Introduction
Entrepreneurship – starting and managing a business for the purposes of growth, income and personal satisfaction – contributes importantly to employment, economic growth and innovation in most developed and developing countries around the globe (Hisrich, Langan-Fox and Grant Reference Hisrich, Langan-Fox and Grant2007; Mead and Liedholm Reference Mead and Liedholm1998; Reynolds et al. Reference Reynolds, Bosma, Autio, Hunt, De Bono, Servais, Lopez-Garcia and Chin2005). In the context of demographic changes and the ageing of the workforce in many Western and Eastern countries (Cohen Reference Cohen2003; Shrestha Reference Shrestha2000), the role of age in entrepreneurship and small business settings has received increasing attention from researchers over the past years (Curran and Blackburn Reference Curran and Blackburn2001; de Bruin and Firkin Reference de Bruin, Firkin, de Bruin and Dupuis2003; Gielnik, Zacher and Frese in press; Kean, Van Zandt and Maupin Reference Kean, Van Zandt and Maupin1993; Lévesque and Minniti Reference Lévesque and Minniti2006; Obschonka, Silbereisen and Schmitt-Rodermund Reference Obschonka, Silbereisen and Schmitt-Rodermund2010; Rogoff Reference Rogoff2007; Schmitt-Rodermund Reference Schmitt-Rodermund2004). However, most of these studies did not consider theories and concepts from the ageing and developmental psychology literature and thus neglected potentially important theoretical considerations that follow from a lifespan perspective on entrepreneurship.
In this study, we take an interdisciplinary approach to these issues by suggesting that ageing and developmental psychology research can make a meaningful contribution towards a better understanding of the process of family succession in small and medium-sized family businesses (Ambrose Reference Ambrose1983; Davis and Harveston Reference Davis and Harveston1998; Handler Reference Handler1994; Ibrahim, Soufani and Lam Reference Ibrahim, Soufani and Lam2001; Miller, Steier and Le Breton-Miller Reference Miller, Steier and Le Breton-Miller2003). Specifically, we investigate the developmental construct of generativity as a mediator of the relationship between family business owners' age and succession in German family businesses. Based on generativity theory (Erikson Reference Erikson1950; McAdams and de St. Aubin Reference McAdams and de St. Aubin1992, Reference McAdams and de St. Aubin1998), we conceptualise family business owners' generativity in this study as family business owners' concerns with establishing and guiding members of the younger generation (e.g. their children) in the family business, while focusing less on their own career goals, occupational gains and accomplishments. Consistent with recent research in lifespan psychology, we emphasise the priority of generative goals over one's own career goals in our conceptualisation of business owners' generativity (Lang and Carstensen Reference Lang and Carstensen2002). Even though it is certainly possible to pursue both generative and personal career goals at the same time, lifespan researchers have suggested that it is goal priorities that shift with increasing age (Carstensen, Isaacowitz and Charles Reference Carstensen, Isaacowitz and Charles1999). This means that over the lifespan, generative goals gain in importance over one's own career goals and that with increasing age there is a shift in people's priority from own career goals to generative goals.
Family succession refers to the question of whether family business owners have successfully made arrangements for a family member to take over the business when they retire, or whether they have (yet) failed to do so (Sharma Reference Sharma2004). Understanding family succession is important for two reasons. First, family succession is an important success criterion in family business settings, as a majority of family business owners would like to retain family control over the business past their own tenure (Astrachan, Allen and Spinelli Reference Astrachan, Allen and Spinelli2002). Family businesses are more likely to place a high priority on long-term continuity and sustainability rather than emphasising short-term profits, which increases the importance of family succession (Kets de Vries Reference Kets de Vries1993; Miller and Le Breton-Miller Reference Miller and Le Breton-Miller2006). In addition, research emphasising a resource-based view of family firms suggests that family succession involves the transfer of tacit knowledge (Cabrera-Suárez, De Saa-Pérez and García-Almeida Reference Cabrera-Suárez, De Saa-Pérez and García-Almeida2001) and social and cultural capital (e.g. networks; Steier Reference Steier2001). These intergenerational transfers of resources, in turn, may lead to competitive advantages for family firms that are successful with regard to family succession. However, despite many family business owners' desire for cross-generational sustainability, past research indicates that many family businesses do not prepare properly for family succession (Handler Reference Handler1994; Kertesz and Atalaya Reference Kertesz and Atalaya1999). For instance, only half of family businesses manage the transition to the second generation and only about one-third make it to the third generation (Neubauer and Lank Reference Neubauer and Lank1998; Shanker and Astrachan Reference Shanker and Astrachan1996). Thus, succession planning seems to be an important issue faced by family businesses (Handler Reference Handler1994). Identifying factors that contribute to successful family succession would therefore contribute to a better understanding of important processes in family businesses and would have practical implications for entrepreneurs involved in family business settings.
Second, family succession is a major determinant of the career choices faced by family members of the next generation (Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006; Sharma and Irving Reference Sharma and Irving2005). A better understanding of family succession in small business settings could therefore help answer the question of why some people choose to pursue careers as entrepreneurs. This question is central in the entrepreneurship domain (Baron Reference Baron2004). So far, entrepreneurship researchers investigating this question have paid only little attention to family succession. This is surprising, given that a large proportion of private enterprises are family businesses (Klein Reference Klein2000; Lee Reference Lee2006; Zahra, Hayton and Salvato Reference Zahra, Hayton and Salvato2004). Approximately 95 per cent of all firms in Germany are family firms (Haunschild et al. Reference Haunschild, Wallau, Hauser and Wolter2007) with similar rates in other European countries and the United States of America (USA) (Austrian Institute for SME Research 2008; Lee Reference Lee2006). Altogether, about two-thirds of firms worldwide are family businesses (Gersick et al. Reference Gersick, Davis, Hampton and Lansberg1997). This implies that family succession should be an important mechanism through which a considerable amount of people become entrepreneurs. In this context, it is important to note that early research in entrepreneurship suggested that growing up in a family with an entrepreneurial background increases the likelihood of pursuing a career as an entrepreneur; however, the main mechanism to explain this effect was social learning or role modelling, respectively (Scherer et al. Reference Scherer, Adams, Carley and Wiebe1980; White, Thornhill and Hampson Reference White, Thornhill and Hampson2007). We would like to add to this perspective by showing that family business owners may not only have an influence on their successors' careers by acting as role models but also by being generative and actively guiding members of the younger generation. Thus, identifying family business owners' generativity as an important factor with a direct effect on family succession would provide new insights into the question of why some members of the next generation become entrepreneurs while others do not in the context of family businesses.
Development of hypotheses
Business owners' age and family succession
Upper echelons theory (Hambrick and Mason Reference Hambrick and Mason1984) states that business managers' demographic characteristics, such as age, influence their managerial decisions and actions as these characteristics are associated with core values, capabilities and perceptions that have an impact on behaviour and decision making. Based on the notions of upper echelons theory (Hambrick and Mason Reference Hambrick and Mason1984), we argue that business owners' age is a particularly relevant predictor of family succession in small business settings. As business owners grow older and approach the end of their careers, they should become increasingly interested in transferring their businesses to the next generation and handing over their control of the family businesses. Lansberg (Reference Lansberg, Aronoff and Ward1991) suggested that as business owners approach retirement, they are often reminded of their own mortality. Lifespan psychology researchers have suggested that as people approach and perceive endings in their lives (e.g. retirement, death), they tend to prioritise generative and emotionally meaningful goals over goals instrumental for own advancement (Lang and Carstensen Reference Lang and Carstensen2002). We argue that transferring a family business to the next generation can be considered such a generative and emotionally meaningful goal in family business settings as it includes aspects of caring and taking responsibility for future generations and the sustainability of the business. This line of reasoning is supported by a recent study by Marshall et al. (Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006), who found that older business owners were more likely to make formal succession plans than younger business owners.
● Hypothesis 1: Business owners' age is positively related to family succession.
Business owners' age and generativity
We further propose that older business owners more strongly prioritise generative concerns than younger business owners. When Erikson (Reference Erikson1950) first described the psycho-social conflict between generativity and stagnation (i.e. caring only for oneself) as the seventh out of eight stages in his seminal theory of lifespan development, he suggested that generative concerns first arise in middle or late adulthood. In their comprehensive generativity theory based on Erikson (Reference Erikson1950), McAdams and de St. Aubin (Reference McAdams and de St. Aubin1992) argued that individuals' ‘need to be needed’ and their ‘desire for symbolic immortality’ become stronger with increasing age. They assumed that these mostly subconscious motivations combine with age-related social norms (cf. Neugarten, Moore and Lowe Reference Neugarten, Moore and Lowe1965) to fuel a growing conscious concern for the next generation during mid-life (McAdams and de St. Aubin Reference McAdams and de St. Aubin1992). Empirical findings generally supported the assumption that generative concerns become more important from young to middle adulthood (Keyes and Ryff Reference Keyes, Ryff, McAdams and Aubin1998; McAdams, de St. Aubin and Logan Reference McAdams, de St. Aubin and Logan1993; Sheldon and Kasser Reference Sheldon and Kasser2001; Stewart and Vandewater Reference Stewart, Vandewater, McAdams and de St. Aubin1998). An age-related increase in generativity is further supported by research on socio-emotional selectivity theory (Carstensen Reference Carstensen1995), which suggests that age-related decreases in perceived remaining lifetime lead to a prioritisation of emotionally meaningful and generative life goals. Whereas young adults prioritise goals that advance their own careers and that are instrumental with regard to expanding their breadth of knowledge, older adults place more emphasis on the engagement in meaningful and lasting social activities (Carstensen, Isaacowitz and Charles Reference Carstensen, Isaacowitz and Charles1999; Grant and Wade-Benzoni Reference Grant and Wade-Benzoni2009; Ng and Feldman Reference Ng and Feldman2010). Research showed that decreases in perceived remaining lifetime indeed resulted in a preference for generativity and emotional meaningful goals (Lang and Carstensen Reference Lang and Carstensen2002).
● Hypothesis 2: Business owners' age is positively related to generativity.
Generativity and family succession
Generativity theory (McAdams and de St. Aubin Reference McAdams and de St. Aubin1992) suggests that peoples' generative concerns may translate into a wide variety of behavioural expressions of generativity such as teaching younger people, taking over leadership roles, and helping to improve the community for future generations. Generativity in the work context is linked to greater work satisfaction, wellbeing, subjective career success (Ackerman, Zuroff and Moskowitz Reference Ackerman, Zuroff and Moskowitz2000; Clark and Arnold Reference Clark and Arnold2008) and predicts a variety of prosocial behaviours (Rossi Reference Rossi2001). Even though several researchers have proposed in recent years that generativity is an important psycho-social concern in the work and organisational context (Clark and Arnold Reference Clark and Arnold2008; Grant and Wade-Benzoni Reference Grant and Wade-Benzoni2009; Kanfer and Ackerman Reference Kanfer and Ackerman2004; Mor-Barak Reference Mor-Barak1995; Zacher et al. Reference Zacher, Rosing, Henning and Frese2011), so far no research exists that has investigated the relationship between business owners' generativity and family successions as an important outcome in family business settings.
Investigating the influence of family business owners' generativity on family succession is theoretically important because it introduces a concept from ageing and developmental psychology to gain a better understanding of processes in the management of family businesses and the career choices of potential successors. We suggest that business owners high in generativity are more likely to successfully make plans to transfer the family business to the next generation than business owners low in generativity. Generative business owners have realised and accepted that their careers are coming to an end, and that it is the next generation's turn to take over and to run the business. Generative business owners are also less inclined to seek personal enhancement and more likely to help others to find a purpose in their lives. In addition, they may consider their business an important part of their personal legacy, and handing it over to a member of the next generation may give meaning to their life story (Kotre Reference Kotre1999).
● Hypothesis 3: Generativity is positively related to family succession.
Generativity as a mediator of the relationship between business owners' age and family succession
Research on ageing is sometimes criticised for treating chronological age as if it was a psychologically meaningful construct by itself. For example, Birren argued that,
By itself, the collection of large amounts of data showing relationships with chronological age does not help, because chronological age is not the cause of anything. Chronological age is only an index, and unrelated sets of data show correlations with chronological age that have no intrinsic or causal relationship with each other. (1999: 460)
We suggest that family business owners' generativity is an important psycho-social construct that may help explain the positive relationship between family business owners' age and family succession. In fact, the mechanisms that link business owners' age and family succession are so far not yet well understood. For example, Marshall et al. (Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006) proposed that leadership and conflict management styles function as mediators in this relationship, but they failed to find support for this hypothesis. Based on the theoretical arguments and empirical studies presented in the previous sections, we propose that generativity is a mediating mechanism of the positive relationship between family business owners' age and family succession. In other words, family business owners' generativity is proposed to increase with age, and this increased concern for the next generation is responsible for successfully making succession plans.
● Hypothesis 4: Generativity mediates the relationship between business owners' age and family succession.
Method
Setting of the present study
We investigated our hypotheses using a sample of family business owners in Germany. Family businesses are the dominant category of firms in Germany, and contribute importantly to economic growth and employment (Klein Reference Klein2000). According to a recent report of the German Institute for Small and Medium-sized Enterprises Research (Haunschild et al. Reference Haunschild, Wallau, Hauser and Wolter2007; Institut für Mittelstandsforschung 2008), approximately 95 per cent or three million of all German businesses are family businesses. Family businesses make approximately 42 per cent of Germany's national revenue (€1.9 trillion) and provide 57 per cent (13.4 million) of all jobs in Germany. The most successful German family businesses are highly specialised, they operate in particular market segments and are famous for their niche strategies (Klein Reference Klein2000).
Participants and procedure
The data used in this study came from 155 family business owners in central Germany. Several different definitions of what constitutes a family business exist in the literature (Chua, Chrisman and Sharma Reference Chua, Chrisman and Sharma1999; Heck and Trent Reference Heck and Trent1999; Klein, Astrachan and Smyrnios Reference Klein, Astrachan and Smyrnios2005). This study includes businesses that identify themselves as family businesses by measuring business owners' self-perceptions of the firm. Twelve (7.7%) of the business owners in the sample were female and 142 (91.6%) were male (one owner (0.6%) did not report his or her gender). Age distribution ranged from 26 to 83 years, and the average age was 52.82 years (standard deviation=10.28). In terms of education, one participant (0.6%) had no school degree, 16 (10.3%) had a general education degree, 54 (34.8%) had a middle school degree, 35 (22.6%) had a high school degree, and 47 (30.3%) had a university degree.
As a first step of data collection for this study, we randomly selected 1,500 businesses located in the central German state of Hesse listed in the Hoppenstedt file, a large database of German firms that provides company profiles and financial information. We called each firm and asked the business owner whether he or she would be willing to participate in our study. Participation was voluntary; as an incentive, participants were promised to receive a feedback report on the study results. Subsequently, we sent a questionnaire package to those 821 business owners (55%) who indicated their general interest in participating; 247 business owners mailed their questionnaires back to us in prepaid envelopes (30% response rate; 16.5% overall response rate). The relatively low overall response rate is consistent with response rates of similar studies in the fields of entrepreneurship and small business management. Specifically, studies using a similar approach (i.e. sending questionnaires to business owners or CEOs) reported response rates between 13 and 25 per cent. For example, Keh, Nguyen and Ng (Reference Keh, Nguyen and Ng2007) reported a response rate of 13.1 per cent in Singapore. In the USA, response rates of similar studies were 13.2 per cent (Hmieleski and Ensley Reference Hmieleski and Ensley2007), 24.8 per cent (Hmieleski and Baron Reference Hmieleski and Baron2009) and 19.8 per cent (Baron and Tang Reference Baron and Tang2011). Thus, our overall response rate of 16.5 per cent is within the range of similar previous research. To test the representativeness of our sample, we compared our sample to the population of 1,500 businesses and conducted t-tests on firm age, revenue, number of employees, firm growth in revenue between 2006 and 2009, and firm growth in employees between 2006 and 2009. All t-tests yielded non-significant results except for revenue. Compared to the population, our sample generated less revenue (€4.01 million versus €3.00 million). This suggests that our sample is representative in terms of firm age, number of employees, firm growth in revenue and firm growth in employees, and it is not representative in terms of total revenue.
Of the 247 respondents, 155 questionnaires came from family business owners which provided the data of interest for this study. We treated missing data in our study following the ‘fundamental principle of missing data analysis’ proposed by Newman: ‘Across missing data conditions, the best data-analytic methods for dealing with missing data follow a simple yet fundamental principle: use all of the available data’ (2009: 11, italics in original). According to Newman (Reference Newman, Lance and Vandenberg2009), this principle is superior to the more commonly applied missing data techniques (i.e. listwise and pairwise deletion) because it does not discard empirical information. The number of missing values on the item level ranged from zero to seven (4.5%) in our study. For the multiple-item scales (generativity, family culture and family involvement), we followed the advice by Newman (Reference Newman, Lance and Vandenberg2009) and addressed item-level non-response using mean-item imputation (Roth, Switzer and Switzer Reference Roth, Switzer and Switzer1999), which means averaging across the items with available responses to calculate a scale score. The number of missing values on the scale level ranged between zero and seven (4.5%) in the study variables. We followed the advice by Newman (Reference Newman, Lance and Vandenberg2009) and addressed scale-level non-response by imputing missing data using the SPSS/PASW routine for expectation-maximisation (EM) estimation (Schafer and Graham Reference Schafer and Graham2002). Note that six of our study variables (gender, age, education, experience, firm age, succession) were single-item measures and not ‘scales’ but were also imputed using the EM procedure.
Measures
Business owners' generativity
We assessed business owners' generativity by self-report using three items which we adapted from the three items recently developed by Zacher et al. (Reference Zacher, Rosing, Henning and Frese2011) to measure leader generativity in the university workplace. Specifically, we exchanged the word ‘scientists’ with ‘business owners’ in the first item, the words ‘in my field’ with ‘for the business’ in the second item, and the word ‘academics’ with ‘business owners’ in the third item. The items used in this study were ‘I devote more energy to building up the next generation of business owners than to getting ahead myself’, ‘I am more strongly concerned with establishing successful successors for the business than with working on my own success’ and ‘I use more time for rearing young business owners than for making progress in my own career’.
The items were answered on five-point scales ranging from 1 (‘not true at all’) to 5 (‘very true’). Cronbach's α of this scale was 0.96 (Zacher et al. Reference Zacher, Rosing, Henning and Frese2011, reported an α of 0.89).
Zacher et al. (Reference Zacher, Rosing, Henning and Frese2011) developed their items based on Erikson's classic definition of generativity (versus stagnation) as ‘primarily the concern in establishing and guiding the next generation’ (1963: 267). Furthermore, consistent with recent research on generative goal priorities in the lifespan psychology literature (Lang and Carstensen Reference Lang and Carstensen2002), these items reflect the prioritisation of generativity goals over own career goals. Zacher et al. (Reference Zacher, Rosing, Henning and Frese2011) showed that self-reported leader generativity was positively and moderately correlated with follower ratings of leader generativity and with self-reports of broad generative accomplishments as assessed by the Loyola Generativity Scale (LGS; McAdams and de St. Aubin Reference McAdams and de St. Aubin1992, cf. Clark and Arnold Reference Clark and Arnold2008). Zacher et al. also provided initial evidence for the convergent and discriminant validity of their generativity scale with respect to the Big Five personality traits, narcissism, and altruism, power, and achievement life goals. The newly developed measure of work-related generativity has a number of advantages compared to McAdams and de St. Aubin's (Reference McAdams and de St. Aubin1992) broad measure of generative accomplishments (LGS) which has been widely used by ageing researchers (McAdams and de St. Aubin Reference McAdams and de St. Aubin1998) and some organisational psychologists in the past (Clark and Arnold Reference Clark and Arnold2008). First, the LGS includes 20 very heterogeneous items, some of which do not seem relevant for generativity in the work context (e.g. ‘I believe that society cannot be responsible for providing food and shelter for all homeless people’) or refer to other established constructs in organisational psychology such as creativity or productivity (e.g. ‘I try to be creative in most things that I do’, ‘Other people say that I am a very productive person’). Despite the heterogeneous nature of the scale, all previous studies using the LGS averaged the scores across the 20 items and used an overall score in the analyses (Clark and Arnold Reference Clark and Arnold2008). Second, the LGS assesses past generative accomplishments instead of current concerns with generativity (Clark and Arnold Reference Clark and Arnold2008; Keyes and Ryff Reference Keyes, Ryff, McAdams and Aubin1998). For example, two sample items from the LGS are ‘Others would say that I have made unique contributions to society’ and ‘I have made many commitments to many different kinds of people, groups, and activities in my life’. Finally, the LGS is not consistently related to age (McAdams and de St. Aubin Reference McAdams and de St. Aubin1992) as would be expected from Erikson's (Reference Erikson1950, Reference Erikson1963) theory of lifespan development. Based on these criticisms, we decided to use the measure by Zacher et al. (Reference Zacher, Rosing, Henning and Frese2011).
Family succession
We measured family succession using the question ‘Does a successor who will take over the business exist within the family?’ Business owners answered the question with either yes (1) or no (0). This categorical measure of family succession is consistent with previous studies on the topic (e.g. Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006) and represents an objective and valid measure of family succession because it asks for an actual circumstance.
Demographic and control variables
Business owners reported their age, gender (0=male, 1=female), and their highest German educational degree achieved (0=no degree, 1=general education degree, 2=middle school degree, 3=high school degree, 4=university degree). We controlled for gender because research suggests that the experiences of female business owners in the succession process may be different from the experiences of male owners (Cadieux, Lorrain and Hugron Reference Cadieux, Lorrain and Hugron2002; Covin Reference Covin1994; Harveston, Davis and Lynden Reference Harveston, Davis and Lynden1997). In addition, business owners indicated the year their firm was established and their entrepreneurial experience. Entrepreneurial experience was computed as the sum of years as a manager of the current business and the number of years managing other businesses. Finally, we controlled for family culture and family involvement in the business because these factors may influence family succession. Family culture was assessed with three items adapted from Klein, Astrachan and Smyrnios (Reference Klein, Astrachan and Smyrnios2005). The items were ‘Family members agree with the family business goals, plans, and policies', ‘Family members really care about the fate of the family business’ and ‘I understand and support my family's decisions regarding the future of the family business’. Items were answered on five-point scales ranging from 1 (‘does not apply at all’) to 5 (‘applies completely’). Cronbach's α of this scale was 0.88. Family involvement was assessed with two items taken from Klein, Astrachan and Smyrnios (Reference Klein, Astrachan and Smyrnios2005): ‘How many family members participate actively in the business?’ and ‘How many management board members are family?’ Cronbach's α of this two-item scale was 0.69.
Results
The descriptive statistics and intercorrelations of the study variables are shown in Table 1. Age was positively related to generativity (r=0.54, p<0.01), family succession (r=0.24, p<0.01) and experience (r=0.74, p<0.01). In addition, generativity was positively related to family succession (r=0.51, p<0.01) and experience (r=0.40, p<0.01), and negatively related to education (r=−0.23, p<0.01). Finally, family succession was positively related to both family culture (r=0.33, p<0.01) and family involvement (r=0.36, p<0.01).
Table 1. Means, standard deviations and intercorrelations of study variables
Notes: Sample size: 155. SD: standard deviation. Reliability estimates (α) are shown in parentheses along the diagonal.
Significance levels:
* p<0.05
** p<0.01.
We used a PASW script for mediation analyses provided by Preacher and Hayes (Reference Preacher and Hayes2008) to test our hypotheses. The script incorporates the steps suggested by Baron and Kenny (Reference Baron and Kenny1986) for tests of mediation and additionally tests the significance of the indirect (mediated) effect using the recommended bootstrapping approach (MacKinnon, Fairchild and Fritz Reference MacKinnon, Fairchild and Fritz2007; Shrout and Bolger Reference Shrout and Bolger2002). In all analyses, we included the complete set of control variables. However, we also ran all of the analyses without control variables and an equivalent pattern of results for all of the hypotheses emerged. Thus, we can rule out that the control variables are a potential explanation for our findings (Becker Reference Becker2005). The results are shown in Table 2.
Table 2. Results of mediation analysis
Notes: Sample size: 155.
1. t-Value from linear regression analysis.
2. β-Coefficient from linear regression analysis. SE: standard error. OR: odds ratio. LL: lower limit. CI: bias corrected and accelerated confidence interval. UL: upper limit. Number of bootstrap resamples=5,000.
According to Hypothesis 1, age is positively related to family succession. As shown in Table 2, age positively predicted family succession (B=0.07, standard error (SE)=0.03, Z=2.35, p<0.05, Wald=5.52, odds ratio (OR)=1.08), supporting Hypothesis 1.
According to Hypothesis 2, age is positively related to generativity. As shown in Table 2, age positively predicted generativity (B=0.07, SE=0.01, β=0.53, t=5.14, p<0.01), supporting Hypothesis 2.
Hypothesis 3 states that generativity has a positive influence on family succession. As shown in Table 2, generativity was positively related to family succession (B=1.25, SE=0.28, Z=4.51, p<0.01, Wald=20.36, OR=3.49), supporting Hypothesis 3.
According to Hypothesis 4, generativity mediates the relationship between age and family succession. Table 2 shows that the relationship between business owners' age and family succession became smaller and non-significant when controlling for generativity (B=0.02, SE=0.04, Z=0.55, p=0.58, Wald=0.30, OR=1.02). Thus, the relationship between age and family succession was fully mediated by generativity. Preacher and Hayes’ (Reference Preacher and Hayes2008) script also computes the indirect effect with bias corrected confidence intervals (CI). The results of a bootstrap analysis showed that the indirect effect of age on family succession (through generativity) was significant (indirect effect=0.09, SE=0.04, lower 95% CI=0.03, upper 95% CI=0.16). Together, these results provide support for Hypothesis 4.
To address potential limitations of EM estimation (Newman Reference Newman, Lance and Vandenberg2009; Schafer and Graham Reference Schafer and Graham2002), we ran our analyses with and without imputed data. The results of the analysis without imputed data (listwise N=144) were equivalent to the results of the analysis with imputed data. All hypothesised direct and indirect effects (Hypotheses 1–4) were positive and significant. Furthermore, the standard errors of the coefficients in both analyses were of an equivalent magnitude, indicating that there was no problem with biased standard errors in our study.
Discussion
Demographic changes and the ageing of the workforce in most Western countries (e.g. Australia, Germany, as well other European countries and the USA) and some Eastern countries such as China, India and Japan have led to an increased interest in the role of age in entrepreneurship and small business settings (Curran and Blackburn Reference Curran and Blackburn2001; Gielnik, Zacher and Frese in press; Lévesque and Minniti Reference Lévesque and Minniti2006). The goal of this study was to investigate the developmental construct of generativity as an important psycho-social mechanism in the relationship between business owners' age and succession in family firms in Germany. Germany provides a relevant socio-cultural context for investigating family business succession, as the vast majority of firms are family businesses (Institut für Mittelstandsforschung 2008). This is similar to most other European countries (Austrian Institute for SME Research 2008) as well as the USA (Lee Reference Lee2006; Shanker and Astrachan Reference Shanker and Astrachan1996).
Consistent with previous research, our results showed that business owners' age was positively related to family succession (Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006). Older business owners were more likely to report that they have successfully made succession plans. This positive relationship was fully mediated by business owners' level of generativity. Consistent with generativity (McAdams and de St. Aubin Reference McAdams and de St. Aubin1992) and socio-emotional selectivity theories (Carstensen Reference Carstensen1995), older compared to younger business owners reported a higher priority of concerns regarding the establishment and guidance of members of the next generation than of concerns regarding their own careers and accomplishments. Business owners' generativity, in turn, was positively related to family succession.
Our study aims to contribute to the literature in several ways. First, our findings expand current perspectives in the small business literature by taking a lifespan perspective on family succession in family business settings. Succession is related to the family business owners' ageing (Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006). However, current theoretical frameworks aiming to explain family succession do not consider constructs from the literature on ageing and developmental psychology that might act as mediating mechanisms and thus explain age-related changes (e.g. Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006; Sharma et al. Reference Sharma, Chrisman, Pablo and Chua2001, Reference Sharma, Chrisman and Chua2003). Our study extends the current literature by showing that the concept of generativity helps in understanding succession in family businesses which is an important success criterion in family business settings. Thus, using theories and concepts from the fields of ageing and developmental psychology provided new insights and a new perspective on processes important for a considerable proportion of private enterprises.
Second, our study extends the literature on the different possible career pathways to entrepreneurship. Why some people become entrepreneurs and others do not is an important question in the entrepreneurship domain (Baron Reference Baron2004). So far, three general perspectives on this question exist. First, the individual perspective proposes that people's individual characteristics, interests and personality traits exert a strong influence on whether one will become an entrepreneur (Rauch and Frese Reference Rauch and Frese2007; Scanlan Reference Scanlan1980; Zhao and Seibert Reference Zhao and Seibert2006). Second, the economic perspective suggests that macro-economic factors, such as regional economic diversity or population growth, determine the rate to which people engage in entrepreneurship (Reynolds, Miller and Maki Reference Reynolds, Miller and Maki1995). Third, the socialisation or social learning perspective suggests that parents act as role models and thus influence their children's career plans (Scherer et al. Reference Scherer, Adams, Carley and Wiebe1980). We add to this literature by showing that family business owners may also influence the career paths of their successors by being more or less generative. Thus, attempts to explain the emergence of entrepreneurs in the context of family businesses must take into consideration factors on different conceptual levels. Besides individual and economic factors, the social context in which the successor seeks to pursue a career as entrepreneur exerts a strong influence. Our results are also in line with recent findings on the role of authoritative parenting and mentoring on entrepreneurship (Ozgen and Baron Reference Ozgen and Baron2007; Schmitt-Rodermund Reference Schmitt-Rodermund2004). These findings suggest that pursuing a career as an entrepreneur depends on significant people other than the entrepreneur, such as incumbent business owners, parents and mentors who help with preparing the career path of the future entrepreneur.
Finally, an important contribution of this study is that it helps to shed light on a psychological process underlying the observed positive relationship between chronological age and a culturally important phenomenon, family succession in small businesses (Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006). Ageing research is sometimes criticised for treating chronological age as if it was a psychologically meaningful construct by itself (Birren Reference Birren, Bengtson and Schaie1999). Our study provides an example of how to investigate and understand relationships between age and work-related outcomes more appropriately. Specifically, we treated family business owners' age as a number that is somewhat psychologically arbitrary alone, but is made more meaningful by understanding how the psychological construct of generativity explains the positive relationship between age and family succession. Thus, our study contributes to a richer understanding of the role of business owners' age in family succession above and beyond previous research (Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006). Our findings also extend the developmental literature by showing that the age-related construct of generativity is related to an objective outcome in small business settings, family succession.
Limitations
This study has a number of limitations. First, our cross-sectional design does not allow inferences regarding intra-individual changes in generativity across the lifespan (Van Der Velde, Feij and Van Emmerik Reference Van Der Velde, Feij and Van Emmerik1998). The age-related differences between business owners may also be attributable to differences between birth cohorts (Smola and Sutton Reference Smola and Sutton2002). In addition, our cross-sectional design does not allow definite conclusions about the causal mechanisms connecting business owners' age, generativity and family succession. Future research needs to employ longitudinal and cohort-sequential designs to better disentangle ageing and cohort effects on family succession and to draw definite causal conclusions. Moreover, entrepreneurship researchers agree that succession is a multi-stage process rather than a single event (Handler Reference Handler1989, Reference Handler1994; Miller, Steier and Le Breton-Miller Reference Miller, Steier and Le Breton-Miller2003). Again, longitudinal studies are needed to investigate these dynamics in the succession process (Miller, Steier and Le Breton-Miller Reference Miller, Steier and Le Breton-Miller2003).
Second, our measures of generativity and family succession may be criticised. We decided to use a short, practical and work-related generativity scale to assess business owners' concerns with generativity (Zacher et al. Reference Zacher, Rosing, Henning and Frese2011). We did not use the longer and widely researched LGS (McAdams and de St. Aubin Reference McAdams and de St. Aubin1992) because it has been criticised with regard to its construct validity by both ageing and organisational psychologists (Clark and Arnold Reference Clark and Arnold2008; Keyes and Ryff Reference Keyes, Ryff, McAdams and Aubin1998). Nevertheless, it may be argued that our approach to assessing Erikson's (Reference Erikson1950, Reference Erikson1963) notion of generativity is too narrow. Our conceptual definition and operationalisation limit generativity to a particular situational context – family business settings. Thus, we may not have captured the complexity of the processes involved in generativity as outlined by Erikson (Reference Erikson1950, Reference Erikson1963) and extended by McAdams and de St. Aubin (Reference McAdams and de St. Aubin1992, Reference McAdams and de St. Aubin1998) in their comprehensive theory of generativity. Similarly, even though the items of our generativity scale emphasise different aspects of business owners' generativity (concern, time and energy investment), one might argue that the items focus too homogenously on one specific aspect of generativity. In fact, our analyses showed that the internal consistency of the scale is very high (Cronbach's α=0.96). We think, however, that the high internal consistency of the items does not invalidate our findings. Rather, it shows that the specific aspect of prioritising the guidance of members of the younger generation over focusing on one's own career goals and accomplishments captures generativity in the way it is hypothesised by generativity theory (Erikson Reference Erikson1950; McAdams and de St. Aubin Reference McAdams and de St. Aubin1992, Reference McAdams and de St. Aubin1998). Using the same measure, Zacher et al. (Reference Zacher, Rosing, Henning and Frese2011) showed that the items explain age-related effects in a university setting. Thus, the items capture generativity in accordance with theoretical propositions by generativity theory across different contexts. Furthermore, Zacher et al. (Reference Zacher, Rosing, Henning and Frese2011) provided evidence for the convergent and discriminant validity of the items. We therefore think that the scale is a valid operationalisation of the generativity construct. However, we suggest that future research could include additional items to assess a broader generativity construct or different facets of generativity.
Our one-item categorical measure of family succession is consistent with previous studies on the topic (e.g. Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006) and we consider it an objective and valid measure of family succession, which minimises the potential problem of common method variance. However, future studies might want to assess additional aspects of the family succession process. Indeed, family succession is a complex process and likely to be influenced by actors and conditions inside and outside the family business (Davis and Tagiuri Reference Davis and Tagiuri1989; Handler Reference Handler1994). Besides business owners' age, other factors to be considered are the age of the successor and the quality of the relationship between owner and successor (Davis and Tagiuri Reference Davis and Tagiuri1989).
Finally, the overall response rate in our study was only 16.5 per cent, which might limit the generalisability of our findings. It is important to note, however, that our sample consists of business owners heading the top management of the business. Research showed that the response rate in this group is significantly lower than in other groups (Baruch Reference Baruch1999). Our response rate is consistent with previous research which reported response rates between approximately 13 and 25 per cent (Hmieleski and Baron Reference Hmieleski and Baron2009; Keh, Nguyen and Ng Reference Keh, Nguyen and Ng2007). Even more importantly, testing our sample against the population showed that our sample did not significantly differ in firm age, number of employees, firm growth in revenue and firm growth in employees. These findings suggest that our sample is representative of the population with regard to firm age, number of employees, firm growth in revenue and firm growth in employees. There was a significant difference in total revenue. Thus, our findings may not generalise to firms that generate more revenue than the firms in our sample.
Future research
Future research should investigate which factors explain the age-related increase in generativity among family business owners. So far, no empirical research exists that shows that inner desires and cultural demands indeed combine to create a growing concern for the next generation (McAdams and de St. Aubin Reference McAdams and de St. Aubin1992). In addition, based on socio-emotional selectivity theory (Carstensen Reference Carstensen1995), occupational future time perspective (Zacher and Frese Reference Zacher and Frese2009) might be investigated as a cognitive-motivational mediator between business owners' age and generativity concerns in the family business context. Business owners with a strong focus on future opportunities and much perceived remaining time may be less likely to hand over the business to the next generation because they perceive many occupational opportunities for themselves in the business. There may also be a number of factors that are theoretically important for the family succession process that were not tested in this study. Besides generativity, future research could include important constructs such as the transfer of tacit knowledge and social capital (Cabrera-Suárez, De Saa-Pérez and García-Almeida Reference Cabrera-Suárez, De Saa-Pérez and García-Almeida2001; Steier Reference Steier2001) into more comprehensive models of family succession. In addition, future research could focus on other operationalisations of age than chronological age. For example, ‘subjective age’ (e.g. how old individuals feel; Cleveland, Shore and Murphy Reference Cleveland, Shore and Murphy1997; Kooij et al. Reference Kooij, De Lange, Jansen and Dikkers2007) may also be a useful operationalisation of age.
Future research could also aim to replicate our findings in other countries and cultural contexts. Family businesses and family succession contribute importantly to the German economy (Institut für Mittelstandsforschung 2008), possibly leading to a heightened awareness of the significance of generativity in the population. It would be interesting to investigate whether similar or even stronger relationships can be found in cultures that place an even stronger emphasis on tradition, long-term planning and generativity in older people (e.g. many collectivistic Asian cultures). Germany is the seventh lowest ranking country among 62 countries in terms of family and in-group collectivism and the eighth lowest ranking country in terms of institutional collectivism (Gelfand et al. Reference Gelfand, Bhawuk, Nishi, Bechtold, House, Hanges, Javidan, Dorfman and Gupta2004). Whereas individualistic cultures such as the German culture regard individuals as independent, collectivistic societies emphasise group membership as the primary source of identity (Hofstede Reference Hofstede2001). Hence, it would be interesting to investigate whether our findings depend on the cultural background of the business owners.
In conclusion, our study extends current perspectives on the role of age in family business settings (e.g. Gielnik, Zacher and Frese in press; Marshall et al. Reference Marshall, Sorensen, Brigham, Wieling, Reifman and Wampler2006) by showing that the developmental construct of generativity mediated the positive relationship between business owners' age and family succession. The study demonstrates that theories and concepts from the ageing and developmental literature can make a meaningful contribution towards a better understanding of the process of family succession and career decisions of potential future entrepreneurs. Additional studies taking such an interdisciplinary approach are now needed to further establish the potentially important direct and indirect effects of age and generativity on other important work and business outcomes.
