Succession in family businesses: A meta-analysis
Name: Tom van EschStudent number: S1795058
Master: MSc Business Administration
Specialization: Small Business and Entrepreneurship Word count: 12.254
Supervisor: Andreas Rauch Co-assessor: Evelien Croonen Date: 20/01/2015
Abstract
Despite the importance of family businesses in the global economy, little is known about antecedents influencing the choice of succession within family businesses. In other words, what factors play a roll in making the choice for a successor? Before examining the antecedents the preferred form of succession is determined. After combining the proportions from various studies, intergenerational succession appeared to be the preferred form of succession opposed to external succession (management-buy-in or management-buy-out). The remainder of this research is aimed at identifying relevant antecedents currently apparent in the literature, selecting previous research focused those variables and combining the results using meta-analytical techniques. The results suggest that in family businesses located in individualistic countries a higher proportion of intergenerational succession appears. In contrast, results suggest that family businesses ran by higher educated predecessors show a lower proportion of intergenerational succession. For the remaining identified antecedents, firm size, age of predecessors and gender of successors, no significant correlations were found.
1 Introduction
Family business performance remains an important area of research because of its worldwide economic importance (Schulze & Gedailjolic, 2010). Globally, family firms create an estimated 70% to 90% of the GDP annually (Global Data Points, n.d., para. 5) and comprise a large part of all firms in the world. For example, in the United States family firms comprise 80% to 90% of all businesses (Astrachan & Shanker, 2003). Similarly, in Germany family businesses are estimated to cover 90% of all businesses (Haunschild & Wolter, 2010). In non-Western economies the same image appears. In the Gulf Cooperation Council (Saudi Arabia, Kuweit and most other of the Gulf states) approximately 98% of commercial activities are run by family businesses (Fadhel, 2004). In short, family firms are the dominating type of firm and account for a large part of the employment. However, many family firms, especially small and medium sized, have difficulties in the current globalizing markets. This has to do with the lack of resources, other personal factors and political influences. Yet, chances of successful internationalization increase when younger family members are involved in managing the company (Fernàndez & Nieto, 2005). Similarly, the environment for innovation improves when more generations are involved in the firm (Zahra, 2005). Based on these papers, the conclusion can be made that succession in family businesses helps them in competing with non-family businesses.
Family businesses can be distinguished from non-family businesses by the intention to shape and pursue the vision of the business held by a dominant coalition in a manner that is potentially sustainable across generations of the family (Chua, Chrisman & Sharma, 1999). In order to keep an enterprise within the family, to maintain activity and management through successor, an internal candidate must be found. However, evidence from a number of studies shows high failure rate in intergenerational succession (Davis & Harveston, 1998; Kets de Vries, 1996; Ward, 1987; Handler, 1992; Moris et al., 1997). Then the question arises, is there a suitable successor within the family who meets the required qualifications and shares the appropriate vision to take over the enterprise? Furthermore, the potential successor must be willing to take over the family business. When options to find an intergenerational successor are insufficient, a successor can be found externally to the family. Different external alternatives for maintaining a business are an external executive successor (management-buy-in) or an internal executive successor (management-buy-out). A lot of research has focused on factors influencing the outcome of succession (Sharma, Chrisman & Chua, 2003; Sharma et al., 2001; Venter et al., 2005). Up to now, only a limited number of studies have researched the influence of various antecedents influencing succession in family businesses and the findings are contradictory. Therefore, this paper is aimed at finding all the relevant available data on the decision concerning succession in family businesses. By using a meta-analysis, which operates independently from model composition, the results from these studies will be aggregated and eventually more generalizable statements will be derived.
2 Research Objective
The objective of this paper is to contribute to the current literature in several ways:
1. It identifies the preferred form of succession, intergenerational succession opposed to the alternative external forms of succession (management buy-in or management buy-out).
2. It identifies universal antecedents. 3. It proposes directions for future research.
The main area of research underlying these objectives is formulated in two parts:
I. What is the preferred of succession within family businesses, intergenerational succession on the one hand or external succession on the other hand (management in or management buy-out)?
II. What antecedents influence the preference for intergenerational succession (opposed to management buy-in or management buy-out) within family businesses?
3 Literature Review
In this chapter relevant antecedents affecting intergenerational succession, mentioned previously in the literature, will be discussed. But first, succession within family businesses in general will be touched upon. Since succession is the starting point for researching the antecedents affecting intergenerational succession. Succession planning is essential regarding the success and continuity of businesses (Pitcher, Cherim & Kisfalvi, 2000). In particular for family businesses, since only a relatively small percentage of those businesses last longer than one generation (Kets de Vries, 1993). Nepotism is generally viewed to be an important reason to internally transfer a family business to offspring or other family members (Allen & Langowtiz, 2003; Ambrose, 1983; Barach et al., 1988). However, this can be a harmful decision, since offspring might be favored over external potential successors in possession of superior qualities and may not be the desire of the shareholders (Barach et al., 1988).
characterized by a high level of mutual trust and honesty (Dyer & Handler, 1994), as a consequence there is a favorable transaction atmosphere among family members. This transaction cost atmosphere consists of sociocultural and technical factors that affect the transaction costs (coordination and motivation instruments)(Picot, Dietl & Franck, 2005). Altogether, in the light of the transaction cost theory, intergenerational succession has advantages over external succession because of the likelihood of a favorable transaction atmosphere in intergenerational succession, resulting in lower transaction costs.
According to Jensen and Meckling (1976) the agency theory deals with the agency problem occurring when cooperating individuals have different goals and vision of labor. This theory is directed at the agency relationship, wherein one party (principal) delegates word to another (agent). Two problems can arise during an agency relationship. The first problem entails different goals of the principal and the agent or when it is difficult for the principal to audit what the agent is doing. The second problem arises when the principal and the agent have different ideas about risk. There are several reasons why family businesses can avoid agency costs with intergenerational succession. Firstly, interests about growth opportunities and risks are automatically aligned, reducing the incentive to behave opportunistic. Therefore, costly mechanisms for separating management and control of decisions can be avoided (Fama & Jensen, 1983a). Secondly, because of private ownership property rights are mostly restricted to internal decision agents. As a consequence of their personal involvement, intergenerational successors are less likely to use shareholders wealth for their own interest. Again, costs associated with preventing such behavior can be avoided (Fama & Jensen, 1983a). Thirdly, shares of family businesses tend to be shared by agents with a special bond with other agents. Hence, agency problems can be settled without separating the management and the control decision (Fama & Jensen, 1983b).
Hypothesis 1: In family businesses intergenerational succession is more likely to occur opposed to management buy-in or management buy-out (external succession).
3.1 Firm size
Stavrou (1999) found that the intention of potential successors to take over a business, is positively correlated with firm size. This article suggests that smaller businesses may have smaller scopes, operating in niche markets and therefore may lose the interest of potential successors. The larger the parents’ business is, the greater the chances of joining and taking over the parents’ business. A possible explanation might be that larger businesses not only tend to be larger sources of wealth, leaders can expect to earn more income (Tosi et al., 2000). Similarly, Ambrose (1983) found that smaller businesses offer less financial reward and security. This is in line with findings of Venter et al. (2005) concluding that a decrease in firm size may lead potential successors to leave the firm, as a result of a perception that future monetary and non-monetary rewards may be unattractive. The previous arguments fit into the transaction cost theory perspective. Based on those arguments, willingness among potential successors in large businesses will be higher, which may lower negotiation costs. Furthermore, larger businesses in general reflect a larger amount of sacrifices resulting in more emotional commitment; therefore the tendency among owners to internally pass on the business increases (Denison, Lief and Ward, 2004). Another argument from different perspective comes from Stavrou (1999). Where he suggests that smaller businesses often have a more narrow product scope and/or operate in niche markets possibly reducing the interest of potential successors. Only Birley (1986) did not find a significant effect of firm size on potential successors intentions. Despite Birley’s (1986) result, previous arguments are summarized in the following hypothesis: Hypothesis 2: The proportion of intergenerational succession, opposed to management buy-in or management buy-out (external succession), increases parallel to firm size.
3.2 Culture
are more likely to be succeeded by an intergenerational successor (Stavrou, Merikas & Vozikis, 2008). From a transaction cost theory point of view, more collectivistic cultures can benefit more from appointing an intergenerational successor. Since, collective cultures are more loyal, less costs may be involved in the negotiation phase. From the agency cost perspective, the likelihood, of family businesses in more individualistic cultures, spending more during the alignment of opportunities and costs is higher. Since individualistic cultures are less focused on sustainability across generations. Based on this line of reasoning, the following relationship is expected:
Hypothesis 3: The proportion of intergenerational succession, opposed to management buy-in or management buy-out (external succession), increases parallel to collectivism.
3.3 Age predecessor
Between the age of 40 and 45, incumbents mainly focus on firm expansion. However, between the age of 50 and 60 they increasingly emphasize business and family philosophies and become less controlling (Davis, 1968). Similarly, DeTienne and Cardon (2010) argue that the intention to pass on the business to a family member is positively related to the age of the business owner. In the same line of reasoning, Santarelli and Lotti (2005) found that, as entrepreneurs age, the likelihood that a business will be succeeded to younger family members increases. Additionally, Battisti and Okamuro (2010) found that age has a negative impact on the choice of selling the business as an exit route. Underlying these findings, there are two main explanations. First, younger entrepreneurs are less likely to consider exiting the business since they may be a lack of adult potential intergenerational successors. Secondly, older entrepreneurs are more likely to have noneconomic motives for running a business. Most importantly, leaving a viable business for the next generation. Especially the second argument fits well in the agency theory perspective. This can be explained by the following reasoning. As a consequence of changing noneconomic goals of predecessors – passing on the firm – fewer costs will be associated with the process of alignment with the potential successor. Altogether, the following assumption is made:
Hypothesis 4: The proportion of intergenerational succession, opposed to management buy-in or management buy-out (external succession), increases as the family business incumbent transcend the age of 50.
3.4 Gender successor
Bukowitz, 1990). In contrast to son’s succession intentions, daughter’s succession intentions can be described as complementarity (Dumas, 1992). Furthermore, Dumas (1992) states that the relationship between fathers and daughters may be influenced by at-home behaviors due to a lack of previous experiences of collaboration. In contrast, sons are often groomed to become eventual successors. From a transaction cost theory perspective it is less costly to fit in male successors with previous business experience, which possibly makes this the preferred choice of succession. Partly, the exclusion of daughters in family businesses can be explained by parental concern and the desire to protect children form privation (Hollander & Bukowitz, 1990). The desire to protect children can be attributed to both male and female. However, it is less an issue for sons, since they are more expected to be independent and go their own way. In the same trend, Goldberg and Wooldridge (1993) found that firstborn, especially males, are more likely to join the family business opposed to female potential successors. Furthermore, Francis (1999) stated that family businesses are predominantly inherited by male successors, daughters are often offered less promising choices. On top of that, Dawley et al. (2004) found that traits typically associated with men are: achievement orientation, competency, forcefulness, activeness, emotional stability and rationality. Those traits are believed to be important for business success, and are believed to be lacking in women. However, another publication stated that family business owners in North America are more shifting towards creating pools of siblings sharing ownership (Gersick, Lansberg & Dunn, 1999). More recently, Bennedsen et al. (2005) concluded that family firms having a male first-child are 32.7 percent more likely to appoint an intergenerational successor than female first-child family firms. Therefore, the following hypothesis is formulated as follows:
Hypothesis 5: The proportion of intergenerational succession, opposed to management buy-in or management buy-out (external succession), increases with the presence of a potential male successor.
3.5 Education predecessor
potential successors external to the family. The family logic suggests that an incumbent treats succession as a provider of the family (Miller et al., 2011). Following a family logic intergenerational succession is the most obvious choice, since control over the business stays within the family (Berrone, Cruz & Gomez-Meija, 2012) and future job opportunities can be secured (Miller et al., 2011). Since they expected higher educated incumbents to act to a greater extent like CEOs. Searching for a family-related successor shifts to searching for the most capable successor. Higher educated predecessors are likely better able to judge successor’s abilities. Furthermore, it is likely that higher educated predecessors possess more knowledge about how to regulate non-related successor’s behavior (Wang & Si, 2011). The two previously given explanations fit well in the transaction cost theory and agency theory. Higher education will result in a relatively lower bounded rationality, which is one of the underlying assumptions of the transaction cost theory. In line with the agency theory, higher educated entrepreneurs are likely to be better aware of the differences between their own goals and vision of labor in contrast to a potential successor. Which can be hard to align and as a consequence entail extra costs. Finally, research by Klyver (2007) pointed out that, when entrepreneurs has between three and four years of higher education, the odds involving family members decreases by 43%. According to Klyver (2007), this may be because higher educated entrepreneurs are involved in more complex businesses. As a consequence, when choosing a less well-educated intergenerational successor, opposed to a better-well-educated external successor (out of a larger pool), a large information asymmetry is likely which entails higher transaction costs. Therefore, the following hypothesis is formulated as followed:
4 Data Collection and Methodology
The meta-analysis used conducting this research is aimed at developing a reasonable summary of quantitative findings (Lipsey & Wilson, 2001). This technique enables researchers to integrate and compare several first source results. Simultaneously, with this technique quality differences can be compensated by modifying results for different sample sizes and artifacts (Hunter and Schmidt, 1990). In this meta-analysis, a number of metafactors influencing family business succession will be examined. But first, the most common form of succession of family businesses will be determined, either intergenerational or external succession (management buy-in or management buy-out). The results from different studies are based on various theoretical streams, only family succession is the common denominator. The selected studies at least report the number of observed successions and the number of observed intergenerational or external successions. A reported proportion of intergenerational of external successions is feasible as well. In the next section the composition of the sample is further elaborated.
4.1 Select Studies as Input for the analysis
Initially the idea for investigating the preferred form of succession was initiated by my supervisor Andreas Rauch. After assisting the Institut für Mittelstandsforschung (IfM) in Bonn with a similar study based on German data, he was interested in this matter on a global scale. Naturally, the German sample, consisting of 17 studies, belongs the scope of this research and therefore this database was used as a starting point.
The first two steps of the procedure described below are predominantly pragmatic. Only the third and final step is based on actual data presented in studies. This choice is made since one individual undertook the selection of studies. Due to time constraint, it was not feasible to assess the presented data of the whole initial potentially relevant sample, consisting of 3818 studies. The first part of the examination consisted of reading the titles of the initial sample to determine whether they were potentially relevant. After this first step in the procedure, the number of potentially relevant studies was scaled back to 263 studies. This considerable drop in the number of studies was caused by studies focusing on the procedures and different stages of transferring family businesses instead of antecedents affecting the choice of succession. Secondly, all the abstracts of the remaining publications were examined. After this second step in the procedure, the number of potentially relevant studies was reduced to 44 studies. Finally, the remaining articles were checked for the presence of the required data. As previously mentioned, the number of observed successions had to be reported as well as the number of intergenerational and/or external successions. A proportion of the intergenerational and/or external succession also sufficed. This procedure resulted in 16 additional publications. However, one of those studies already existed in the German sample, it was a study related to the German government. In the end, the sample for this meta-analysis consisted of 32 studies. However, Ryan and Power (2012) use data from both Schotland and Ireland. Therefore, their publication is mentioned twice in the overall sample. Appendix 1 gives an overview of the research used in this meta-analysis.
4.2 Coding Studies
Prior to the analysis, the studies in the final sample were coded. First of all, the numbers of the different types of succession, intergenerational, management buy-in and management buy-out were reported. Some studies divide external succession in management buy-in and management buy-out. However, in contrast to the German sample, only two studies report in such a detailed fashion. Therefore, the choice is made to code subgroups of succession using two subgroups, namely: intergenerational succession and external succession. Otherwise, only two studies would be added to the initial German sample, strongly limiting the value of this research.
a subgroup with one study (medium size) and a subgroup with two studies (large size). To increase sample size within subgroups, calculations are also made using two subgroups. Here, the small and micro size groups are combined into one group and the medium and large groups are combined into one subgroup. Results for both of the coding version are displayed in Table 1.
Table 1 – Overview coding based on SME definition EU
Company category Employees Turnover Balance sheet
Large >250 > € 50 Million > € 43 Million
Medium <250 ≤ € 50 Million ≤ € 43 Million
Small <50 ≤ € 10 Million ≤ € 10 Million
Micro <10 ≤ € 2 Million ≤ € 2 Million
As clarified previously, culture is split into two dimensions based on Hofstede’s (1983) work, namely: collectivistic culture and individualistic culture. Hofstede (1983) rated 50 countries, all but one of the countries appearing in my sample are mentioned in this index. Countries rated among the upper 25 countries in this index are coded as individualistic countries. In contrast, countries rated among the lower 25 countries in this index are coded as collectivistic countries. Consequently, only three countries are coded as collectivistic opposed to 28 individualistic countries. However, coding based on mean scores in the index would result in coding individualistic countries as relatively collectivistic, caused by the high amount of individualistic countries in this sample. This would affect the results of the relationship at present undesirably. Therefore, the index developed by Hofstede (1983) is withheld. China is the only country not mentioned in this index. To code China, the website www.geert-hofstede.com was consulted. This website offers a tool, with which data can be requested from specific countries. The data for China was requested and the score of 20 out of 100 appeared on the individualism scale, consequently China is coded as a collectivistic country (Hofstede, n.d.).
Coding the incumbent’s age is based on Davis (1963), similarly to the theoretical framework. The studies in the sample were analyzed for the mean age. Studies reporting a mean age exceeding 50 years of age are coded as old; in case of a lower mean age studies are coded as young. Studies using age categories were coded based on the distribution over age categories. Studies coding the majority of their sample in categories exceeding 50 years of age are coded as old, when the majority of their sample is coded in categories lower than 50 years of age the studies are coded as young.
reporting a percentage exceeding this cut-off point are coded as masculine; studies reporting a percentage under this cut-off point are coded as feminine.
Empirically, the relationship between education and succession is determined in several publications. However, no clear overarching cut-off point formulated in exact years of education was found (Glauben et al., 2009; Kimhi & Nachlieli, 2001; Pérez-González, 2006; Stiglbauer & Weiss, 2000; Sifei, 2013; Zacher et al., 2012). Therefore, the decision is made to base the cut-off point on what information is available in the sample. Most of the studies reporting about education use categories. The level of aggregation varies from two to six categories. However, every study reporting about education distinguishes at least the difference between high school and further education. To be able to assign every study to a category, two categories are used to code education. Studies reporting the majority only finished high school were coded lowly educated, studies reporting the majority within their samples finished further education were rated as highly educated.
4.3 Protocol for Meta-analysis
As stated in the introduction a meta-analysis technique is used, operating independently from the composition model. Lipsey and Wilson’s (2001) protocol is used for undertaking this analysis. This protocol enables to compare across different research studies. The relationship studied in this paper concerns a one-variable relationship. A pattern of observations across categories and values of a single variable, succession, is researched. As mentioned previously, all findings involve succession operationalized in the same way, specified into intergenerational and/or external succession. To be able to perform a met-analysis, the data must enable you to define effect size and to determine the standard error (Lipsey & Wilson, 2001). For this meta-analysis the proportions of intergenerational succession are used to determine the overall effect size. According to Lipsey and Wilson (2001) there are two approaches available for generating the effect size for proportions: based on proportions and based on proportions conversed to logits. The latter is appropriate when proportions are less than .20 or greater than .80, due to the compression of the standard error. One proportion in the entire sample is not within those boundaries. Nevertheless, the first method is chosen since the other 32 proportions are within boundaries. The statistics used for calculating the effect size, sampling error and inverse variance weight are:
Effect size Standard error
where p refers to proportion, k is the number of subjects (intergenerational successions) in the category of interest and n is the total number of subjects in the sample (intergenerational and external successions).
The selected sample of studies was ordered using the software Excel. Similarly, the calculations were done using Excel since the absence more advanced expensive software to perform a meta-analysis.
This meta-analysis involves two categories of variables: effect sizes and the descriptive variables. The descriptive variables characterize the effect size and the study that generates the effect size. The descriptive variables reflect the antecedents, which are specified in the hypotheses.
Effect sizes are derived from the statistics reported in the studies in the sample, their statistical properties depend on their underlying sample size. Large samples contain less sampling error and therefore are more reliable and precise. This is why all the effect sizes are weighted by the inverse variance of the sampling error variance. So that its contribution is proportionate to its reliability.
According to Lipsey and Wilson (2001), the analysis of the effect size data is undertaken in two stages: 1) calculating the weighted mean effect size and its statistical significance and 2) assess the adequacy of the weighted mean effect size by a homogeneity test.
First, the mean effect size is calculated based on previously mentioned statistics. The weighted mean effect size is calculated with the following formula:
where ESi are the values on the effect size statistics used, wi is the inverse variance weight for effect size and i is equal to 1 to k, k being the number of effect sizes.
Thereafter, a test is performed to assess whether the weighted mean size is significant. More specifically, the confidence interval is calculated. The confidence interval indicates the range within which the population mean is likely to be. If the weighted mean effect size falls within this range this indicates a significant result. A confidence interval is based on the weighted mean effect size, the standard error and a critical value from the z-distribution. The standard error of the weighted mean effect size is calculated as the square root of the sum of the inverse variance weights (Hedges & Olkin, 1985):
Weighted mean effect size
!
After completing this calculation the confidence interval was determined by calculating the lower and upper boundaries based on calculations recommended by Hunter and Schmidt (2004). The confidence intervals are based on the weighted effect size. Adding and subtracting the square root of the standard error multiplied by 1.96, for a 1% significance level.
The second stage of the analysis consists of an assessment of the homogeneity of the effect sizes. To test whether the various effect sizes averaged into the weighted mean effect size estimate, to a certain extent, the same population effect size (Hedges, 1982). A homogeneous sample indicates that the distribution of effect sizes around the mean is no greater than the expected sampling error. If the null hypothesis is rejected, the distribution of is greater than the expected sampling error. In other words, individual effect sizes do not estimate a common population mean. The homogeneity test is based on the Q statistic, distributed as a chi-square with k-1 degrees of freedom, where k is the number of effect sizes (Hedges & Olkin, 1985):
Hunter and Schmidt (1990) propose a different approach, also described in Lipsey and Wilson’s (2001) protocol, to test whether effect sizes are homogeneous. Their method divides the observed variability in two parts: one part variability attributable to sampling error and the other part attributable to between study differences. A sample is labeled homogeneous if the sampling error accounts for 75% or more of the variability. The calculation is based on the following formula:
Both homogeneity tests are performed. First, the tests were performed to assess the homogeneity of the overall choice of succession, reflecting hypothesis 1. Next, the homogeneity of all the antecedents and the quality checks was tested one by one.
Finally, the significance between groups within antecedents was tested. Using a statistical significance test, described by Hunter and Schmidt (1990). This method is based on the difference in effect size between subgroups within antecedents. Subsequently, this difference is divided by the squared root of the sum of the standard error divided by the number of observations for both groups. The critical value for this test is z=1.96 at a 5% confidence interval.
! Q-test
Table 2 – Results meta-analysis
K n Weighted mean effect size Standard error mean effect size Confidence intervalI^2 Q value Z value
H1: Overall succession 33 18876 0,5932 0,0033 0,4981 to 0,6293 0,9865 3422,5602 H2: Firm size (four categories)
Micro 5 576 0,4027 0,0120 0,22437 to 0,5617 0,6501 27,1238 Micro vs. Small -0,6327 Micro vs. Medium -1,2123 Micro vs. Large 0,4786 Small vs. Medium -0,9655 Small vs. Large 0,9929 Medium vs. Large 1,4301 Small 5 7529 0,4477 0,0054 0,2608 to 0,6347 0,9915 1110,3932 Medium 1 1010 0,5722 0,0156 0,5555 to 0,5890 0,0000 Large 2 564 0,3482 0,0180 0,0492 to 0,6472 -4,55131 0,6917
H2: Firm size (two categories)
Old 6 1483 0,6382 0,0120 0,4359 to 0,6698 0,9020 112,9408 H5: Gender sample Masculine 5 1058 0,4300 0,0150 0,2775 to 0,5652 0,8132 50,8271 0,6947 Feminine 3 6220 0,3810 0,0060 0,2148 to 0,7193 0,9868 453,5615 H6: Education predecessor Low 3 5654 0,7581 0,0057 0,3622 to 0,8178 0,9373 95,4569 2,2214* High 4 1470 0,6027 0,0120 0,3970 to 0,7237 0,9570 181,6370 * p < .05
K=number of correlations; n=accumulated sample size; I^2=heterogeneity test Hunter and Schmidt (1990)
Table 3 – Results quality checks
K n Weighted
mean effect size
Standard error mean effect size
Confidence interval I^2 Q value Z value
Quality check: actual vs. intended
Actual succession 29 18195 0,5966 0,0034 0,5265 to 0,6667 0,9874 3292,3314 1,4934
Intended succession 4 681 0,4954 0,0179 0,3160 to 0,6748 0,9213 99,3550
Quality check: Tom vs. German
Tom 16 15197 0,5853 0,0036 0,4842 to 0,6862 0,9921 3171,2071 -1,6765
German 17 3679 0,6290 0,0077 0,5728 to 0,6853 0,8832 225,0891
Quality check: farming vs. no farming
No farming 27 12932 0,5069 0,0041 0,4356 to 0,5781 0,9806 1995,1320 -6,2304**
Farming 4 5944 0,7509 0,0055 0,5555 to 0,9464 0,8523 169,2899
** p < .01
5 Results
The meta-analytical results are presented in tables 2 and 3. Results are presented according to the basis of the sequence held in the theoretical framework. First, with computing the proportion of intergenerational succession within the whole sample hypothesis 1 is tested. The corrected proportion measure is set at p=0,5932, confirming the expected preference for intergenerational succession. However, the result does not pass the Lipsey and Wilson’s homogeneity test (Qtotal=3422,5602; df=46,19). Therefore, an analysis of the antecedents specified in hypotheses 2 until 6 was necessary.
Hypothesis 2 was tested using two different coding procedures. The first procedure divides the antecedent size in four categories; the second procedure divides the antecedent size in two categories. The first procedure reports the following corrected proportions of intergenerational succession: micro p=0,4027, small p=0,4477, medium p=0,5722 and large p=0,3482. The differences between all size categories are not significant at a 5% significance level. Comparing the size categories in range of the European definition, using four size categories, rejects hypothesis 2. The results do not pass the homogeneity test (Qsize(fourcategories)=1138,2088; df=21,03). However, the Q-value is smaller than the Q-value for the whole sample (Qtotal=3422,5602). The second procedure reports the following corrected proportions of intergenerational succession: micro+small p=0,4447 and medium+large p=0,5042. Again, the difference between the categories is not significant at a 5% significance interval. Therefore, also with the use of two categories, hypothesis 2 is rejected. The results do not pass the homogeneity test (Qsize(twocategories)=1209,1540; df=21,03). However, the Q-value for the sub-sample is smaller than the Q-value for the whole sub-sample (Qtotal=3422,5602).
The result of the computations for hypothesis 3 report the following corrected proportions of intergenerational succession: individualistic p=0,5999 and collectivistic p=0,3787. Results are significant at a 5% significance level. However, the significant effect is opposed to the expected effect. Therefore, hypothesis 3 is rejected. Furthermore, the results do not pass the homogeneity test (Qculture=3290,1596; df=43,78). However, the value for the sub-sample is smaller than the Q-value for the whole sample (Qtotal=3422,5602).
The result of the computations for hypothesis 4 report the following corrected proportions of intergenerational succession: young p=0,6946 and old p=0,6382. Results are not significant. Additionally, the direction of the effect is opposed to the expected effect. Therefore, hypothesis 4 is rejected. Furthermore, the results do not pass the homogeneity test (Qage=200,9040; df=14,07). However, the Q-value for the sub-sample is smaller than the Q-value for the whole sample (Qtotal=3422,5602).
Furthermore, the results do not pass the homogeneity test (Qage=200,9040). However, the Q-value for the sub-sample is smaller than the Q-value for the whole sample (Qtotal=422,5602).
The result of the computations for hypothesis 6 report the following corrected proportions of intergenerational succession: low education p=0,7581 and high education p=0,6027. The effect of the dividing the sample based on education is in line the expected effect. Additionally, the results are significant at a 5% significance level. Therefore, hypothesis 5 is confirmed. Furthermore, the results do not pass the homogeneity test (Qeducation=277,0939; df=12,59). However, the Q-value for the sub-sample is smaller than the Q-value for the whole sample (Qtotal=422,5602).
Additionally to the computations to test the hypotheses based on the theoretical framework, three quality checks were performed (Table 3). Therefore, the initial sample was divided in sub-samples based on outstanding characteristics to check whether differences would appear. First of all, it was notable that four publications within the sample did not focus on actual succession but on intended succession. The results of the computations report the following corrected proportions of intergenerational succession: actual succession p=0,5966 and intended succession p=0,4954. Although a difference in proportions appears, the difference is not significant. Secondly, the sample was divided based on the one hand on the initial German sample of the Institut für Mittelstandsforschung (IfM) and on the other hand on the additional added publications by myself. The results of the computations report the following corrected proportions of intergenerational succession: Tom’s sample p=0,5853 and German sample p=0,6290. Again, a small difference between the two subgroups exists. However, the difference is not significant. Finally, it was remarkable that four studies were focused specifically on businesses engaged in farming. Since no other industry was unduly apparent in the sample, the sample was divided in the sub-groups no-farming and farming. The results of the computations report the following corrected proportions of intergenerational succession: no-farming p=0,5069 and farming p=0,7509. This difference is significant at a 5% significance level, even at a 1% significance level.
6 Discussion
This research examines the preferred form of succession within family businesses and what antecedents influence the preferred form of succession. The aim of this research is to develop a better understanding of the various forms of succession within family firms. Up to a certain extend, expectations are fulfilled. However, some of the hypothesized effects of antecedents influencing the choice of succession are not confirmed. Below, the above-mentioned results will be discussed.
and confirmed based on a database in which 33 smaller samples are combined. This finding provides global support for the idea that costs can be reduced or eliminated when intergenerational successors are appointed. Costs related to the transaction theory, such as: search and information costs, the risk of losing trade secrets, negotiation costs and enforcing agreements. Since family businesses are characterized by a high level of trust and honesty (Dyer & Handler, 1994). And costs related to the agency theory can be avoided, such as costs associated to a difference between the agent and the principal. For example differences in goals and ideas about risks (Jensen & Meckling, 1979).
The results concerning the effect of firm size on the preferred form of succession rejects hypothesis 2. As explained before, hypothesis 2 is tested using two distinctive forms of coding. Independent of the form of coding used, hypothesis 2 is not confirmed. Except for the category large (using four categories), the proportions show the expected effect on the proportions. Namely, an increase in firm size causes a higher proportion of intergenerational succession. However, none of the differences found is significant. A possible explanation for the fact that large businesses are less likely to appoint intergenerational successors might be the indication Le Breton-Miller et al. (2004) give. According to their research, succession is more important in smaller family businesses compared to large publicly traded businesses (possibly family owned). Since the pool from which potential external successors can be selected is smaller. Besides, family relationships and attendant emotional factors are more complicated than just business relationships in large firms. Furthermore, there is a methodological issue. The number of studies coded medium (K=1) and large (K=2) is very small. This causes problems in the generalizability of this result. Therefore, the second form of coding was applied. As explained before, the groups micro and small (K=10) were combined. The same applies to the groups medium and large (K=3). Yet, the group combining medium and large firms is still relatively small. Altogether, the results do not confirm previous findings that future monetary and non-monetary rewards may be unattractive in relatively smaller firms (Ambrose, 1983; Stavrou, 1999; Tosi et al., 2000; Venter, Boshoff & Maas, 2005). This finding confirms Birley’s (1986) research outcome in which no significant relation between firm size and succession was found.
this entrepreneurial mindset, potential successors would be more willing to take-over a family business.
The result concerning the effect of age on the preferred form of succession rejects hypothesis 4. Firms with predecessors aged 50 years and older are less likely to appoint an intergenerational successor opposed to firm with predecessors aged younger than 50, however the result is not significant. The result rejects Davis’s (1968) findings, which found that predecessors between 50 and 60 years of age increasingly emphasize family philosophies and become less controlling. Research by Detienne and Cordon (2010) and Santarelli and Lotti (2005) confirmed this result. However, based on the dataset gathered for this meta-analysis the expected effect based on Davis (1968) is not confirmed. In fact, although not significant, the effect found is the opposed. It should be emphasized that the group coded “young” only entails two studies, lowering the explanatory power. A possible theoretical explanation comes from Guinanne’s (1992) research based on Irish data. Wherein he suggests that as a predecessor ages, the time to prepare a potential intergenerational successor decreases. As a result, a predecessor may be argue that the potential intergenerational successor is not sufficiently prepared to take-over the business and does not choose for intergenerational succession. Kihmi (1994) formulated another possible explanation. He suggests that intergenerational succession declines from a certain age. Finding an intergenerational successor becomes harder since potential successors start looking for alternative job opportunities. Glauben et al. (2004) confirms this by concluding that the likelihood of intergenerational succession increases up to 53 years age after which the likelihood decreases.
The result concerning the effect of gender on the preferred form of succession rejects hypothesis 5. Although the result shows the expected effect, it is not significant. Previous findings underlying the hypothesis cannot be confirmed. For example the influence of at-home behavior on the preferred form succession (Dumas, 1992). And the influence of parental concern and the desire to protect children on preferred succession (Hollander & Buckowitz, 1990). The finding that specific traits are associated with men and believed to be important for business success (Dawley et al., 2004) cannot be confirmed either. In the theoretical framework a shift towards the creation of pools of siblings sharing ownership was mentioned (Gersick, Lansberg & Dunn, 1999). In case this shift further developed might explain why no significant difference between the effects of gender on succession was found. This shift was confirmed by recent literature objecting Dawley et al.’s (2004) findings, namely the relationship between feminine traits and managerial effectiveness (Appelbaum et al., 2003; Eagly and Carli, 2007; Thomas et al., 2008).
encourages predecessors to make decisions increasingly based on the corporate logic, decreasingly taking family ties into account. In other words, more educated predecessors base their choice more objectively on abilities and skills of potential successors instead of appointing family members as successors.
The only significant result among the quality checks was the difference between the proportion of intergenerational succession among non-farming family businesses and farming family businesses. The result shows that family businesses operating in the agricultural sector are more likely to appoint an intergenerational successor. A possible explanation is given by Glauben et al. (2009), in which they suggest that intergenerational succession is more likely to happen in asset heavy farming businesses. This is consistent with Pesquin et al. (1999) suggesting to look at succession as a Nash bargaining game. From their point of view, having an intergenerational successor raises potential future farm income, which creates a surplus for the predecessor and the successor. In other words, an intergenerational successor is a more profitable option and therefore preferred.
7 Limitations and Future Research
This meta-analysis has a number of limitations. First of all, a broad definition of family business (Chua, Chrisman & Sharma, 1999) is used to select studies for this meta-analysis. The advantage of this choice is that as little studies as possible are excluded from the initial database. The downside of this choice is that authors in this literature field, who use a narrow definition, may judge the inclusion of some studies wrongly. Consequently, they could argue that this meta-analysis is not specifically focused on family businesses and does not sufficiently address the declared literature gap. However, taking into account the immatureness of this literature field, the broad definition was used. In the future, a similar meta-analysis using a more narrow definition would be interesting to determine whether different findings occur compared to this research.
Second, an issue occurring in every meta-analysis is the file drawer problem (Rosenthal, 1979). This issue occurs when journals are more likely to be publicized when significant results are reported, and as a consequence authors are more likely to submit research with significant results. The advantage in this study is that the initial German database mainly consists out of research never publicized. Therefore, the authors did not have to take into account the possible judgment of editors. A quality check was included to see if a significant difference between the initial German sample and the added publications occurred. However, no difference was detected. Therefore, no extra effort was made to find additional non-publicized research.
examined again more thoroughly, a different sample might occur. If a comparable study is to be done in the future, this is an issue to take into account. In case there is enough capacity within a research group, a more thorough selection procedure is recommended.
Fourth, the European definition was used to code firm size. However, not all publications are based upon European data. Research coded large in this meta-analysis might be differently coded with the use of another definition. As a consequence, different results might be found. For example, the North American definition used by the SBA Office of Advocacy includes all enterprises with fewer than 500 employees. Using this definition would result in a different result.
Fifth, a small number of publications specifically focus on certain industries or sectors. At least, it is often not reported in publications. This makes it difficult to study this factor in analysis. In the future it is recommended to gather and report data concerning this issue. So that, meta-analytical studies in the future can cope with this issue.
Finally, in the introduction two important advantages were put forward for intergenerational succession within family businesses, namely: coping with globalizing markets (Fernàndez & Nieto, 2005) and innovation (Zahra, 2005). After performing the meta-analysis two antecedents appear to be significantly influencing intergenerational succession. Based on the findings of this report, it would be interesting to research whether culture and predecessor’s education influence globalization and innovation, using intergenerational succession as a mediating variable.
8 Practical and Theoretical Implications
The findings of this research have several implications. First, the practical implications will be discussed. In many countries special institutes exist aimed at supporting family businesses. Those institutes offer networks and organize conferences and meetings where family businesses owners share and learn ideas. With the results of this meta-analysis more knowledge is created concerning antecedents affecting intergenerational succession. By sharing this knowledge, family business institutes can create awareness about among family business owners. As a consequence, family business owners may be able to take precaution when making succession related decisions.
Furthermore, governments may use the outcomes in shaping regulations and creating a favorable environment for family businesses. As previously discussed, intergenerational succession is less likely in collectivistic countries. This may cause competitive difficulties compared to individualistic countries. As mentioned in the introduction, involving more generations into a family business leads to an improved innovative environment. For example, countries may shape favorable transfer conditions to stimulate intergenerational succession.
regarding succession. The choice can be made to focus on specific parts if the succession procedure. In collectivistic countries predecessors can stimulate entrepreneurial behavior of potential successors. And highly educated predecessors can start well on time with assessing the requirements for potential successors and the associated training and education.
Besides the above-mentioned practical implications, the findings of this research have some theoretical implications. One part of the objective is to determine the most common or preferred form of succession. After combining the various samples, the proportion of intergenerational succession appears to be the largest. The hypothesis, based on the transaction cost theory (Williamson, 1989) and agency theory (Jensen & Meckling, 1976) is confirmed based on global data. Previously, datasets were more narrowly focused on smaller regions. In this sense, the contribution of this research is that previous findings are found to be generalizable across countries. However, caution should taken in generalizing since this sample still does not cover all regions.
The other contributions this research offers are related to the antecedents researched. Five different antecedents prominent apparent in the literature were identified and researched. Remarkable is that only one of the five hypotheses, education, was found to be correct. Culture was found to have a significant influence opposed to the hypothesized. The other three antecedents, firm size, age predecessor and gender successor, did not show a significant effect. An explanation for not finding significant results concerning three of the five hypotheses may be that those antecedents are not globally generalizable. The antecedents previously described in the literature may hold for specific regions, but they possibly may not be of any influence in different regions. However, caution must be taken in adopting the statement because the number of studies included in this meta-analysis is for some antecedents very limited. Furthermore, the global distribution of the studies included is not proportional.
9 Summary
Appendices
Appendix 1
Overview of studies included in the meta-analysis
Author Year Source
Ambrose 1983 Journal of Small Business
Management
Bennedsen & Nielsen 2008 Centre for Economic and
Business Research
Bennedsen, Nielsen, Perez-Gonzalez & Wolfenzon 2007 The quarterly Journal of Economics
Berlemann, Engelmann, Lessmann & Schmalholz* 2006 Mittelstandsbericht 2005/2006
Bocatto, Gispert & Rialp 2010 Journal of Small Business
Management
Chung & Luo 2013 Strategic Management
Journal
Cucculelli & Micucci 2008 Journal of Corporate
Finance
Doblhammer, Salzmann & Westphal* 2009 Projektbericht im Auftracht der Handwerkskammer Schwerin
Freund * 2000 Deutsche Universtitätsverlag
Freund * 2002 Schriften zur
Mittelstandsforschung
Freund & Kayser * 2007 Schriften zur
Mittelstandsforschung
Freund, Kayser & Schröer * 1996 Institut für
Mittelstandsforschung
Glauben, Petrick, Tietje & Weiss 2009 Applied Economics
Gottschalk. Höwer, Licht, Niefert, Hauer, Keese &
Woywode * 2010 Institut für
Mittelstandsforschung
Kayser & Freund * 2004 Handelskammer Hamburg
des BDI, der Ernst & Young und der IKB Deutsche Industriebank AG
Kerkhoff, Ballarini & Keese * 2004 Institut für
Mittelstandsforschung der Universität Mannheim
Kimhi & Nachlieli 2001 Journal of Agricultural
Economics
Lamsfuss & Wallau * 2012 Ergebnisse der
Frühjahrsbefragung 2012
Müller * 1997 Göttinger
Handwerkswirtschaftliche Studien Band 50
Müller, Kay, Felden, Moog, Lehmann, Suprinovic, Meyer, Mirabella, Boerger, Welke & Coritnaia 1 *
2011 Göttinger
Handwerkswirtschaftliche Studien Band 83
Müller, Kay, Felden, Moog, Lehmann, Suprinovic, Meyer, Mirabella, Boerger, Welke & Coritnaia 2 *
2011 Göttinger
Handwerkswirtschaftliche Studien Band 78
Müller, Koschmieder, Trombska, Zapfe & Rötzler * 2009 Empirische Personal- und Organisationsforschung Band 38
Pérez-González 2006 The American Economic
Review
Potter & Lobley 1996 European Society for Rural
Sociology
Regionomica * 2000 Gutachten im Auftrag des
Ministeriums für Wirtschaft des Landes Brandenburg
Ryan & Power 1 2012 Irish Journal of Management
Ryan & Power 2 2012 Irish Journal of Management
Sifei 2013 Beijng Foreign Studies
Stiglbauer & Weiss 2000 Department of Food
Studies
Uebe-Emden * 2011 Anforderingen und
Herausforderungen unter besonderer Berücksichtigung potenzieller Strukturbrüche der Unternehmensnachfolge
Wiklund, Nordqvist, Hellerstedt & Bird 2013 Entrepreneurship Theory and Practice
Zacher, Schmitt & Gielink 2012 Cambridge Journals
* Studies already present in the initial German sample
References
Allen, I.E. & Langowitz, N.S. (2003). Women in family-owned businesses. Boston: MassMutual Financial Group and Center for Women’s Leadership.
Ambrose, D. M. (1983). Transfer of the family-owned business. Journal of Small Business Management, 21(1): 49–56.
Appelbaum, S., Audet, L. & Miller, J. (2003). Gender and leadership? Leadership and gender? A journey through the landscape of theories. Leadership and Organization Development Journal, 24(1): 43-51.
Astrachan, J. H. & Shanker, M. C. (2003). Family businesses’ contribution to the U.S. economy: A closer look. Family Business Review, 16(3): 211–219.
Barach, J.A., Gatinsky, J. & Doochin, B.A. (1988). Entry of the next generation: Strategic challenge for the family business. Journal of Small Business Management, 26(2): 49-56.
Battisti, M. & Okamuro, H. (2010). Selling, passing on or closing? Determinants of entrepreneurial intentions on exit modes. Mimeo.
Begin, L. (2007). Motivations et freins a la reprise de l’entreprise familiale en Suise romande. Economies et Societes, 41(1): 11-36.
Bennedsen, M. & Nielsen, K. (2008). Vækst via generationsskifte. Centre of Economic and Business Research, September: 1-79.
Bennedsen, M., Nielsen, K., Perez-Gonzalez, F. & Wolfenzon, D. (2007). Inside the Family Firm: The Role of Families in Succession Decisions and Performance. Quarterly Journal of Economics, 122(2): 647-691.
Berrone, P., Cruz, C. & Gomez-Mejia, L.R. (2012). Socioemotional wealth in family firms: Theoretical dimensions, assessment approaches and agenda for future research. Family Business Review, 25(3): 258-279.
Birley, S.J. (1986). Succession in the family firm: The inheritor’s view. Journal of Small Business Management, 24(3): 36-43.
Birley, S. (2001). Owner-manager attitudes to family and business issues: A 16 country study. Entrepreneurship Theory and Practice, 26(2): 63–76.
Bocatto, E., Gispert, G. & Rialp, J. (2010). Family-owned business succession: The influence of pre-performance in the nomination of family and nonfamily members: Evidence from Spanish firms. Journal of Small Business Management, 48(4): 497-523.
Chua, J.H., Chrisman, J.J. & Sharma, P. (1999). Defining the family business by behaviour. Entrepreneurship Theory and Practice, 23(4): 19-37.
Chung, C. & Luo, X.R. (2012). Leadership succession and firm performance in an emerging economy: Successor origin, relational embeddedness and legitimacy. Strategic Management Journal, 34(3): 338-357.
Commission Recommendation. (European Commission, 2003). Covering the area concerning the definition of micro, small and medium-sized enterprises.
Cucculelli, M. & Micucci, G. (2008). Family succession and firm performance: Evidence from Italian family firms. Journal of Corporate Finance, 14(1): 17-31.
Davis, S. (1968). Entrepreneurial succession. Administrative Science Quarterly, 13(3): 402–416. Davis, P. & Harveston, P. (1998). The influence of family on the family business succession process: A multi-generational perspective. Entrepreneurship, Theory and Practice, 22(3): 31-53.
Dawley, D., Hoffman, J. & Smith, A. (2004). Leader succession: Does gender matter? The Leadership and Organization Development, 25(8): 678-690.
DeTienne, D. & Cardon, M. (2010). Impact of founder experience on exit intentions. Small Business Economics, 38(4): 351-374.
Doblhammer, G, Slazmann, T. & Westphal, C. (2009). Prognose über die entwicklung des übergabe- und nachfolgepotenzials von handwerksunternehmen in Mecklenburg-Vorprommern vor dem hintergrund der regionalen bevökerungsenticklung bis zum jahr 2020. Technical report for: handwerkskammer schwerin in cooperation mit der handswerkskammer Ostmecklenburg-Vorpommern.
Dumas, C. (1992). Integrating the daughter into family business management. Entrepreneurship Theory and Practice, 16(4): 41–55.
Dyer, W.G. & Handler, W. (1994). Entrepreneurship and family business: Exploring the connection. Entrepreneurship Theory and Practice, 19(1), 71-83.
Eagly, A. & Carli, L. (2007). Women and the labyrinth of leadership. Harvard Business Review, 85(9): 62-71.
Fadhel, M. (2004). Gulf family businesses urged to find investors. Retrieved July 18, 2014, from: http://www.middle-east-online.com/english/business/.
Fairclough, S. & Micelotta, E.R. (2013). Beyond the family firm: Reasoning the influence of the family institutional logic across organizations. Research in the Sociology of Organizations, 39(B): 63-98.
Fama, E. & M. C. Jensen. (1983a). Agency problems and residual claims. Journal of Law and Economics, 26(2): 325-344.
Fama, E. & M. C. Jensen. (1983b). Separation of ownership and control. Journal of Law and Economics, 26(2): 301-325.
Fernández, Z. &, M. J. (2005). Internationalization strategy of small and medium-sized family businesses: Some influential factors. Family Business Review, 18(1): 77–89.
family-Freund, W. (2000). Familieninterne unternehmensnachgefolge: erfolgs- und risikofaktoren. Wiesbaden.
Freund, W. (2002). Frauen in der unternehmensnachgefolge. In: Institüt für Mittelstandsforschung Bonn: Jahrbuch zur Mittelstandsforschung 1/2004, schriften zur Mittelstandsforschung Nr. 91, Wiesbaden.
Freund, W. & Kayser, G. (2007). Unternehmensnachfolgen in Bayern. Technical report for: des Bayerischen staatsministeriums für wirtschaft, infrastruktur, verkehr und technologie. In: Institüt für Mittelstandsforschung Nr. 173, Bonn.
Freund, W., Kayser, G. & Schroër, E. (1995). Generationwechsel im mittelstand: Unternehmensübertragungen und übernahmen 1995 bis 2000. In: Institüt für Mittelstandsforschung Nr. 109, Bonn.
Ganter, M., Kammerlander, N. & Zellweger, T. (2014). The incumbent’s dilemma when exiting the firm: torn between the family and the corporate logic. Academy of Management Journal Annual Meeting Proceedings, 2014(1): 965-971.
Georgas, J., Christakopoulou, S., Poortinga, Y.H., Angleitner, A., Goodwin, R. & Charalambous, N. (1997). The relationship of family bonds to family structure and function across cultures. Journal of Cross-Cultural Psychology, 28(3): 303-320.
Gersick, K. E., Davis, J. A., Hampton, M. M. & Lansberg, I. (1997). Generation to generation: Life cycles of the family business. Cambridge, MA: Harvard Business School Press.
Glauben, T., Tietje, H. & Weiss, C.R. (2004). Intergenerational succession in farm households: Evidence from upper Austria. Review of the Economics of the Household, 2(4): 443-461.
Glauben, T., Petrick, M, Tietje, H. & Weiss, C. (2009). Probability and timing of succession or closure in family firms: A switching regression analysis of farm households in Germany. Applied economics, 41(1): 45-54.
Goldberg, S. & Wooldridge, B. (1993). Self-confidence and managerial autonomy: Successor characteristics critical to succession in family firms. Family Business Journal, 6(1): 55-72.
Gottschalk, S., Höwer, D., Licht, G., Niefert, M., Hauer, A., Keese, D. & Woywode, M. (2010). Generationenwechsel im mittelstand: Herausforderungen un erfolgsfaktoren aus der perspective der nachfolger. In association with: Verband der Vereine Creditreform, Baden-Württembergische Bank, Stuttgart.
Guinanne, T.W. (1992). Intergenerational transfers, emigration and the rural Irish household system. Explorations in Economic History, 29(4): 456-476.
Handler, W. (1992). The succession experience of the next generation. Family Business Review, 5(3): 283-307.
Haunschild, L. & Wolter, H.J. (2010). Volkswirtschaftliche bedeutung von familien- und frauenunternehmen. In: Institut fur Mittelstandsforschung Bonn.
Hedges, L. V. (1982). Estimating effect size from a series of independent experiments. Psychological Bulletin, 92(2): 490-499.
Hedges, L.V. & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, FL: Academic Press. Hofstede, G. (1983). National culture in four dimensions. International Studies of Manegement and
Organization, 13(2): 46-74.
Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions and organizations across nations. 2nd Edition, Thousand Oaks CA: Sage Publications.
Hofstede, G. (n.d.) What about China? Retrieved October 24, 2014, from: http://geert-hofstede.com/china.html.
Hollander, B. & Bukowitz, W. (1990). Women, family culture and family business. Family Business Review, 3(2): 139-151.
Hunter, J.E. & Schmidt, F.L. (1990). Methods for meta-analysis: Correcting error and bias in research findings. Newbury Park, CA: Sage.
Iannarelli, C. (1992). The Socialization of Leaders: A Study of Gender in Family Business. Unpublished Doctoral Dissertation, University Pittsburgh.
Jensen, M.C. & Meckling, W.H. (1976.) Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3(4): 305-360.
Kayser, G, & Freund, W. (2004). Unternehmensnachgefolge in der Hansestadt Hamburg. In: Handelskammer Hamburg: Unternehmensnachgefolge in Hamburg. Technical report for: Handlungsnotwendigkeiten und Begründungen, Hamburg.
Kayser, G., Wallau, F. & Adenäuer, C. (2006). BDI-mittelstandspanel: Ergebnisse der online-mittelstandsbefragung, Frühjahr 2006. Technical report for: BDI, Ernst & Young and IKB Deutsche Industriebank AG, Berlin, Düsseldorf and Bonn.
Kerkhoff, E., Ballarini, K. & Keese, D. (2004). Generationwechsel in mittelstandischen Unternehmen. In: Institut für Mittelstandsforschung der Universität Mannheim. Technical report for: Mittelstandforschung Band 7, Karlsruhe.
Kets de Vries, M. (1993). The dynamics of family controlled firms: The good news and the bad news. Organizational Dynamics, 21(3): 59-68.
Kets de Vries, M. (1996). Family businesses: Human dilemmas in the family firm. Cornwall: International Thomson Business Press.
Khimi, A. (2000). Optimal timing of farm transferal from parent to child. American journal of Agricultural Economics, 76(2): 228-236.
Kihmi, A. & Nachlieli, N. (2001). Intergenrational succession on Israeli family farms. Journal of Agricultural Economics, 52(2): 42-58.
Le Breton-Miller, I., Miller, D. & Steier, L.P. (2004) Toward an integrative model of effective FOB succession. Entrepreneurship Theory and Practice, 28(4): 305-328.
Lamsfuß, C. & Wallau, F. (2012). Die größten familienunternehmen in Deautschland: Daten, fakten, potenziale und ergebnisse der Früjahrsbefragung 2012. Technical report for: der Deutsche Bank AG und des Bundesverband des Deutschen industrie, Berlin and Frankfurt.
Miller, D., Le Breton-Miller, I. & Lester, R.H. (2011). Family and lone founder ownership and strategic behavior: Social context, identity and institutional logics. Journal of Management Studies, 48(1): 1-25.
Morris, M. & Schindehutte, M. (2005). Entrepreneurial values and the ethnic enterprise: An examination of six subcultures. Journal of Business Management, 43(5): 453-479.
Müller, K. (1997). Generationwechsel im handwerk: Eine untersuchung über das niedersächsische handwerk. In: Göttinger Handwerkschriftschäftliche Studien Band 50, Göttingen.
Müller, K., Kar, R., Felden, B., Moog, P., Lehman, S., Suprinovic, O., Meyer, S., Mirabella, D., Boerger, S., Welge, B. & Coritnaia, I. (2011). Der generationswechsel im mittelstand im demografischen wandel. In: Göttinger Handwerkswirtschaftliche Studien Band 78, Duderstadt.
Müller, K., Koschmieder, K.D., Trombska, D., Zapfe, A. & Rötzler, K. (2009). Unternehmensnachgefolge im Thüringerhandwerk: Eine analyse im zeichen des demografischen wandel. In: Göttinger Handwerkswirtschaftliche Studien Band 78, Duderstadt.
Pérez-González, F. (2006). Inherited control and firm performance. The American Economic Review, 96(5): 1559-1588.
Pesquin, C., Kimhi, A. & Kislev, Y. (1999). Old age security and intergenerational transfer of family farms. European Review of Agricultural Economics, 26(1): 19-37.
Picot, A., Dietl, H. M. & Franck, E. (2005). Organisation: Eine ökonomische Perspektive, 4th ed. Stuttgart: Schäffer-Poeschel Verlag.
