basic corporate governanceroles, each one with specific tasks. First, theshareholders’ meeting has an ownership roleof shareholders’ control over the board; thatis, the main ownership tasks include thechoice of board members and the control andevaluation of the board and its performance(Charkham and Simpson 1999). Second, theBOD is assigned a monitoring role on behalfof the owners, which implies control overstrategy and management (Charkham and

Simpson 1999; McNulty and Pettigrew 1999).So the monitoring tasks include, in essence,choosing the CEO, controlling and evaluatingthe CEO and his/her team, taking strategicdecisions, and approving strategic plans(Hillman, Cannella, and Paetzold 2000;Hillman and Dalziel 2003; Johnson, Daily, andEllstrand 1996; Nicholson and Kiel 2004).Third, the CEO is delegated his/her own roleof leading the firm by elaborating the strategyand by coordinating and controlling how theorganization executes it, which includes suchtasks as choosing and evaluating managers,formulating strategic plans, and formulatingand controlling budgets to organize andmonitor strategy implementation on an annualbasis (Fayol 1949; Finkelstein and Hambrick1996; Minichilli, Corbetta, and MacMillan2010). External corporate governance mecha-nisms, such as regulations, markets for man-agers, and markets for control, are alsoconsidered by agency theory, but they aretypical of public corporations and they mightbe either absent or not developed in SMEs(Fama and Jensen 1983).

Various agency scholars (for exampleEisenhardt 1989; Fama and Jensen 1983; Jensenand Meckling 1976) argue there are no relevantagency costs in owner-managed firms, that is,in many family SMEs (Gersick et al. 1997; Ward1987 and 2004). In fact, in these firms theowners’ and managers’ roles are often playedby the same people, which naturally alignstheir interests and makes the principal–agenthierarchy much less defined, mixing up differ-ent governance roles and tasks (Fama andJensen 1983, p. 307). Nevertheless, internal cor-porate governance mechanisms remain subjectto agency-based norms that must be accom-plished, whether such mechanisms are workingor not. What is more, as soon as the companyevolves from the pure “owner–manager”model, the typical agency threats reemerge,and so does the need to make the principal–agent hierarchy work. This happens, forinstance, in family firms (including SMEs) whennonfamily managers are hired, or when own-ership becomes differentiated between manag-ing and nonmanaging family owners withgenerational transitions (Chua, Chrisman, andSharma 2003; Corbetta and Montemerlo 2003;Gnan and Montemerlo 2006; Gubitta andGianecchini 2002).

The family nature of firms is also tradition-ally assumed to reduce agency costs, as the


family relations between the key decision-making actors imply higher interest alignmentand fewer information asymmetries betweenowners and managers. Besides, these relationsdevelop over a long-term horizon and they arebased on kinship, sentiments, trust, and recip-rocal altruism, which are supposed to counter-balance the opportunistic, profit-maximizingbehaviors assumed to be intrinsic by agency; allthis is supposed to reduce the moral hazardand related monitoring requirements (Famaand Jensen 1983; Harvey 1999; Jones 1983;McConaughy 2000). However, other studiesshow that the family nature can actually bringabout family-specific agency problems that maystem from asymmetric altruism (Anderson andReeb 2003; Chrisman, Chua, and Litz 2004;Schulze, Lubatkin, and Dino 2003; Schulzeet al. 2001), conflicts of interest between familymembers in different roles (Chrisman, Chua,and Litz 2004; Schulze et al. 2001), conflicts ofinterest between family and non-familymembers (Chrisman, Chua, and Litz 2004;Daily and Dollinger 1993; Villalonga and Amit2006), and conflicts of interest between family(majority) and family or nonfamily (minority)shareholders (Villalonga and Amit 2006). Spe-cific agency problems include nepotism, incon-gruity between the executives’ and the family’sgoals, a lack of market discipline, self-control,adverse selection, managerial entrenchment,and moral hazard (Bruce and Waldman 1990;Buchanan 1975; Gomez-Mejia, Nunez-Nickel,and Gutierrez 2001; Jensen 1998; Morck,Schleifer, and Vishny 1988).

In summation, agency theory still remains areference theory for family SMEs. Not all ofthem are “purely” owner-managed despite theirsmall and medium size; that is, agency issuesmight be sometimes, but not always, reduced,as a small or medium size can coexist witharticulated ownership and management. More-over, family-specific agency problems may existanyway. Also, agency theory inspires the hier-archical form and the legal obligations of gov-ernance systems, regardless of size, in mostcountries. However, no definite conclusionshave yet been reached about the “net effect” ofsize and family factors on the overall agencycosts and on the consequent corporate gover-nance needs. Besides, the extant researchshows that corporate governance mechanismsare often held “on paper,” questioning theactual efficacy of agency-based systems(Corbetta and Tomaselli 1996; Huse 2000;

Melin and Nordqvist 2000). In addition, agencytheory is generally acknowledged not toventure deep into the family nature of firms—which far exceeds its contractual-based logic—and particularly not to capture all thedistinctive traits of family ownership and theirimplications for family business governanceand management.

To contribute to solving such controversies, itis essential to check the extent to which corpo-rate governance mechanisms such as sharehold-ers’ meetings, BODs, and CEOs play theiragency-based control roles. In any case, agencycan hardly be considered the only theoreticalreference for family business governance. Arelational perspective, and particularly a stew-ardship one, is required as a complement, andwe deal with it in the following section.

The Relational PerspectiveThe relational perspective focuses on

common interests as purposes for whichhuman beings are intrinsically motivated andon how to accomplish them by means of gov-ernance systems that operate by a cooperative,noncontractual logic. As this logic is prominentin family firms, the relational perspective hasemerged as a particularly interesting onethrough which to integrate agency into familybusiness studies (Corbetta and Salvato 2004;Mustakallio and Autio 2001; Miller and LeBreton-Miller 2006).

The relational perspective encompassesvarious theories, such as stewardship, socialcapital, and resource dependence. We focus onstewardship theory, which traditionallydescribes agents’ behavior in terms opposed toagency theory. In fact, it represents agents withcooperative and pro-organizational attitudesand with a natural propensity to align theirgoals with those of the principal, due to anumber of conditions like intrinsic personalfeatures, needs and motivations; identificationwith the company and commitment tocompany values; power intended as a service; acollectivistic company culture; and a participa-tive and trust-oriented management philosophy(Donaldson and Davis 1991; Miller and LeBreton-Miller 2006; Salvato 2002). Stewardship-based values and behaviors are acknowledgedto favor family firms’ success in the long run(Aronoff and Ward 1992; Salvato 2002).

To develop stewardship-based values andbehaviors, family governance mechanisms arefundamental, as they keep the interface


between family and business healthy by faci-litating social interaction among familymembers. The most important family gover-nance mechanism, as stated above, is the familycouncil; other mechanisms may be present,especially in large and complex family groups,such as family assemblies, family committees,and family offices (Carlock and Ward 2001;Gallo and Kenyon 2004; Lank and Ward 2000;Lansberg 1999; Ward 2004). Given their limiteddiffusion in family SMEs, we did not considerthem.

The family governance role is composed oftwo tasks typically assigned to family councils(Carlock and Ward 2001; Gallo and Kenyon2004). One task relates to designing and man-aging the relationship between family and busi-ness in the controlling generation. The otherone involves planning the family–firm relation-ship for the next generation: in essence, it isabout planning generational transition. Thesetwo tasks encompass a number of moredetailed and interdependent activities (whichmay take place jointly and thereby be difficultto isolate in practice), such as discussing andupdating the family values; setting principlesand policies about family, ownership, and busi-ness for leading and the next generation; car-rying on educational and recreational initiativesfor present and future family owners; managingcommunication within the owning family; for-mulating the owning family’s shared vision ofthe family business and communicating it to theBOD; and keeping spouses, who are notowners but who groom future owners,informed about the business (Carlock and Ward2001; Gallo and Kenyon 2004; Lansberg 1999).

In performing the mentioned tasks, familycouncils act as relational contexts that allowfamily members to express and discuss opin-ions, develop mutual trust, share values andvisions, and translate them into collective plansand actions more effectively than might happenwith corporate governance mechanisms for-mally regulated by law and practice, like share-holders’ meetings and BODs (Habbershon andAstrachan 1997; Mustakallio and Autio 2001;Miller and Le Breton-Miller 2006). In turn, trust,shared values, and vision nurture the unity andcommitment of family and nonfamily actors,considered by scholars to be key conditions forfamily firms’ success in the long run (Galloet al. 2001; LaChapelle and Barnes 1998).

It has to be noticed that the studies men-tioned mostly analyze family councils in large

families in control of large firms, showing aprogressive institutionalization process of suchmechanisms (Melin and Nordqvist 2007),whereas hardly any research exists on familycouncils in SMEs. However, the very fewexploratory studies available suggest thatfamily SMEs often make use of family councilsfor family governance (Gnan and Montemerlo2006; Suàre and Santana-Martìn 2004), whichmakes them worth investigating for this type ofcompany as well.

In addition, the literature suggests that, dueto their relational strengths, family councilsmay exceed their family governance role andbecome involved in corporate governance(Gersick et al. 1997; Lansberg 1999; Lank andWard 2000; Ward 1987 and 1991). This makes itessential to check which governance controlroles are actually played by corporate gover-nance mechanisms and to explore whetherfamily councils may act as partial substitutes forthese mechanisms in family SMEs as well. Wemaintain that substitution occurs when a gov-ernance role (like ownership) is totally or par-tially covered by governance mechanisms thatare not the correspondingly entitled ones (forownership, by the family council and not by theshareholders’ meeting). Substitution is total if agovernance role is not played at all by itsentitled mechanism but by others; it is partial ifa governance role is played partly by itsentitled mechanism and partly by others. In ourcase we assume that in the presence of a familycouncil, the corporate governance control rolediminishes and that the substitution of corpo-rate governance mechanisms by the familycouncil can only be partial, as the familycouncil has no legal standing and so some formof agency-based hierarchy must always be inplace, at least to ratify the family council’sdecisions.

Hypotheses about Substitution Effectsbetween Family Councils and CorporateGovernance Mechanisms

Based on our literature analysis, we proposethree hypotheses about whether family coun-cils may partially substitute the shareholders’meeting, the BOD, and the CEO in performingtheir respective corporate governance controlroles.

As regards the possible substitution of theshareholders’ meeting by the family council,the shareholders’ meeting is sometimes consid-ered as an owners’ council acting as a subset of


the broader family council. At other times, it ishighlighted as an ownership control mecha-nism that may either formally ratify some deci-sions taken by the family council or in any caseconsult the family council before taking its gov-ernance decisions, that is, decisions aboutchoosing, controlling, and evaluating the boardand its performance (Charkham and Simpson1999; Lansberg 1999; Lank and Ward 2000). Infamily SMEs, ownership is most often closelyheld and concentrated within the family(Gersick et al. 1997; Ward 2004), so ownershipissues are closely and almost “naturally” relatedto family governance ones: for instance, boardand company performance issues are directlyconnected to family wealth ones; new boardmembers’ appointment and control issues areoften intertwined with the education of thenext generation to govern. Moreover, SMEs’entrepreneurs hardly perceive ownership as adistinct role, as they “mix it up” with manage-ment; when they have to play the role in share-holders’ meetings, they often feel “outside” thefirm (Melin and Nordqvist 2000). In thiscontext, following stewardship theory, a familycouncil may be a better place for all the familyshareholders to take corporate governancedecisions, leading to partial substitution asfollows:

H1a: When a family council is present, the cor-porate governance role of ownership playedby the shareholders’ meeting diminishes.

Substitution between the family council and theBOD might occur as, due both to firm size andto family ownership concentration and involve-ment, corporate governance mechanisms oftenoverlap in family SMEs (Johnson and Scholes2002). In particular, the shareholders’ meetingtends to feature the same composition as theBOD (Corbetta and Montemerlo 1999; Famaand Jensen 1983; Gersick et al. 1997; Melinand Nordqvist 2000; Ward 1991). Thus, thecorporate governance roles of both ownershipand monitoring might be played at the sametable, and the family might not only control asthe shareholder, but also direct the firm byapproving the strategy and plans, choosingand evaluating the CEO, and controlling thewhole executive team’s behavior (Corbettaand Tomaselli 1996; Huse 2000; Melin andNordqvist 2000; Ward 1991). Moreover, when aBOD copes with firm complexity and/or own-ership evolution, entrepreneurs may feel

outside their normal business activity and,again, perceive the board as closer to the familythan to the firm. According to stewardship,family councils could then be preferred byfamily directors to BODs—also because boardsare subject to a number of legal obligations aswell as shareholders’ meetings—for taking cor-porate governance decisions in a less formalcontext that facilitates information exchangeand discussion. Thus follows:

H1b: When a family council is present, the cor-porate governance role of monitoring playedby the BOD diminishes.

Finally, family owners–directors are often soinvolved in the business that they share asco-CEOs the leading role of choosing andevaluating managers, formulating strategicplans, and defining and controlling budgets(Melin and Nordqvist 2000; Minichilli, Corbetta,and MacMillan 2010; Sharma 2004). Again, forstewardship-based reasons, the family councilmight be used for taking CEO-level governancedecisions as well, leading to:

H1c: When a family council is present, the cor-porate governance role of leading thecompany played by the CEO diminishes.

MethodsResearch Design and Definitions of“Family Firm” and “SME”

To test our hypotheses, we investigated thegovernance roles played by various governancemechanisms of family SMEs with family councilscompared with family SMEs without familycouncils. We used a quantitative deductivedesign, and the hypotheses were tested throughquestionnaires from a sample of 243 Italianfamily SMEs. The respondents were in mostcases the CEOs, and 94.3 percent of thembelonged to the owning family. We used abroad, control-based definition of family firms—that is, more than 50 percent of the shares had tobelong to the owning family (Astrachan andShanker 2003; Corbetta 1995; Sharma 2004)—toinvestigate the largest variety of ownershipstructures, and thereby of governance systems.Analogous criteria were applied to define sizeranges: we used a definition of SMEs broaderthan that of the European Union to increase thelikelihood of embracing a varied range of own-ership structures and governance systems and ofobtaining a better response rate. So, based on


the North American parameters of Canada andthe United States (U.S. Small Business Adminis-tration and Canadian Ministry of Industry), weassumed small firms to be companies with fewerthan 250 employees and turnover of 50 millioneuros and medium-sized ones to be thoseemploying 251 to 500 employees and turnovertotaling 50 to 250 million euros.

SampleItaly features a very relevant presence of

family SMEs, which represent around 84 percentof incorporated companies (Gnan andMontemerlo 2008). By civil law, Italian busi-nesses can take either the incorporated or theunlimited form. The first one includes joint stockcompanies and companies with limited liability,whereas the second one implies unlimited liabil-ity of the owners. We conducted our survey bysending a questionnaire to Italian incorporatedSMEs, as these are more likely to feature articu-lated governance systems, and they represent asignificant player in the Italian economy (49.1percent of the total Italian employment; IstitutoNazionale di Statistica 2004).

The empirical base for the study was drawnfrom the AIDA database (Bureau Van Dijk Elec-tronic Publishing), containing data on manufac-turing and nonmanufacturing incorporatedSMEs, representative of the Italian population. Atotal of 15,157 firms were randomly extracted tobe representative of the population by region,range of employees, and industries. The ques-tionnaire was mailed in October 2000; it wasaddressed to top executives, and particularly toeither the CEO or an equivalent senior execu-tive, that is, to knowledgeable people who musthave an in-depth understanding of the gover-nance system and whose answers have proved

to be reliable, even if reporting self-perceptions,in anonymous surveys (Dillman 1978; Starbuckand Milliken 1988). Two follow-up letters andone replacement questionnaire were mailedafter the initial mailing. The questionnaire con-sisted of six sections on firms’ demographics,ownership, governance systems, strategy, per-formance, and succession.

The responses were collected throughoutJanuary 2001. The final data set includes 480SMEs; the response rate was 3.2 percent, whichis low in absolute terms, but in line with thosenormally obtained in Italy (Corbetta andMontemerlo 1999; Giacomelli and Trento2005). We conducted two tests that reassuredus regarding the absence of sample biases.First, we evaluated nonresponse biases withchi-square tests between our sample and themailing list, finding no significant differencesfor industries (χ2 = 0.025835, df = 5), turn-over (χ2 = 0.047543, df = 4), or employees (χ2 =0.073285, df = 3). Moreover, across industries,classes of turnover, and classes of employees,the response rates vary, but the proportions ofrespondents are not statistically different (Fish-er’s exact test) in the sample and in the mailinglist (industries vs. turnover, p = .0234; indus-tries vs. employees, p = .3751; turnover vs.employees, p = .03228). Second, we comparedthe early respondents (first half) with the laterespondents (second half) (Armstrong andOverton 1977), finding no significant differ-ences in industries, size (employees and turn-over), firm age, or market characteristics.

Among the 480 firms responding to thesurvey, the median number of employees was22 and the mean number was 58 with a stan-dard deviation of 89 (see Tables 1 and 2). Themedian turnover was €3 million. Almost half of

Table 1Sector and Size of Firms in the Mailing List and the Sample

Mailing List Sample ResponseRate

(Percentage)Number Percentage Number Percentage

Manufacturing Small 6,048 39.9 197 41.0 3.3Medium-Sized 887 5.9 33 6.9 3.7

Nonmanufacturing Small 6181 40.8 190 39.6 3.1Medium-Sized 2041 13.5 60 12.5 2.9

Total 15157 100.0 480 100.0 3.2


the sample was composed of manufacturingfirms. Out of the 480 sample firms in total, 398firms (82.9 percent) were classified as familySMEs.

We reduced the number of family SMEs from398 to 243 firms by identifying the sampledfamily firms with a BOD, based on whether therespondents declared that the firm had a boardor not. In fact, having a BOD is not mandatoryin Italy, as owners of incorporated firms canappoint a sole CEO instead of a board (whereashaving a shareholders’ meeting is compelledby law). It is interesting to notice that thesubsamples of firms with a BOD and firmswithout a BOD feature no significant differ-ences, as reported in Table 3.

Our analysis proceeded by dividing the 243sampled firms with a BOD into two groups: (1)family SMEs with a BOD and a family council(108 cases); (2) family SMEs with a BOD but nofamily council (135 cases). Descriptive statisticsof both groups are presented in Table 4.

The existence of a family council was alsobased on whether the respondents declaredthat the firm had a family council or not. Fol-lowing the definition of a family council, weasked the respondents to acknowledge thepresence of a family council in the case thatregular gatherings took place involving adultfamily members and were recognized as afamily council.

Firms with a family council represent 27.1percent of the total sample of family SMEs (108out of 398 cases) and 44.4 percent of family firmswith a BOD (108 out of 243 cases). In Italy, no

other empirical data exist about family councilsin SMEs, but qualitative empirical referencessuggest that these percentages are reasonable.Moreover, similar data are reported by a surveyon Spanish family firms, featuring 25 percentof cases with a family council (Suàre andSantana-Martìn 2004), regardless of whether aBOD is present or not; given the similaritiesbetween the family business systems in the twocountries, we were reassured about the familycouncil incidence we found.

Firms with family councils are on averageolder than those without family councils(p = .1) and have more family shareholders(p = .01). No other significant differences existbetween the two groups of firms with respectto the number of employees, generationinvolved, number of shareholders, percentageof family ownership, percentage of shares heldby the largest shareholder, number of boardmembers, number of family board members,number of shareholders’ meetings, or numberof board meetings. Particularly, the meannumber of board members in family firms withand without a family council is 3.8, which iscomparable with figures on the number ofboard members in other countries’ SMEs(Fiegener, Brown, and Derux 2000; Gabrielssonand Winlund 2000).

Variables and MeasuresDependent Variables. The three dependentvariables are the ownership role of the share-holders’ meeting, the monitoring role of theBOD, and the leading role of the CEO. We chose

Table 2Number of Employees of Firms in the Mailing List and the Sample

Number ofEmployees

Mailing List Sample

Number Percentage Number Percentage

From 1 to 10 6,513 43.0 211 44.0From 11 to 20 3,261 21.5 103 21.5From 21 to 50 3,234 21.3 102 21.3From 51 to 100 1,194 7.9 35 7.3From 101 to 150 391 2.6 11 2.3From 151 to 200 200 1.3 6 1.3From 201 to 350 266 1.8 8 1.7From 351 to 500 99 0.7 4 0.8Total 15,157 100.0 480 100.0


Tab

le3

Char

acte

rist

ics

of

Sam

ple

dFam

ily

Fir

ms

wit

hout

and

wit

ha

Fam

ily

Counci

l

Wit

hout

Fam

ily

Counci

l(1

35

case

s)W

ith

Fam

ily

Counci

l(1

08

case

s)

Mea

nM

edia

nSta

ndar

dD

evia

tion

Min

imum

Max

imum

Mea

nM

edia

nSta

ndar

dD

evia

tion

Min

imum

Max

imum

Yea

rof

Foundat

ion

1970

1965

22.2

819

0019

9719

67*

1959

32.1

518

0719

95N

o.

of

Em

plo

yees

27.0

074

.010

2.04

1.00

500.

0030

.00

67.0

98.9

52.

0049

6.00

Gen

erat

ion

Invo

lved

2.10

2.0

0.85

1.00

5.00

2.26

2.0

0.94

1.00

6.00

No.

of

Shar

ehold

ers

4.19

3.00

3.75

2.00

26.0

03.

814.

001.

522.

009.

00N

o.

of

Fam

ily

Shar

ehold

ers

2.50

3.00

3.27

1.00

25.0

03.

46**

4.00

1.50

1.00

18.0

0

Per

centa

geO

wned

by

Fam

ily

90.4

510

0.00

16.5

550

.00

100.

0097

.58

100.

008.

5354

.00

100.

00

Per

centa

geof

Shar

esH

eld

by

the

Larg

est

Shar

ehold

er

44.6

244

.00

23.6

95.

0099

.60

41.0

439

.00

18.8

810

.00

91.0

0

No.

of

Boar

dM

ember

s3.

783.

001.

572.

0011

.00

3.80

4.00

1.21

2.00

8.00

No.

of

Fam

ily

Boar

dM

ember

s2.

913.

001.

331.

008.

003.

503.

001.

131.

007.

00

No.

of

Shar

ehold

ers

Mee

tings

1.88

2.00

1.09

1.00

6.00

2.76

2.00

2.82

1.00

12.0

0

No.

of

Boar

dM

eetings

2.98

3.00

2.03

1.00

12.0

03.

262.

002.

991.

0012

.00

No.

of

Fam

ily

Counci

lM

eetings

0.00

0.00

0.00

0.00

0.00

12.7

16.

0027

.54

1.00

18.0

0

*Tw

o-tai

led

p<

.1.

**Tw

o-tai

led

p<

.001

.


Tab

le4

Char

acte

rist

ics

of

Sam

ple

dFam

ily

Fir

ms

wit

hout

and

wit

ha

Boar

dof

Dir

ecto

rs

Wit

hout

aB

oar

d(1

55

case

s)W

ith

aB

oar

d(2

43

case

s)

Mea

nM

edia

nSta

ndar

dD

evia

tion

Min

imum

Max

imum

Mea

nM

edia

nSta

ndar

dD

evia

tion

Min

imum

Max

imum

Yea

rof

Foundat

ion

1980

1974

18.9

618

7419

9519

6919

6227

.23

1807

1997

No.

of

Em

plo

yees

22.3

042

.173

.93

1.00

479.

0027

.50

70.9

100.

521.

0050

0.00

Gen

erat

ion

Invo

lved

1.82

2.0

0.83

1.00

6.00

2.17

2.0

0.90

1.00

6.00

No.

of

Shar

ehold

ers

2.87

3.00

1.30

2.00

17.0

04.

023.

002.

972.

0026

.00

No.

of

Fam

ily

Shar

ehold

ers

2.54

2.00

1.22

1.00

15.0

03.

553.

002.

631.

0025

.00

Per

centa

geO

wned

by

Fam

ily

93.4

010

0.00

14.4

550

.00

100.

0093

.62

100.

0014

.01

50.0

010

0.00

Per

centa

geof

Shar

esH

eld

by

the

Larg

est

Shar

ehold

er

54.2

150

.00

29.1

51.

0010

0.00

43.1

240

.00

21.8

35.

0099

.60

No.

of

Boar

dM

ember

s0.

000.

000.

000.

000.

003.

793.

001.

422.

0011

.00

No.

of

Fam

ily

Boar

dM

ember

s0.

000.

000.

000.

000.

003.

183.

001.

271.

008.

00

No.

of

Shar

ehold

ers’

Mee

tings

2.71

1.50

3.52

1.00

12.0

02.

362.

002.

241.

0012

.00

No.

of

Boar

dM

eetings

0.00

0.00

0.00

0.00

0.00

3.10

3.00

2.48

1.00

12.0

0

No.

of

Fam

ily

Counci

lM

eetings

4.93

4.50

30.5

81.

0022

.00

5.12

5.00

18.5

11.

0018

.00


role coverage as the measure of our variables inorder to explore how much each governancemechanism plays its distinctive role, indepen-dently of both the process by which the role isplayed (for example, in terms of frequency) andthe relative importance of the tasks that areincluded in the role itself. The variables werebuilt up through the following steps. First, foreach of the three governance roles, we catego-rized the corresponding governance tasks basedon the literature analysis presented in section 1.Second, using dummy coded questions in an adhoc section of the questionnaire, the respon-dents were asked to indicate whether or noteach governance mechanism (the shareholders’meeting, the BOD, and the CEO) performedall the tasks composing the governance role.Every performed task was coded as 1 and everyunperformed task was coded as 0. Third, wecalculated every role’s coverage as the mean ofthe number of performed tasks out of the totaltasks composing every governance mecha-nism’s role. In particular:

• Regarding the ownership role of the share-holders’ meeting, two items were related tothis variable based on the ownership role’stasks: Q1 on choosing board members; Q2on controlling and evaluating the board andits performances (Charkham and Simpson1999). The variable is calculated as follows:the mean of Q1’s and Q2’s performedtasks = (Q1 + Q2)/2 (Cronbach’s α = 0.74).

• Regarding the monitoring role of the BOD,four items were related to this variablebased on the monitoring role’s tasks: Q3 ontaking strategy decisions; Q4 on approvingstrategic plans; Q5 on choosing and evalu-ating the CEO; Q6 on controlling the CEO

and managers (Charkham and Simpson1999; Hillman, Cannella, and Paetzold 2000;Hillman and Dalziel 2003; Johnson, Daily,and Ellstrand 1996; McNulty and Pettigrew1999; Nicholson and Kiel 2004). The vari-able is calculated as follows: the mean ofQ3’s, Q4’s, Q5’s, and Q6’s performedtasks = (Q3 + Q4 + Q5 + Q6)/4 (Cronbach’sα = 0.88).

• Regarding the leading role of the CEO, threeitems were related to this variable based onthe leading role’s tasks: Q7 on formulatingstrategic plans; Q8 on defining and control-ling annual budgets; Q9 on choosing andevaluating the management (Fayol 1949;Finkelstein and Hambrick 1996; Minichilli,Corbetta, and MacMillan 2010). The variableis calculated as follows: the mean of Q7’s,Q8’s, and Q9’s performed tasks = (Q7 +Q8 + Q9)/3 (Cronbach’s α = 0.93).1

All the measures of role coverage haveCronbach’s αs above the 0.6 value. The mea-sures were also evaluated in terms of individualitem reliability, internal consistency, and dis-criminant validity (see Appendix 1). Thepsychometric properties of the multi-item mea-sures (ownership role, monitoring role, andleading role) were assessed simultaneously inone confirmatory factor analysis (CFA)2 usingLISREL 8.53. We interpreted the goodness of fitusing the comparative fit index (CFI), thenormed fit index (NFI), and the standardizedroot-mean-square residual (SRMR) (see Hairet al. 1998; Hu and Bentler 1999). Additionally,we used the commonly accepted cutoff values(CFI < 0.90, NFI < 0.90, and SRMR < 0.06) asindicative of poor fit (for example, Hair et al.1998; Zhao et al. 2007). The model presents a

1For completeness’s sake, we also measured the family governance role, that is, how the family council coversits governance role. Two items were related to this variable based on family governance tasks: Q10 ondesigning and managing family–firm relationships and Q11 on planning generation transition (Carlock andWard 2001; Gallo and Kenyon 2004; Lansberg 1999; Mustakallio and Autio 2001). The variable was calculatedas follows: mean of Q10’s and Q11’s performed tasks = (Q10 + Q11)/2 (α = 0.91). In addition, we measuredhow each of the three corporate governance roles and the family governance role were covered by all theother, noncorresponding mechanisms (for example the monitoring role by shareholders’ meetings and theCEO). Actually, multiple responses were allowed, so that the same task could be attributed to differentgovernance mechanisms. In other words, each respondent was asked each of the eleven questions (Q1–Q11)four times, that is, once for each of the four family and corporate governance mechanisms considered (familycouncil, shareholders’ meeting, BOD, and CEO).2We also proceeded to test the CFA. No signs of problems (for example nonconvergence, nonpositive definitematrices, unreasonable standard errors, and so on) emerged. The model is empirically identified as shown bya converged solution, by the absence of any out-of-bounds or unexpected parameter estimates and by theabsence of any warning or error messages.


root-mean-square error of approximation(RMSEA) value of 0.057 and an SMSR of 0.032.The goodness-of-fit index values are as follows:χ2 = 64.31, df = 24; CFI = 0.78, NFI = 0.76. Weassessed the reliability by calculating a compos-ite reliability for each construct (Fornell andLarcker 1981). Along with the reliability calcu-lations, we also examined the parameter esti-mates and their associated t-values as well asthe average variances extracted (Anderson andGerbing 1988). The factor loadings range from0.45 to 0.88 (p < .05), and the average variancesextracted range from 73 to 85 percent. Theitems were also found to be reliable and validwhen evaluated based on each item’s errorvariance, modification index, and residualcovariance. We established discriminant valid-ity by calculating the shared variance betweeneach pair of constructs and verified that it waslower than the variances extracted for theinvolved constructs (Fornell and Larcker 1981).The shared variances between the pairs ofall possible scale combinations indicated thatthe average variances extracted were higherthan the associated shared variances in all thecases.

Independent Variable. The independent vari-able was the presence of a family council in thefamily firm (FAMILY COUNCIL). The existenceof a family council was coded as 1; nonexis-tence was coded as 0.

Control Variables. In order to account forpotential biases, several control variables wereadopted in the analysis:

• The year of foundation, as family firmsmight confuse governance roles and mecha-nisms and make substitutions because theyare younger, that is, they have little gover-nance experience and few governanceneeds; conversely, older family firms mightbe more familiar with both family councilsand the agency-based hierarchy as well ashaving a greater need to respect their dis-tinctive roles (Lank and Ward 2000; Ward1987)

• The number of employees, as smaller familyfirms—young or not—might face little com-plexity in the family and company and somight devote less attention to governanceand especially to the agency-based hierar-chy (Fama and Jensen 1983; Melin andNordqvist 2000)

• The industry, as a proxy for complexitysuch as the number of employees (Famaand Jensen 1983)

• The generation involved, as a proxy forgovernance experience such as the year offoundation (Gersick et al. 1997; Lansberg1999; Ward 2004)

• The number of shareholders, also as aproxy for experience such as the year offoundation and the generation involved. Infact, the number of shareholders tends toincrease over time and generations (Gersicket al. 1997; Lansberg 1999; Ward 2004)

• The number of family shareholders, forthe same reasons as the number ofshareholders

• The number of board members, as too highor too small a number might make it diffi-cult for the board to work properly andthereby encourage substitution (Corbettaand Salvato 2004; Corbetta and Tomaselli1996)

• The percentage of family ownership, as thehigher the percentage of shares, the greaterthe power of the family and the morediscretion the family might apply in using(and substituting) governance mechanisms(Miller and Le Breton-Miller 2006)

• The number of family board members, as aproxy for family power like the percentageof family ownership

Analyses and ResultsWe tested the hypotheses by comparing the

governance roles as covered by various gover-nance mechanisms in the two groups of familyfirms composed, respectively, of those withoutfamily councils and those with family councils.

We expected that if a substitution effect ispresent, all things being equal, the corporategovernance roles’ coverage by their corre-spondingly entitled mechanisms (for examplethe ownership role’s coverage by the share-holders’ meeting) would diminish when thefamily council exists. The univariate test resultsare presented in Table 5, which shows theextent to which each corporate governancemechanism plays its role. That is, for eachcombination of corporate governance mecha-nism and governance role, the mean percent-ages of role coverage featured in group 1(family firms without family councils) and ingroup 2 (family firms with family councils) arecompared, and the significance levels of thedifferences are reported. In addition, the table


reports the percentage of corporate governanceroles covered by the family council.

Hypothesis 1aThe corporate governance role of ownership

played by the shareholders’ meeting is signifi-cantly diminished from coverage of 36.9percent in group 1 to almost half, that is, 18.9percent, in group 2, in which the family councilis present.

We also looked at the ownership role’s cov-erage by the family council (53.9 percent) ingroup 2; this shows that the family council notonly replaces the shareholders’ meeting in per-forming the role of ownership, but is also muchmore involved in this role than the sharehold-ers’ meeting itself. The other corporate gover-nance mechanisms partly cover the ownershiprole as well, with percentages that diminishfrom group 1 to group 2. In particular, BODs’coverage of the ownership role passes from12.7 percent (when there is no family council)to 7.0 percent (when the family council ispresent) and the CEO’s coverage of the owner-ship role passes from 14.5 percent (with afamily council) to 9.8 percent (without a familycouncil). Thus, Hypothesis 1a is supported.

Hypothesis 1bThe monitoring role’s coverage by the BOD

is significantly diminished when the familycouncil appears, from coverage of 19.3 percentin group 1 to coverage of 8.8 percent ingroup 2.

Besides, the research findings show that thefamily council is the biggest performer of themonitoring role, as it covers it at a rate of 42.7percent in group 2. As we did for ownership,we also measured whether other corporategovernance mechanisms cover the monitoringrole, and we found that in this case such cov-erage also occurs with decreasing percentagesfrom group 1 to group 2. In particular, themonitoring role’s coverage by the shareholders’meeting passes from 16.7 to 10.1 percent andthe monitoring role’s coverage by the CEOpasses from 26.8 to 23.6 percent in group 1 andgroup 2, respectively. So, Hypothesis 1b is sup-ported as well.

Hypothesis 1cThis hypothesis is not supported. The

leading role’s coverage by the CEO actuallydiminishes from 32.8 percent in group 1 to 23.8percent in group 2, but the difference betweenthe two percentages is not significant.

Tab

le5

Gove

rnan

ceM

echan

ism

san

dG

ove

rnan

ceR

ole

sin

Fam

ily

Fir

ms

wit

hout

Fam

ily

Counci

ls(G

roup

1)

and

Fam

ily

Fir

ms

wit

hFam

ily

Counci

ls(G

roup

2)

(Mea

nP

erce

nta

ge)

Corp

ora

teG

ove

rnan

ceR

ole

sShar

ehold

ers’

Mee

ting

Boar

dof

Dir

ecto

rsC

EO

Fam

ily

Counci

l

Gro

up

1G

roup

2G

roup

1G

roup

2G

roup

1G

roup

2G

roup

1G

roup

2

Ow

ner

ship

Role

36.9

a18

.9**

a12

.77.

0**

14.5

9.8

0.0

53.9

Monitori

ng

Role

16.7

10.1

*19

.3b

8.8*

*b26

.823

.60.

042

.7Le

adin

gRole

13.3

8.3

16.6

9.2*

*32

.8c

23.8

c0.

038

.8

*Tw

o-tai

led

p<

.1.

**Tw

o-tai

led

p<

.05.

a Fig

ure

use

dto

test

H1a

.bFi

gure

use

dto

test

H1b

.c F

igure

use

dto

test

H1c

.


It has to be noticed that the family councilpartly covers this role as well (38.8 percent),and so do the BOD and the shareholders’meeting, always in percentages that arereduced passing from group 1 to group 2. Spe-cifically, the shareholders’ meeting’s coverageof the CEO’s role passes from 13.3 to 8.3percent and the BOD’s coverage of the CEO’srole passes from 16.6 to 9.2 percent, in group 1and group 2, respectively.

We also conducted multivariate analyses ofvariance (MANOVA, with Wilks’s lambda test)to control for possible sources of heterogeneityin the two groups (see Table 6), using thecontrol variables mentioned above. TheMANOVA confirmed the univariate results,finding no impact from the control variables.Only the presence of a family council intro-duces significant differences concerning theownership role of the shareholders’ meetingand the monitoring role of the board of direc-tors (p = .000), that is, the dependent variablesof H1a and b.

Follow-Up AnalysesTo follow up the results reported so far, we

carried out further analyses of which tasks aretaken over by family councils within corporategovernance roles. These analyses, presented inTable 7, show in more detail how extensivelythe family council goes beyond its own familygovernance role.

In fact, the family council partly covers theownership role, both choosing board members(at 60.3 percent) and controlling and evaluatingthe board (at 52.4 percent). As for the monitor-ing role, the family council especially coverstwo out of the three tasks within this role:choosing the CEO (at 66.0 percent) and takingstrategic decisions (at 45.4 percent). The familycouncil also plays some of the leading role ofthe CEO by formulating strategic plans (44.7percent). It is noticeable that it is exactly forthese roles and tasks that their correspondingcorporate governance mechanisms mostlydiminish their coverage in the presence of thefamily council. That is, when the family councilexists, the shareholders’ meeting reduces bothits ownership tasks, falling from coverage of43.4 percent to coverage of 24.1 for choosingboard members and from 39.7 percent to 16.5percent for controlling and evaluating them.The board mostly diminishes its coverage oftaking strategic decisions (from 37.2 to 12.0percent) and choosing and evaluating the CEO

(from 15.5 to 6.1 percent). The CEO also losessome of his/her roles, especially in formulatingstrategic plans (from 31.2 to 7.9 percent). Theseresults reinforce H1a and 1b, both verified, andgive some support to H1c, although it was notverified.

Robustness ChecksIn order to check the possible bias related to

the low response rate of our sample, we con-ducted two different checks: (1) two separateMANOVA analyses in two different subsamples,(a) manufacturing family firms and (b)nonmanufacturing family firms; and (2) twoseparate MANOVA analyses in two differentsubsamples, (1) family firms in which the mostimportant shareholder owns more than 50percent and (2) family firms in which the mostimportant family shareholder owns 50 percentof equity or less.

These four MANOVA analyses confirm thehypothesis tests, and they do not significantlydiffer in terms of results (see Tables 8 and 9),except for the following:

• The year of foundation and the number ofemployees introduce a significant differ-ence concerning the monitoring role ofBODs in manufacturing family firms. Thatis, the monitoring role of the board does notonly depend on the presence of a council;in particular, the larger the number ofemployees and the younger the company,the more active and focused we may expectthe BOD to be. Size might make it neces-sary to have a BOD that adequately copeswith complexity; a young age might help insetting up such a board without the con-straints of a long family tradition of gover-nance undistinguished from management;

• The generation involved and the number offamily shareholders introduce a significantdifference concerning the ownership role ofshareholders’ meetings in family firms inwhich no shareholder owns more than 50percent of equity. Again, the ownership roleof the shareholders’ meeting is not onlyinfluenced by the presence of the familycouncil; in particular, the later the genera-tion and the more numerous the familyowners, the less “substituted” we mayexpect the shareholders’ meeting to be byother mechanisms. This might be explainedby the need of the shareholders’ meeting tobe accountable enough to represent the


Tab

le6

MA

NO

VA

Res

ult

sfo

rC

om

par

ison

of

Fam

ily

Fir

ms

wit

hout

Fam

ily

Counci

ls(G

roup

1)

and

Fam

ily

Fir

ms

wit

hFam

ily

Counci

ls(G

roup

2)

Wilks’

sLam

bda

Fp

Ow

ner

ship

Role

of

the

Shar

ehold

ers’

Mee

ting

(H1a)

Monit

ori

ng

Role

of

the

Boar

dof

Dir

ecto

rs(H

1b)

Lea

din

gR

ole

of

the

CEO

(H1c)

Typ

eII

ISum

of

Squar

es

Fp

Typ

eII

ISum

of

Squar

es

Fp

Typ

eII

ISum

of

Squar

es

Fp

Corr

ecte

dM

odel

2.48

72.

171

0.02

2**

1.74

22.

091

0.02

7**

2.56

41.

425

0.17

2In

terc

ept

0.96

72.

013

0.11

40.

013

0.11

20.

738

0.02

70.

321

0.57

20.

926

5.14

80.

024*

*Fa

mily

Counci

l0.

901

6.41

90.

000*

*1.

221

10.6

610.

001*

*0.

477

5.72

40.

018*

*0.

374

2.07

80.

151

Yea

rof

Foundat

ion

0.96

82.

010

0.11

50.

000

0.00

00.

991

0.02

90.

345

0.55

80.

456

2.53

40.

124

No.

of

Em

plo

yees

0.99

20.

499

0.68

40.

013

0.11

70.

732

0.09

91.

188

0.27

70.

014

0.07

70.

782

Man

ufa

cturi

ng

0.99

70.

183

0.90

80.

003

0.02

30.

880

0.03

00.

363

0.54

70.

018

0.10

20.

750

Gen

erat

ion

Invo

lved

0.98

80.

707

0.54

90.

083

0.72

10.

397

0.02

40.

292

0.59

00.

247

1.37

10.

243

No.

of

Shar

ehold

ers

0.99

40.

350

0.78

90.

033

0.29

20.

590

0.05

80.

698

0.40

40.

037

0.20

50.

651

No.

of

Fam

ily

Shar

ehold

ers

0.98

90.

629

0.59

70.

027

0.24

00.

625

0.12

01.

444

0.23

10.

087

0.48

20.

489

No.

of

Boar

dM

ember

s0.

974

1.58

20.

196

0.06

50.

568

0.45

20.

156

1.87

30.

178

0.02

90.

159

0.69

1Per

centa

geO

wned

by

Fam

ily

0.96

62.

012

0.11

30.

334

2.91

60.

190

0.00

00.

002

0.96

70.

273

1.51

90.

219

No.

of

Fam

ily

Boar

dM

ember

s0.

967

2.01

30.

114

0.24

92.

177

0.14

20.

234

2.80

90.

119

0.00

10.

007

0.93

3

Err

or

20.2

7514

.742

31.8

45Tota

l37

.563

20.5

5650

.500

Corr

ecte

dTota

l22

.762

16.4

8434

.410

*Tw

o-tai

led

p<

.1.

**Tw

o-tai

led

p<

.05.

***T

wo-tai

led

p<

.001

.


Tab

le7

Gove

rnan

ceM

echan

ism

s,R

ole

san

dTas

ks

inFam

ily

Fir

ms

wit

hout

Fam

ily

Counci

ls(G

roup

1)

and

Fam

ily

Fir

ms

wit

hFam

ily

Counci

ls(G

roup

2)

(Mea

nP

erce

nta

ge)

Shar

ehold

ers’

Mee

ting

Boar

dof

Dir

ecto

rsC

EO

Fam

ily

Counci

l

Gro

up

1G

roup

2G

roup

1G

roup

2G

roup

1G

roup

2G

roup

1G

roup

2

Ow

ner

ship

Role

Choosi

ng

Boar

dM

ember

s43

.4a

24.1

a16

.710

.117

.08.

30.

060

.3

Contr

ollin

gan

dEva

luat

ing

the

Boar

dan

dits

Per

form

ance

39.7

a16

.5a

8.7

10.1

12.0

8.3

0.0

52.4

Monitori

ng

Role

Tak

ing

Stra

tegy

Dec

isio

ns

12.9

8.5

37.2

b12

.0b

37.0

28.9

0.0

45.4

Appro

ving

Stra

tegi

cPla

ns

12.9

10.1

16.1

b13

.9b

36.7

26.4

0.0

39.5

Choosi

ng

and

Eva

luat

ing

the

CEO

29.2

14.0

15.5

b6.

1b19

.76.

30.

066

.0

Contr

ollin

gCEO

and

Man

ager

s11

.87.

86.

0b5.

6b13

.832

.70.

028

.6

Lead

ing

Role

Form

ula

ting

Stra

tegi

cPla

ns

14.0

8.3

20.9

11.2

31.2

c7.

9c0.

044

.7

Defi

nin

gan

dContr

ollin

gA

nnual

Budge

ts11

.17.

411

.88.

216

.0c

24.5

c0.

031

.9

Choosi

ng

and

Eva

luat

ing

Man

agem

ent

14.8

9.2

17.1

8.2

31.4

c23

.5c

0.0

41.3

Fam

ily

Gove

rnan

ceRole

Des

ignin

gan

dM

anag

ing

Fam

ily–

Firm

rela

tionsh

ips

25.8

7.8

8.7

5.9

15.6

8.7

0.0

66.8

Pla

nnin

gG

ener

atio

nTra

nsi

tion

23.4

7.8

7.1

2.9

10.2

8.7

0.0

68.8

a Fig

ure

use

dto

test

H1a

.bFi

gure

use

dto

test

H1b

.c F

igure

use

dto

test

H1c

.


Tab

le8

MA

NO

VA

Res

ult

sfo

rM

anufa

cturi

ng

Fam

ily

Fir

ms

and

Nonm

anufa

cturi

ng

Fam

ily

Fir

ms

Nonm

anufa

cturi

ng

Wilks’

sLam

bda

Fp

Ow

ner

ship

Role

of

the

Shar

ehold

ers’

Mee

ting

(H1a)

Monit

ori

ng

Role

of

the

Boar

dof

Dir

ecto

rs(H

1b)

Lea

din

gR

ole

of

the

CEO

(H1c)

Typ

eII

ISum

of

Squar

es

Fp

Typ

eII

ISum

of

Squar

es

Fp

Typ

eII

ISum

of

Squar

es

Fp

Corr

ecte

dM

odel

2.18

12.

367

0.01

8**

1.87

02.

222

0.02

9**

0.73

10.

386

0.94

9In

terc

ept

0.96

31.

262

0.29

20.

030

0.29

40.

589

0.04

70.

627

0.43

00.

505

3.05

80.

083*

Fam

ily

Counci

l0.

865

5.10

10.

003*

*0.

681

6.65

50.

011*

*0.

336

4.46

70.

053*

0.25

61.

549

0.21

7Y

ear

of

Foundat

ion

0.95

21.

638

0.18

60.

055

0.53

50.

466

0.04

50.

597

0.44

20.

646

3.90

90.

051*

No.

of

Em

plo

yees

0.98

10.

646

0.58

80.

151

1.47

20.

228

0.00

90.

118

0.73

20.

048

0.29

10.

591

Gen

erat

ion

Invo

lved

0.96

11.

334

0.26

80.

081

0.79

10.

376

0.04

80.

638

0.43

80.

016

0.09

60.

757

No.

of

Shar

ehold

ers

0.98

50.

503

0.68

10.

015

0.14

30.

707

0.05

10.

679

0.41

20.

162

0.98

00.

325

No.

of

Fam

ily

Shar

ehold

ers

0.97

90.

705

0.55

10.

011

0.10

40.

747

0.07

20.

963

0.32

90.

239

1.44

50.

232

No.

of

Boar

dM

ember

s0.

976

0.80

80.

492

0.03

30.

320

0.57

30.

159

2.11

50.

149

0.00

20.

013

0.90

8

Per

centa

geO

wned

by

Fam

ily

0.95

41.

589

0.19

70.

157

1.53

50.

218

0.01

10.

140

0.70

90.

408

2.46

90.

119

No.

of

Fam

ily

Boar

dM

ember

s0.

962

1.30

70.

277

0.23

82.

324

0.13

10.

123

1.63

30.

204

0.00

90.

055

0.81

5

Err

or

10.2

377.

522

16.5

28Tota

l20

.875

10.6

6729

.000

Corr

ecte

dTota

l12

.418

8.33

920

.264


Tab

le8

Con

tin

ued

Man

ufa

cturi

ng

Corr

ecte

dM

odel

2.06

32.

327

0.01

9**

1.83

02.

214

0.03

1**

0.63

10.

356

0.95

2In

terc

ept

1.00

00.

000

0.00

00.

000

Fam

ily

Counci

l0.

890

2.75

60.

049*

*0.

639

4.68

40.

034*

*0.

385

4.19

60.

044*

*0.

063

0.32

00.

574

Yea

rof

Foundat

ion

0.92

81.

733

0.16

90.

087

0.63

50.

428

0.29

23.

184

0.07

9*0.

160

0.81

50.

370

No.

of

Em

plo

yees

0.94

71.

248

0.29

90.

016

0.11

90.

731

0.27

63.

008

0.08

7*0.

068

0.34

70.

558

Gen

erat

ion

Invo

lved

0.95

21.

134

0.34

20.

000

0.00

00.

982

0.10

71.

161

0.28

50.

364

1.84

90.

178

No.

of

Shar

ehold

ers

0.98

20.

414

0.74

40.

101

0.73

90.

393

0.05

00.

546

0.46

30.

000

0.00

10.

981

No.

of

Fam

ily

Shar

ehold

ers

0.97

30.

631

0.59

80.

069

0.50

90.

478

0.13

81.

505

0.22

40.

007

0.03

60.

850

No.

of

Boar

dM

ember

s0.

978

0.49

20.

689

0.00

00.

003

0.95

60.

138

1.50

80.

224

0.00

40.

022

0.88

2

Per

centa

geO

wned

by

Fam

ily

0.97

70.

525

0.66

70.

187

1.36

70.

246

0.02

60.

288

0.59

30.

030

0.15

30.

697

No.

of

Fam

ily

Boar

dM

ember

s0.

969

0.70

80.

551

0.02

20.

159

0.69

10.

179

1.95

30.

167

0.01

20.

063

0.80

2

Err

or

9.41

86.

336

13.5

78Tota

l16

.938

9.88

921

.500

Corr

ecte

dTota

l10

.386

8.16

614

.209

*Tw

o-tai

led

p<

.1.

**Tw

o-tai

led

p<

.05,

***T

wo-tai

led

p<

.001

.


Tab

le9

MA

NO

VA

Res

ult

sfo

rM

anufa

cturi

ng

Fam

ily

Fir

ms

wit

ha

Lar

ges

tFam

ily

Shar

ehold

erw

ith

More

than

or

atLea

st50

Per

cent

of

the

Shar

esan

dM

anufa

cturi

ng

Fam

ily

Fir

ms

wit

ha

Lar

ges

tFam

ily

Shar

ehold

erw

ith

Les

sth

an50

Per

cent

of

the

Shar

es

Lar

ges

tFam

ily

Shar

ehold

erw

ith

More

than

or

atLea

st50

Per

cent

of

the

Shar

es

Wilks’

sLam

bda

Fp

Ow

ner

ship

Role

of

the

Shar

ehold

ers’

Mee

ting

(H1a)

Monit

ori

ng

Role

of

the

Boar

dof

Dir

ecto

rs(H

1b)

Lea

din

gR

ole

of

the

CEO

(H1c)

Typ

eII

ISum

of

Squar

es

Fp

Typ

eII

ISum

of

Squar

es

Fp

Typ

eII

ISum

of

Squar

es

Fp

Corr

ecte

dM

odel

2.83

52.

310

0.01

8**

1.93

02.

293

0.02

8**

1.33

40.

732

0.69

3In

terc

ept

0.97

30.

867

0.46

10.

027

0.21

90.

641

0.14

11.

495

0.22

40.

105

0.57

40.

451

Fam

ily

Counci

l0.

934

2.18

40.

095*

0.57

54.

686

0.03

3**

0.45

64.

838

0.03

0**

0.01

90.

103

0.74

9Y

ear

of

Foundat

ion

0.97

40.

811

0.49

10.

004

0.03

30.

856

0.13

61.

444

0.23

30.

120

0.66

10.

418

No.

of

Em

plo

yees

0.97

80.

694

0.55

80.

047

0.38

40.

537

0.02

40.

256

0.61

40.

274

1.50

20.

223

Man

ufa

cturi

ng

0.99

50.

163

0.92

10.

044

0.35

50.

553

0.01

10.

113

0.73

70.

014

0.07

60.

784

Gen

erat

ion

Invo

lved

0.98

50.

458

0.71

20.

028

0.22

60.

635

0.00

10.

006

0.93

80.

218

1.19

50.

277

No.

of

Shar

ehold

ers

0.97

50.

794

0.50

00.

153

1.24

80.

267

0.03

60.

379

0.53

90.

206

1.13

20.

290

No.

of

Fam

ily

Shar

ehold

ers

0.96

90.

986

0.40

30.

142

1.15

90.

285

0.06

70.

710

0.40

20.

292

1.60

30.

209

No.

of

Boar

dM

ember

s0.

994

0.19

00.

903

0.01

00.

080

0.77

80.

001

0.01

10.

917

0.09

70.

532

0.46

7

Per

centa

geof

Fam

ily

Ow

ner

ship

0.94

61.

734

0.16

60.

233

1.90

00.

171

0.00

30.

035

0.85

20.

166

0.91

20.

342

No.

of

Fam

ily

Boar

dM

ember

s0.

989

0.35

60.

785

0.01

00.

079

0.77

90.

005

0.05

20.

820

0.17

40.

955

0.33

1

Err

or

11.5

348.

854

17.1

28Tota

l22

.500

12.1

1126

.750

Corr

ecte

dTota

l13

.050

9.77

418

.462


Tab

le9

Con

tin

ued

Lar

ges

tFam

ily

Shar

ehold

erw

ith

Les

sth

an50

Per

cent

of

the

Shar

es

Corr

ecte

dM

odel

2.47

62.

018

0.02

0**

1.86

72.

219

0.02

9**

3.49

21.

912

0.06

1*In

terc

ept

0.92

41.

552

0.21

10.

091

0.78

30.

380

0.03

30.

437

0.51

10.

742

4.06

00.

048*

*Fa

mily

Counci

l0.

849

3.37

30.

024*

*0.

587

5.03

60.

029*

*0.

325

4.30

40.

042*

*0.

094

0.53

20.

458

Yea

rof

Foundat

ion

0.91

31.

805

0.15

70.

122

1.04

70.

310

0.03

30.

434

0.51

30.

845

4.62

70.

036*

*N

o.

of

Em

plo

yees

0.90

12.

091

0.11

10.

145

1.24

60.

269

.017

0.22

60.

636

0.09

90.

542

0.46

4M

anufa

cturi

ng

0.96

70.

653

0.58

40.

129

1.10

40.

298

0.01

50.

197

0.65

90.

091

0.49

80.

483

Gen

erat

ion

Invo

lved

.940

1.20

70.

315

0.37

73.

234

0.07

7*0.

068

0.89

90.

347

0.00

50.

026

0.87

2N

o.

of

Shar

ehold

ers

0.95

10.

974

0.41

10.

317

2.72

10.

104

0.00

60.

076

0.78

40.

066

0.36

10.

550

No.

of

Fam

ily

Shar

ehold

ers

0.94

21.

180

0.32

60.

383

3.28

60.

075*

0.01

00.

133

0.71

70.

002

0.01

10.

918

No.

of

Boar

dM

ember

s0.

917

1.71

80.

173

0.24

02.

057

0.15

70.

103

1.36

50.

247

0.19

41.

063

0.30

7

Per

centa

geO

wned

by

Fam

ily

0.98

50.

290

0.83

20.

029

0.24

80.

621

0.00

00.

005

0.94

60.

122

0.66

60.

418

No.

of

Fam

ily

Boar

dM

ember

s0.

927

1.49

00.

227

0.26

62.

280

0.13

60.

080

1.06

30.

307

0.10

80.

590

0.44

6

Err

or

6.87

34.

450

10.7

79Tota

l14

.375

7.22

222

.500

Corr

ecte

dTota

l8.

661

5.69

714

.271

*Tw

o-tai

led

p<

.1.

**Tw

o-tai

led

p<

.05.

***T

wo-tai

led

p<

.001

.


various interests of an extended group thatmight be composed of family owners fromdifferent generations and with differentpositions (managing and nonmanaging) inthe family business.

Discussion and ConclusionsContributions and Implications

Our study reveals that family councils doexist in family SMEs and that they partly sub-stitute corporate governance mechanisms, andparticularly the shareholders’ meeting and theBOD.

From the empirical point of view, this studyis one of the very first to depict corporategovernance systems and family councils infamily SMEs. On one hand, we found that afamily council exists in about 27 percent of thecases in our sample, which shows how thisgovernance mechanism is not typical of onlylarge family firms. Moreover, the family councilnot only features high coverage of its ownfamily governance role, but it also covers alarge part of the ownership and monitoringroles, to the extent of being the governancemechanism with the highest coverage of theseroles. It even covers—although to a lesserextent—the leading role of the CEO. On top ofthat, we expected all this to be stronger thehigher the percentage of family ownership;instead, we found that the percentage of familyownership does not make any difference. Thus,family councils are actually both family andcorporate governance mechanisms that cutacross the agency-based hierarchy (which isreflected in the numerous meetings they holdper year with respect to other mechanisms; seeTable 4) whatever the percentage of familyownership. On the other hand, the presence offamily councils and the related substitutioneffect characterize only a minority percentageof family SMEs, which highlights how familySMEs are articulated and varied in terms ofgovernance systems. We assume that the pres-ence of family councils and the related substi-tution effect might take place more often wheneither the same or most family members areboth owners and directors. In these situations,the use of family councils could be especiallyappreciated, as decisions might be taken moreefficiently using one mechanism only, in afamily atmosphere and with no legal proce-dures to fulfill. In fact, we found that theoverlap between family membership, owner-ship, and board membership is high, but not

total. In this respect, we asked some questionsabout adult family members’ roles with respectto the company, and we found that 83.9percent of all family members are also share-holders of the family company, 68.9 percent ofall family members are also BOD members inthe family company, and 59.5 percent of allfamily members are at the same time share-holders and BOD members in the familycompany.

From a theoretical point of view, our find-ings confirm the need to match agency withstewardship in family business governancestudies (Johannisson and Huse 2000; Miller andLe Breton-Miller 2006). Agency seems to beapplied little in family SMEs, given the lowcoverage the shareholders’ meeting, the BOD,and the CEO respectively perform of theiragency-based roles; moreover, we found thatmost CEOs (94.3 percent) belong to the owningfamily, which reduces the agency threats due tothe owner–manager separation. The relationalperspective of stewardship seems to be veryhelpful in explaining this low coverage vis-à-visthe diffusion of family councils and the partialsubstitution effect that may often bring them toreplace the shareholders’ meeting and the BODin playing the ownership and monitoring roles.This does not mean that agency ceases to berelevant. Firstly, it inspires the legal setting upof corporate governance systems, whichimplies that corporate governance mechanismsmust, even if partially, work, and in the pres-ence of a family council’s substitution effectthey must ratify the family council’s decisions.Secondly, and even more importantly, agencyissues might be intensified in the future offamily SMEs, as the ownership complexity willincrease: the number of family shareholders isexpected to grow, bringing about morenonmanaging owners; at the same time, anumber of family owners’ exit processes andnonfamily owners’ entry processes will have tobe managed, and the managerialization pro-cesses might increase as well (Corbetta andMontemerlo 1999; Gnan and Montemerlo2008). Last, but not least, family businessrelational-based studies might broaden theirfocus and direct attention to family councils infamily SMEs, and not only in large familygroups.

Moving from the empirical and theoreticalcontributions to the practical implications, ourstudy’s findings offer some hints on the designand functioning of governance systems from


the family and business side. In fact, manyfamily firms seem to underutilize their corpo-rate governance mechanisms to give space toless formal family ones, like the family council,for a number of corporate governance tasks.From a stewardship perspective, this may bringa number of advantages, as mentioned above,to keep commitment and unity that are mostcritical factors for family business continuity(Gallo et al. 2001). On the contrary, disadvan-tages might occur as long as the overlapbetween family members, family owners, andfamily directors is reduced over time, whichnormally happens from generation to genera-tion, and a clearer distinction of family andcorporate governance roles and mechanisms isthen required (Gallo and Kenyon 2004; Lankand Ward 2000); our study shows that theoverlap of family members’ roles is already apartial one. In summation, owning families mayrun the risk of inertia, that is, they might keepthe ownership and monitoring roles coveredinsufficiently to cope with the intensification ofagency problems that might occur in the futureof family SMEs (Corbetta and Montemerlo 2003;Gnan and Montemerlo 2006). Besides, substi-tution is always partial, which might also gen-erate some duplication of governance roles,with consequent inefficiencies. From this per-spective, owning families and the professionalswho assist them might want to check whetherthe current design and functioning of theirshareholders’ meetings and boards of directorsare adequately active and focused to addressthe present, but above all the future, complex-ity of family and business. This does not meanthat family councils need to lose relevance, butas long as complexity increases they will needto be used as complements to corporate gover-nance mechanisms—and so to concentrate onfamily governance, which is the fundamentalbasis on which good corporate governance liesin family firms—rather than as substitutes(Corbetta and Montemerlo 2003; Ward 2004).Consistently with what has been recommendedsince the early phases of the family businessliterature (Ward 1987), other studies suggestthat complementarity can generate importantsynergies between different governance mecha-nisms and thereby generate far more positiveeffects than substitution (Poppo and Zenger2002). In our case, good family governancewould prepare good owners, directors, andleaders, and vice versa; good corporate gover-nance would keep the family satisfied about the

business and motivated, united, and committedas the owner.

Limitations and Directions forFuture Research

This study stimulates various future researchdirections, also in relation to some of itslimitations.

We conducted the empirical study of Italianfamily SMEs between 2000 and 2001. In fact,the existence of family councils may varyacross different types of family businesses(Corbetta 1995), across countries and also overtime. Therefore, our findings would have beenstronger if similar results had been found inother contexts as well as more recently.

Another important limitation is that we didnot precisely measure the composition andoverlap of family and corporate governancemechanisms; we only collected information onfamily members’ multiple roles as mentionedabove, which can only be considered as aproxy. So, further developments of our studyshould study in greater depth how family coun-cils are composed with respect to shareholders’meetings and BODs, as well as the “complexitythreshold” beyond which family councilsshould be clearly differentiated from corporategovernance mechanisms and should comple-ment rather than substitute them.

Another direction for future research con-cerns the impact of family councils’ presence.Does such a presence lead companies to takedifferent decisions from companies withoutfamily councils? Does it lead to superiorperformance?

Also, we actually considered the agency-based hierarchy as the reference model forcorporate governance roles and definition oftasks. Considering that the functioning proce-dures that companies have to comply with areagency-based and also that many boards arejust held “on paper,” we used the stewardshiptheory to test whether corporate governancecontrol mechanisms may be substituted byfamily governance mechanisms like the familycouncil, but not to analyze corporate gover-nance roles based on the stewardship perspec-tive. In particular, we did not measure theservice and advice roles of the BOD, as they areanalyzed by stewardship and other relationaltheories (Corbetta and Salvato 2004). Are theservice and advice roles actually played byfamily SMEs’ boards and does the substitutioneffect concern these roles as well?


Finally, we suggested above that our studyreinforces that agency cannot be a fullyexplanatory theory. Actually, agency might alsowork in a “hidden way.” That is, the numberof meetings of a certain corporate governancemechanism might hide those additionalmeetings in which board or shareholders’meeting members gather informally. When thishappens, board or shareholders’ meetingmembers do take the decisions they are entitledto take, but to make processes more fluid andnot to have to comply with all the bureaucracyinvolved otherwise, they do not take them intheir official capacity. In this case, informalmeetings feature the same members, but typi-cally exclude the invited members such as audi-tors. Future research developments mightexplore the informal functioning of the agency-based hierarchy as well and its impact on thesubstitution effect involving family councils.

ReferencesAnderson, J. C., and D. W. Gerbing (1988).

“Structural Equation Modelling in Practice:A Review and Recommended Two-StepApproach,” Psychological Bulletin 103, 411–423.

Anderson, R. C., and D. M. Reeb (2003).“Founding-Family Ownership and Firm Per-formance: Evidence from the S&P 500,” TheJournal of Finance 58(3), 1301–1328.

Armstrong, J. S., and T. S. Overton (1977).“Estimating Non-Response Bias in MailSurveys,” Journal of Marketing Research 14,396–402.

Aronoff, C. E., and J. L. Ward (1992). “TheCritical Value of Stewardship,” Nation’sBusiness 49, 22–32.

Astrachan, J. H., and M. C. Shanker (2003).“Family Businesses’ Contribution to the USEconomy: A Closer Look,” Family BusinessReview XVI(3), 211–221.

Braun, M., and P. Sharma (2007). “Should theCEO Also Be Chair of the Board? An Empiri-cal Examination of Family-Controlled PublicFirms,” Family Business Review XX(2), 111–126.

Bruce, N., and M. Waldman (1990). “The RottenKid Theorem Meets the SamaritanDilemma,” Quarterly Journal of Economics105, 155–165.

Brunninge, O., M. Nordqvist, and J. Wiklund(2007). “Corporate Governance and StrategicChange in SMEs: The Effects of Ownership,

Board Composition and Top ManagementTeams,” Small Business Economics 29(3),295–308.

Buchanan, J. M. (1975). “The Samaritan’sDilemma,” in Altruism, Morality and Eco-nomic Theory. Ed. E. S. Phelps. New York:Russell Sage Foundations, 71–85.

Carlock, R. S., and J. L. Ward (2001). StrategicPlanning for the Family Business. ParallelPlanning to Unify the Family and the Busi-ness. Basingstoke: Palgrave MacMillan.

Castaldi, R., and M. S. Wortman (1984). “Boardsof Directors in Small Corporations: AnUntapped Resource,” American Journal ofSmall Business IX(2), 1–10.

Charkham, J. P. (1994). Keeping GoodCompany. Oxford: Clarendon.

Charkham, J. P., and A. Simpson (1999). FairShares. Oxford: Clarendon.

Chrisman, J. J., J. H. Chua, and R. A. Litz (2004).“Comparing the Agency Costs of Family andNon-Family Firms: Conceptual Issues andExploratory Evidence,” EntrepreneurshipTheory and Practice 28(4), 335–354.

Chua, J. H., J. J. Chrisman, and P. Sharma(2003). “Succession and Nonsuccession Con-cerns of Family Firms and Agency Relation-ships with Nonfamily Managers,” FamilyBusiness Review XVI(2), 89–107.

Corbetta, G. (1995). “Patterns of Developmentof Family Businesses in Italy,” Family Busi-ness Review VIII(4), 255–265.

Corbetta, G., and D. Montemerlo (1999). “Own-ership, Governance and Management Issuesin Small and Medium Sized Family Busi-nesses: A Comparison of Italy and the UnitedStates,” Family Business Review XII(4), 361–374.

——— (2003). “Leading Family Firms: ADouble Challenge,” in New Frontiers inFamily Business Research: The LeadershipChallenge. Eds. P. Z. Poutzioris and L. Steier.Manchester and Edmonton: ManchesterBusiness NetworkSchool and Alberta Schoolof Business, 432–449.

Corbetta, G., and C. Salvato (2004). “The Boardof Directors in Family Firms: One Size FitsAll?” Family Business Review XVII(2), 119–134.

Corbetta, G., and S. Tomaselli (1996). “Boardsof Directors in Italian Family Businesses,”Family Business Review IX(4), 403–421.

Daily, C. M., and M. J. Dollinger (1993). “Alter-native Methodologies for Identifying Family-


Versus Non-Family-Managed Businesses,”Journal of Small Business Management2(31), 79–90.

Dillman, D. A. (1978). Mail and TelephoneSurveys: The Total Design Method. NewYork: John Wiley.

Donaldson, L., and J. H. Davis (1991). “Stew-ardship Theory or Agency Theory: CEO Gov-ernance and Shareholder Returns,”Australian Journal of Management 16,49–64.

Dyer, W. G. Jr (2006). “Examining the ‘FamilyEffect’ on Firm Performance,” Family Busi-ness Review XIX(3), 253–273.

Eisenhardt, K. M. (1989). “Agency Theory: AnAssessment and Review,” Academy of Man-agement Review 14, 57–74.

Fama, E. (1980). “Agency Problems and theTheory of the Firm,” Journal of PoliticalEconomy 88, 288–307.

Fama, E., and M. Jensen (1983). “Separation ofOwnership and Control,” Journal of Lawand Economics 26, 301–325.

Fayol, H. (1949). General and Industrial Man-agement. London: Pitman.

Fiegener, M., M. K. Brown, and D. R. Derux(2000). “CEO Stakes and Board Compositionin Small Private Firms,” EntrepreneurshipTheory and Practice 24, 5–25.

Fiegener, M. K. (2005). “Determinants of BoardParticipation in the Strategic Decisionsof Small Corporations,” EntrepreneurshipTheory and Practice 29(5), 627–650.

Finkelstein, S., and D. C. Hambrick (1996).Strategic Leadership. Top Executives andTheir Effects on Organizations. St. Paul, MN:West Publishing Company.

Forbes, D. P., and F. J. Milliken (1999). “Cog-nition and Corporate Governance: Under-standing Boards of Directors as StrategicDecision-Making Groups,” Academy of Man-agement Review 24, 489–505.

Fornell, C., and D. Larcker (1981). “EvaluatingStructural Equation Models with Unobserv-able Variables and Measurement Error,”Journal of Marketing Research 18, 39–50.

Gabrielsson, J., and H. Winlund (2000). “Boardof Directors in Small and Medium SizedIndustrial Firms: Examining the Effectsof the Board’s Working Style on BoardTask Performance,” Entrepreneurship andRegional Development 12, 311–330.

Gallo, M., and D. Kenyon (2004). “The Impor-tance of Family and Business Governance,”in Family Business: Key Issues. Eds. D.

Kenyon and J. L. Ward. Basingstoke, UK:Palgrave MacMillan, 102–126.

Gallo, M. A., G. Corbetta, K. Cappuyins, W. G.Dyer Jr., D. Montemerlo, and S. Tomaselli(2001). Success as a Function of Love, Trustand Freedom in Family Businesses. Barce-lona: Ediciones IESE.

Gersick, K. E., J. A. Davis, I. Lansberg, and M.McCollon Hampton (1997). Generation toGeneration. Life Cycles of the Family Busi-ness. Cambridge, MA: Harvard BusinessSchool Press.

Giacomelli, S., and S. Trento (2005). “Proprietà,controllo e trasferimenti nelle impreseitaliane. Cosa è cambiato nel decennio 1993–2003,” Working Paper No. 550. Rome: Bankof Italy.

Gnan, L., and D. Montemerlo (2006). “Family–Firm Relations in Italian SMEs: Ownershipand Governance Issues in A Double-FoldTheoretical Perspective,” in Handbook ofResearch on Family Business. Eds. P. Z.Poutziouris, K. Smyrnios, and S. B. Klein.Northampton, MA: Elgar-IFERA, 501–516.

——— (2008). Le PMI familiari in Italia tratradizione e novità. I risultati di unaricerca. Milan: EGEA.

Gomez-Mejia, L. R., M. Nunez-Nickel, and I.Gutierrez (2001). “The Role of Family Ties inAgency Contracts,” Academy of Manage-ment Journal 44, 81–95.

Gubitta, P., and M. Gianecchini (2002). “Gov-ernance and Flexibility in Family OwnedSMEs,” Family Business Review XV(4), 277–297.

Habbershon, T. G., and J. H. Astrachan (1997).“Perceptions Are Reality: How Family Meet-ings Lead to Collective Action,” Family Busi-ness Review X(1), 37–52.

Hair, J. F., R. E. Anderson, R. L. Tatham, and W.C. Black (1998). Multivariate Data Analysis.Englewood Cliffs, NJ: Prentice Hall.

Harvey, J. S. Jr (1999). “What Can the FamilyContribute to the Business? Examining Con-tractual Relationships,” Family BusinessReview XII(1), 61–71.

Hillman, A. J., and T. Dalziel (2003). “Boards ofDirectors and Firm Performance: IntegratingAgency and Resource Dependence Perspec-tives,” Academy of Management Review28(3), 383–396.

Hillman, A. J., A. A. J. Cannella, and R. L.Paetzold (2000). “The Resource DependenceRole of Corporate Directors: Strategic Adap-tation of Board Composition in Response to


Environmental Change,” Journal of Manage-ment Studies 37(2), 235–255.

Hu, L. P., and M. Bentler (1999). “Cutoff Criteriafor Fit Indexes in Covariance StructureAnalysis: Conventional Criteria Versus NewAlternatives,” Structural Equation Modeling6, 1–55.

Huse, M. (2000). “Boards of Directors in SMEs:A Review and Research Agenda,” Entrepre-neurship and Regional Development 12,271–290.

——— (2003). “Renewing Management andGovernance: New Paradigms of Gover-nance?” Journal of Management and Gover-nance 7, 211–221.

Istituto Nazionale di Statistica (2004). Strutturae dimensione delle imprese—Archiviostatistico delle imprese attive. Rome: IstitutoNazionale di Statistica, Ufficio diComunicazione.

Jensen, M. C. (1998). “Self-Interest, Altruism,Incentives, and Agency,” in Foundations ofOrganizational Strategy. Ed. M. C. Jensen.Cambridge, MA: Harvard Business SchoolPress, 48–62.

Jensen, M. C., and W. H. Meckling (1976).“Theory of the Firm: Managerial Behavior,Agency Costs and Capital Structure,” Journalof Financial Economics 3, 305–360.

Johannisson, B., and M. Huse (2000). “Recruit-ing Outside Board Members in the SmallFamily Business: An Ideological Challenge,”Entrepreneurship and Regional Develop-ment 12, 353–378.

Johnson, G., and K. Scholes (2002). ExploringCorporate Strategy, 6th ed. London: PrenticeHall.

Johnson, J. L., C. M. Daily, and A. E. Ellstrand(1996). “Boards of Directors: A Review and aResearch Agenda,” Journal of Management22, 409–438.

Jones, G. R. (1983). “Transaction Costs, Prop-erty Rights, and Organizational Culture: AnExchange Perspective,” AdministrativeScience Quarterly 28, 454–467.

LaChapelle, K., and L. B. Barnes (1998). “TheTrust Catalyst in Family-Owned Businesses,”Family Business Review XI(1), 1–17.

Lank, A. G., and J. L. Ward (2000). “Governingthe Business-Owning Family,” FBN Newslet-ter 26, 10–16.

Lansberg, I. (1999). Succeeding Generations.Boston: Harvard Business School Press.

McConaughy, D. L. (2000). “Family CEOs vs.Nonfamily CEOs in the Family-Controlled

Firm: An Examination of the Level and Sen-sitivity of Pay to Performance,” Family Busi-ness Review 13(2), 120–131.

McNulty, T., and A. Pettigrew (1999). “Strate-gists on the Board,” Organization Studies20(1), 48–74.

Melin, L., and M. Nordqvist (2000). “CorporateGovernance Processes in Family Firms. TheRole of Influential Actors and the StrategicArena,” ICSB World Conference Proceedings86–94.

——— (2007). “The Reflexive Dynamics ofInstitutionalization: The Case of the FamilyBusiness,” Strategic Organization 5(3), 321–333.

Miller, D., and I. Le Breton-Miller (2006).“Family Governance and Firm Performance:Agency, Stewardship, and Capabilities,”Family Business Review 19, 73–87.

Minichilli, A., G. Corbetta, and I. C. MacMillan(2010). “Top Management Teams in FamilyControlled Companies: ‘Familiness,’‘Faultlines,’ and Their Impact on FinancialPerformance,” Journal of ManagementStudies 47(2), 205–222.

Morck, R., A. Schleifer, and R. W. Vishny(1988). “Management Ownership andMarket Valuation: An Empirical Analysis,”Journal of Financial Economics 20, 293–315.

Mustakallio, M. A., and E. Autio (2001). “Effectsof Formal and Social Controls on StrategicDecision Making in Family Firms,” in TheRole of Family in Family Business. Eds. G.Corbetta and D. Montemerlo. Milan: EGEA,89–107.

Nicholson, G. J., and G. C. Kiel (2004). “AFramework for Diagnosing Board Effective-ness,” Corporate Governance 12(4), 442–460.

Pieper, T. M., S. B. Klein, and P. Jaskiewicz(2008). “The Impact of Goal Alignment onBoard Existence and Top Management TeamComposition: Evidence from Family-Influenced Businesses,” Journal of SmallBusiness Management 46(3), 372–394.

Poppo, L., and T. Zenger (2002). “Do FormalContracts and Relational Governance Func-tion as Substitutes or Complements?” Strate-gic Management Journal 23, 707–725.

Salvato, C. (2002). “Towards A StewardshipTheory of the Family Firm,” in The Future ofFamily Business. Values and Social Respon-sibility. Eds. M. Koiranen and N. Karlsson.Family Business Network 2002 Research


Forum Proceedings, Jyväskylä, Finland: Uni-versity of Jyväskylä, 121–145.

Schulze, W. S., M. H. Lubatkin, R. M. Dino, andA. K. Bucholtz (2001). “Agency Relation-ships in Family Firms: Theory and Evi-dence,” Organization Science 12, 99–116.

Schulze, W. S., M. H. Lubatkin, and R. M. Dino(2003). “Exploring the Agency Conse-quences of Ownership Dispersion amongPrivate Family Firms,” Academy of Manage-ment Journal 46, 179–194.

Sharma, P. (2004). “An Overview of the Field ofFamily Business Studies: Current Status andDirections for Future,” Family BusinessReview XVII(1), 1–36.

Starbuck, W. H., and F. J. Milliken (1988).“Executives’ Perceptual Filters: What TheyNotice and How They Make Sense,” in TheExecutive Effect: Concepts and Methods forStudying Top Managers. Ed. D. Hambrick.Greenwich, CT: JAI Press, 35–65.

Suàre, K. C., and D. J. Santana-Martìn (2004).“Governance in Spanish Family Business,”International Journal of EntrepreneurialBehaviour and Research 10(1/2), 141–163.

Van den Heuvel, J., A. Van Gils, and W.Voordeckers (2006). “Board Roles in Smalland Medium-Sized Family Businesses:Performance and Importance,” Corporate

Governance: An International Review 14(5),467–485.

Villalonga, B., and R. Amit (2006). “How DoFamily Ownership, Control and Manage-ment Affect Firm Value?” Journal of Finan-cial Economics 80, 385–417.

Ward, J. L. (1987). Keeping the Family BusinessHealthy. How to Plan for ContinuingGrowth, Profitability, and Family Leader-ship. San Francisco, CA: Jossey Bass.

——— (1991). Creating Effective Boards forPrivate Enterprises. San Francisco, CA:Jossey Bass.

——— (2004). Perpetuating the Family Busi-ness. 50 Lessons Learned from Long-LastingSuccessful Families in Business. New York:Palgrave MacMillan.

Zahra, S. A., and J. A. Pearce (1989). “Board ofDirectors and Corporate Financial Perfor-mance: A Review and Integrated Model,”Journal of Management 15, 291–334.

Zahra, S. A., D. O. Neubaum, and M. Huse(2000). “Entrepreneurship in Medium SizeCompanies: Exploring the Effects of Owner-ship and Governance Systems,” Journal ofManagement 26, 947–976.

Zhao, H., H. Bang, H. Wang, and P. E. Pfeifer(2007). “On the Equivalence of SomeMedical Cost Estimators with CensoredData,” Statistics in Medicine 26, 4520–4530.


Appendix 1 Measurement model results for the threegovernance rolesLatent Variables Number of

ItemsInternal

ConsistencyAverageVarianceExtracted

Correlationsbetween

Latent Variables

1 2 3

1—Ownership Role of theShareholders’ Meeting

2 0.917 0.85 0.920

2—Monitoring Role of theBoard of Directors

4 0.914 0.73 0.410 0.855

3—Leading Role of theCEO

3 0.917 0.79 0.170 −0.010 0.460

Item Measured QuestionNumber

UnstandardizedEstimates

t ErrorVariance

p ItemReliability

1—Ownership Role of theShareholders’ Meeting

Choosing BoardMembers

Q1 0.88 31.30 0.110 *** True

Controlling andEvaluating the Boardand Its Performances

Q2 0.78 32.91 0.140 *** True

2—Monitoring Role of theBOD

Taking StrategyDecisions

Q3 0.62 40.26 0.099 *** True

Approving StrategicPlans

Q4 0.59 40.65 0.100 *** True

Choosing andEvaluating the CEO

Q5 0.45 36.39 0.114 *** True

Controlling CEO,Managers

Q6 0.48 36.61 0.117 *** True

3—Leading Role of theCEO

Formulating StrategicPlans

Q7 0.54 35.04 0.071 *** True

Defining andControlling AnnualBudgets

Q8 0.56 34.94 0.070 *** True

Choosing andEvaluatingManagement

Q9 0.48 32.35 0.086 *** True

Values of the critical ratio greater than 1.64, 1.96, and 2.32 are statistically significant at the 90percent, 95 percent, and 99 percent confidence levels, respectively.*p < .1.**p < .05.***p < .01.


Documents

Governance Systems in Family SMEs: The Substitution ... · governance mechanisms and family gover-nance mechanisms. Particularly, the paper’s main objective is to explore the roles