PAST ECONOMIC PERFORMANCE AS A DRIVER OF CEO SATISFACTION IN SMALL FAMILY FIRMS

Transcription

1 PAST ECONOMIC PERFORMANCE AS A DRIVER OF CEO SATISFACTION IN SMALL FAMILY FIRMS 1. Introduction The entrepreneurship literature emphasizes that understanding entrepreneur s satisfaction with the firm is relevant since satisfaction determines, to a great extent, the decision to continue and invest more time and resources in the firm or to exit the company (Cooper and Artz, 1995; Gimeno et al, 1997; Delgado-Garcia et al., 2012). Among these studies, several authors point to the need of going beyond firm economic performance to understand entrepreneurial satisfaction, highlighting the role played by entrepreneur s noneconomic goals (Cooper and Artz, 1995; Van Gelder et al, 2007; Khelil, 2016). According to Cooper and Artz (1995), given that entrepreneurs differ in their goals (economic vs noneconomic), they may react differently to a given level of economic performance. Specifically, for a given level of economic performance, economically driven entrepreneurs would perceive a higher gap and thus would be less satisfied with the business than non-economically driven entrepreneurs. The interplay between economic and noneconomic goals in determining individual s satisfaction with the firm is particularly relevant in the context of family businesses, given the importance that family owners concede to socioemotional wealth (SEW), that represents the affective endowment derived from family control of the firm (Gomez-Mejia et al, 2007). Extant studies following the SEW approach show how family firm outcomes depend on the existing gap between family owner s goals and firm performance (Gomez-Mejia et al, 2010; Chrisman and Patel, 2012). However to our knowledge, satisfaction in the context of

2 family firms has been restricted to particular aspects such as satisfaction with the job (Daily and Near, 1999; Block et al, 2012) or with the succession process (DiMassis et al, 2012). As such, although the importance of noneconomic goals in family firms is widely accepted, how the existing gap between financial performance and individuals goals and aspirations affect personal satisfaction with the progress of the family firm is still unknown. This neglect is important, given the deterministic relationship between satisfaction with the firm and firm survival and growth (Cooper and Artz, 1995). Our paper represents one step in this direction by comparing the impact of past economic performance on CEO s personal satisfaction in small family and non-family firms. Our reasons to choose this particular context are twofold: First, extant evidence shows the central role played by owners and CEOs preferences in small firms decision making process (Brockmann and Simmonds, 1997; Jennings and Beaver, 1997; Powell and Eddleston, 2013). Hence, understanding CEO levels of satisfaction with the firm is important to comprehend key strategic decision in small firms (Cooper and Artz, 1995; Gilmore, Carson & O'Donnell, 2004: Fama and Jensen, 1983). Second, the importance of owner s noneconomic goals seem to be especially salient in this context (Daily and Dalton, 1992). Focusing on the CEO as a unit of analysis allows us to explore how family ownership and family management (and thus the presence of family noneconomic goals) influence the performance-satisfaction relationship among CEOs. We draw on Gimeno s et al work (1997) who argue that when entrepreneurs are motivated by noneconomic goals, they obtain a greater psychic income (p. 758), and thus are willing to accept lower economic returns when they gain personal satisfaction from the business. Based on that, and building on the SEW approach we conjecture that since the majority of

3 controlling families emphasize SEW protection goals over pure economic goals (Arregle et al., 2007; Gómez-Mejía et al., 2007; Berrone et al., 2010; Mahto et al., 2010), the impact of firm economic performance on CEO s personal satisfaction with the firm will be weaker in family than in non-family firms. Moreover, following calls to study behavioral differences among family firms (Chua et al, 2012), we analyze how two of the main sources of heterogeneity within family firms, namely the family or non-family membership of the CEO and the generational stage (Casillas et al, 2010; Cruz and Norqvist, 2012), impact the relationship between past performance and CEO s personal satisaction with the firm in the case of family owned firms. Our work makes several contributions to the literature. First, we bring insights on the determinants of entrepreneur s satisfaction with the firm (e.g. Cooper and Artz, 1995; Gimeno et al, 1997) to the context of family controlled firms, and develop a framework to understand the level of satisfaction with the firm in these companies. In this matter, although entrepreneurial studies frequently highlight the importance of noneconomic goals for entrepreneurs, existing research on owner s satisfaction has focused essentially on satisfaction with firm economic performance (Mahto et al, 2010). By contrast, we do not restrict ourselves to economic performance but focus on general satisfaction with the firm, and therefore our measure reflects satisfaction with both economic and noneconomic aspects. In this vein, the study adds to the growing literature on family firms that looks at how SEW preservation motives drive the perceptions of family firms. Second, we contribute to the scarce literature in the determinants of personal satisfaction with the business for CEOs. Research on satisfaction is done primarily from the perspective of employees (Saari and Judge, 2004) or entrepreneurs (Cooper and Artz, 1995; Carree and

4 Verheul, 2012), with little attention to firm managers. Although the influence of owners on strategic decisions can be strong, CEOs have a direct influence on firm s strategies, and above all in the context of small firms. So understanding CEOs satisfaction with the firm is crucial. Doing so in family business contexts is likely to be of great interest to family business and management scholars alike, given the pervasiveness of this ownership form. Third, distinguishing between family and non-family CEO, our study also contributes to enlarge research on nonfamily CEOs in family firms which remains surprisingly scarce (Astrachan et al., 2002; Chua et al., 2003; Poza, 2004). It also contributes to one of the central topics in the literature of family firms these days: the understanding of the heterogeneity of family firms. Our empirically supported theory proposes that the impact of past performance on CEO s satisfaction with the progress of the family firm will be dependent on whether or not the CEO belongs to the family. It will also be dependent upon the family firm s generational stage. Lastly, our research context offers a unique opportunity to expand existing research on the importance of noneconomic goals in small family business. Most of previous research has been conducted on large listed firms, and has not addressed small firms, whose context may be significantly different (Anderson and Reeb, 2003; Villalonga and Amit, 2006). Analyzing a sample of small family firms we are able to explore out in this study the specificities of privately held firms, which have not been studied with much detail. The paper is structured as follows. The next section presents the theoretical framework and the resulting hypotheses. Section three, contains information about the database and the variables employed to test the hypotheses of the study. Section four presents the results of

5 the empirical analysis. The final section of the study is devoted to the discussion. 2. Theoretical Background The importance of understanding CEO s satisfaction in the context of small firms. The entrepreneurship literature portrays that satisfaction with business performance is a fundamental measure of owner s success (Delgado-Garcia et al, 2012; Simon, Houghton and Savelli, 2003). Extant research emphasized the role of entrepreneur s satisfaction with the firm as a crucial predictor of the entrepreneur s decision about whether to persist or exit, or whether to invest more money and time in the firm (Cooper and Art, 1995; Westhead, Ucbasaran and Wright, 2005). This is particularly true in the case of small and medium sized companies, where the influence of founders and owners may be strong (Powell and Eddleston, 2013). The concept of entrepreneurial satisfaction was first introduced by Cooper and Arzt (1995) as fundamental measure of success for the individual entrepreneur (p 440). Following the emotive approach to business failure (Van Gelder et el, 2007; Shepherd et al, 2009; Khelil, 2016), the authors assign a crucial role to entrepreneur s motivation to explain why some entrepreneurs decide to give up performing firms while others accept to persist despite leading an underperforming firm. Borrowing from existing studies in organizational psychology on the determinants of job satisfaction (i.e. Michalos, 1986), they suggested that entrepreneur s satisfaction is: a) referent dependent, meaning that it is determined by the perceived gap between entrepreneur s goals and actual firm performance and b) contingent upon the balance between entrepreneur s economic and noneconomic goals (Cooper and Art, 1995). Given that entrepreneurs differ in their balance between economic and noneconomic

6 goals, entrepreneurs may react differently to a given level of economic performance. Accordingly, they predict that the sensitivity of satisfaction to business financial performance would be much stronger for economically motivated owners (who will perceive a higher gap between goals and actual performance) than for those emphasizing noneconomic goals (which satisfaction would be much more attributable to noneconomic rewards). Likewise Gimeno et al (1997) posit that entrepreneurs motivated by noneconomic goals obtain greater psychic income, and that consequently are keen to accept lower economic returns. Thus, entrepreneur s satisfaction is contingent upon the balance of economic and noneconomic goals, with the weight given by the entrepreneur to the later reducing the importance of the former on entrepreneur s judgement of satisfaction. Despite important contributions, the above mentioned research is focused on understanding small firm owners and entrepreneur s satisfaction, with little or no attention to firm s managers and particularly CEOs, who are responsible of and accountable for the central strategic decisions made by firms. While it is true that many small firms are led by its founder, this is not always the case since many others could be run by a founder s relative or by professional managers with no ownership and/or family ties to the business. If this is the case, more research is needed to understand how different managers balance economic and noneconomic goals and ultimately, what determines their overall satisfaction with the firm. This need is even greater in the case of small family owned firms given the importanceof noneconomic goals for family members. This is why in the next section we examine how past performance affect CEO s satisfaction in family and non-family firms. 2.1 CEO satisfaction in small family and non-family firms

7 The role of noneconomic rewards is especially relevant in the context of family firms, in which the business is not only a source of income for family owners, but also a framework for family employment, family pride and family identity (Zellweger, Nason, Nordqvist, & Brush, 2013). Indeed, the contention that family firms are motivated by more than the monetary outcome of organizational activity is one of the most prominent observations in the family firm literature (Gomez-Mejia et al., 2007; Berrone et al., 2012). This is particularly true in the case of small firms, in which there is a strong connection between the organization s identity and the founding family s identity (Cruz, Justo and De Castro, 2012). These findings suggest that the utility of family owners is a function of the extent to which economic and noneconomic goals are valued and achieved. Family business scholars have termed the concept of SEW as an umbrella concept that accommodates all noneconomic elements of a family's utility function that directly relate to the family's involvement in the firm (Gomez-Mejia et al., 2007). According to the SEW approach, preservation of the family socioemotional endowment is the primary reference point for family owners. This is to say, family firms would make decisions in such way that they preserve their accumulated socioemotional endowment (what is considered important for personal welfare that is already accrued and can be counted on) within the firm and consequently, family owners will exhibit a heightened concern for noneconomic goals (ultimately protecting the family's own SEW). If this is the case, we would expect that the assessment of the CEO s satisfaction with the progress of the family firm will be determined not only by past financial performance, but also by how the firm is contributing to the family owners SEW (i.e. whether the family SEW is maintained or improved, or it has diminished). In other words, much of the CEO s satisfaction with the progress of the firm may be attributable to the achievement of

8 noneconomic outcomes. Any business with a long term view needs executives with a profile matching the business culture, organization, and strategy (Gallo, 1991; Welch, 2005). This holds especially true for family firms, since they tend to have a specific and distinct business culture (Denison et al., 2004). Thus, as compared with non-family contexts where the CEO relationship with the firm is often more utilitarian, distant and transient (Lubatkin, Schulze, Ling, & Dino, 2005) and where owners face a diversified asset position (Fama, 1980), we would expect that the impact of past performance on CEO s evaluation of satisfaction with the business will be less significant in the case of family firms. Our rationale does not imply that CEO in family firms ignore financial outcomes but that acknowledging the partial trade-offs between economic and socioemotional wealth considerations (Gedajlovic, Carney, Chrisman, & Kellermanns, 2011), they would need to balance the two forces and seek a sustainable mix of satisfactory economic and socioemotional wealth creation. As a result of this balancing effort, we anticipate that the sensitivity of CEO s satisfaction to financial performance would be weaker for CEOs running family firms. Formally stated: Hypothesis 1: The impact of past financial performance on overall satisfaction with the firm of CEOs in family firms will be lower than that of CEOs in non-family firms Family vs Non family CEO s satisfaction In family firms, the desire to appoint family members as CEOs is well known (Karra, Tracey and Philips, 2006). However, they may also opt for appointing a non-family CEO to prevent managerial entrenchment (Gomez-Mejia et al, 2001), and/or to bring more objectivity to the decision-making process (Blumentritt, Keyt, & Astrachan, 2007; Huybrechts, Voordeckers

9 and Lybaert, 2013). Yet the previous hypothesis made no difference between family and nonfamily CEOs satisfaction. Implicit in our reasoning was the idea that family owner s balance between economic and noneconomic goals permeates to the organization, influencing CEO s definition of success regardless of their family ties (Cruz et al, 2010; Berrone et al, 2010) particularly when compared with CEOs managing non-family firms. Despite the familiarization process for non-family CEOs working in family firms, they often have views and assumptions of the world that differ from those of family CEOs. These different expectations can be traced to organizational and occupational socialization experiences (Van Maanen & Schein, 1979; Dyer Jr., 1989; Hofmann, 1991; Le Breton-Miller et al., 2004). Family CEOs do not just learn skills and practices that tend to be idiosyncratic to their particular family business, but they are taught to adhere to the family s values and to respect the role of the family (Dyer, 1989). Hence, the family CEO is more likely to attend to the goals of the controlling family as a means of enhancing the family s noneconomic utilities and/or preventing SEW losses (Berrone et al, 2010). Moreover, evidence shows that the typical CEO of a family firm holds tenure four times longer than that of a non-family CEO (McConaughy, 2000; Gómez-Mejía et al., 2003). This longer term horizon provides them with greater job security and makes them less sensitive to economic outcomes than nonfamily CEOs, who are generally held more accountable for firm performance (Gomez-Mejia et al., 2001). In this vein, recent research suggest that non family CEOs would need to perform at a higher level than family CEOs in order to produce a net income that compensates for the loss of socioemotional wealth incurred from hiring outside the family. Accordingly, family owners would set higher performance expectations for nonfamily managers than for family managers (Chrisman, Memili and Mishra, 2014).

10 Hence, although in hypothesis 1 we argued that CEOs in family firms should look for a sustainable mix of satisfactory economic and socioemotional wealth creation, we now anticipate that because non-family CEOs wealth is more likely to be contingent on higher financial goals (Gomez-Mejia et al., 2001; DeTienne and Chirico, 2013), the balance between SEW and economic outcomes leans more to economic factors when the CEO is not a family member. Consequently, in the case of small family firms, non-family CEO s would give a greater weigh to past financial results when assessing his overall satisfaction with the firm. Formally stated: Hypothesis 2: The impact of past financial performance on overall satisfaction with the firm in family firm will be higher for non- family CEOs than for family CEOs CEO satisfaction in family firm and generational stage Several family firms scholars have addressed the impact of generation in control on family firm outcomes (Casillas et al, 2010; Cruz and Norqvist, 2010; Miller et al 2013; Giovannoni, Maraghini and Riccaboni, 2011). The underlying rationale is that generational stage influences family owners perception about the value of both economic and noneconomic goals (Gomez-Mejia et al, 2007; Björnberg and Nicholson, 2012). Compared to founder-managed firms, in which authority is highly centralized in the founder (Gedajlovic, Lubatkin, & Schulze, 2004), descendants of founders managing family firms face very different challenges (Cruz & Nordqvist, 2012). As the family expands, the number of family owners usually also increases leading to conflicts that arise from siblings competing values and higher diversity of personal goals (Ward, 1997). As a result, it is suggested that family s attachment to the organization is highest when the firm is owned and

11 managed by the founding family (founder lead family firm) and that it tends to weaken as the firm transitions into subsequent generations (descendent lead family firms) (Chua et al., 1999; Mishra & McConaughy, 1999; Schulze, Lubatkin, & Dino, 2003). Further, with the increase in passive family shareholders, more pressure is placed on short-term performance and the payment of dividends (Schulze, Lubatkin, & Dino, 2003). Following this reasoning, the SEW approach is projected on a generational perspective (Gómez-Mejía et al., 2007) emphasizing that losses in SEW should weigh less heavily on a family firm s willingness to give up control as it moves from a founding family-controlled to a firm that is owned by an extended family. Collectively, these studies suggest that the emphasis on preserving the family s socioemotional wealth lessens as the firm moves through generations and that financial considerations become more important as a frame of reference (Gomez-Mejia el al, 2011). It follows from the preceding arguments that generational stage would influence CEO s satisfaction with the firm. Regardless of his family ties, CEOs in descendent lead firms are aware of the diversity of owners interest. Specifically, they know that especially for those not directly involved in the business, ensuring financial performance is key to sustain their private lifestyles (Le-Breton-Miller and Miller, 2009). In contrast, a great emphasis on noneconomic goals coupled with higher ownership concentration in founder family firms, would make it easier for family owners to enforce the fulfillment of noneconomic goals as one of the main priorities of the firm. Hence we would expect that the impact of past financial performance on CEO satisfaction in descendent lead family firms should be higher than in the founder generation. Formally stated: Hypothesis 3: The impact of past financial performance on CEO s overall satisfaction with

12 the firm will be lower for founder lead family firms than for descendent lead family firms. 3. Methods 3.1.Data collection The hypotheses of the study are going to be tested on a unique representative sample of Spanish small firms in high and medium technology manufacturing and service industries. This setting is particularly suitable to test our predictions for several reasons. First, technology sectors are characterized by their higher degree of uncertainty and stiff competition (Makri and Scandura, 2010). Hence, in these sectors pressures for survival and economic performance are particularly salient. Second, small size firms cannot take advantage of scale economies and may experience difficulties accessing to financial resources and distribution channels (Coleman, 2002). Consequently, small firms face a constant struggle for reaching the minimum performance levels that can guarantee their survival. Third, given the substantial influence of owners on firm outcomes in the case of small enterprises (Jennings and Beaver, 1997), we would expect family owner s noneconomic goal to be especially salient in this context. Finally, the power and importance of the CEO in the decision making process is even greater than in bigger firms, in which CEO power can be balanced by the power and supervision of the firm s board of directors. Thus, the influence of the CEO s satisfaction with firm in firm level strategic decision is remarkably relevant. The population of Spanish small firms in these industries was initially identified using the SABI database, the most comprehensive dataset of incorporated firms in Spain. First high

13 and medium-high technology sectors (in both manufacturing and services industries) were identified using the classification of the Organization for Economic Co-operation and Development (OECD) and the Instituto Nacional de Estadística (INE). 1 Based on this industry classification we searched for firms between 10 and 50 employees whose primary or secondary activity code corresponded to one of those sectors. In addition we removed the few firms that were not incorporated businesses or limited partnerships (Wiklund et al., 2009), obtaining a total population of firms. A representative sample of 1500 firms was selected to guarantee industry and legal form representativeness (sampling error was ±2.34% with a confidence level of 95%). Firms were randomly selected within each industry segment for a phone interview using Computer-Assisted Telephone Interviewing (CATI) software. Interviews were conducted between November and December of 2010 by a firm specialized in market studies with a large experience conducting similar research oriented interviews. 97% of firms in the population were contacted and asked to participate agreed to participate and responded to the questionnaire (14.20% response rate). The survey was answered by the CEO of the firm, with interviews having an average duration of 27 minutes. Missing values reduce our effective sample to 1314 for multivariate analyses (12.43% effective response rate). We found no differences in terms of size or industry between those that participate and those that refused to do so. Primary data was obtained from the survey questionnaire answered by the CEOs during the 1 INE is the Spanish National Institute of Statistics. Sectors 21 (manufacturing of pharmaceutical products), 26 (manufacturing of optical and electronic devices) and 303 (manufacturing of aeronautic and aerospace machines and products) are manufacturing high-tech sectors. Sectors 20 (chemical industry), 254 (weapon and ammunition manufacturing), 27 (manufacturing of electric products), 28 (manufacturing of machines and equipment), 29 (car manufacturing), 302 (manufacturing of railway products), 304 (manufacturing of military vehicles), 309 (manufacturing of other transportation materials) and 325 (manufacturing of medical instruments and supplies) are the medium-high technology manufacturing sectors. Finally sectors 59 (image and music recording and editing), 60 (radio and television broadcast), 61 (telecommunication), 62 (software programming and consulting), 63 (information services) and 72 (research and development) are high-technology service sectors.

14 interviews. This was the core source of information to measure several key constructs of our model. This information was complemented with some secondary information obtained from the SABI data base achieving a unique and original collection of data. 3.2.Variables Dependent variable. Satisfaction captures the CEO s level of satisfaction with the progress of the company. Specifically, and in line with previous studies such as Cooper and Artz, (1995), we asked respondents to assess their "personal overall satisfaction with their business." The item was measured with a single item using a 5-point Likert scale (1=totally satisfied, 5= Not satisfied at all) (Block et al., 2005, Carree and Verheul, 2012). In contrast to multi-item approaches that rate satisfaction with different facets of the business (i.e. financial performance or sales), our measure request a single global assessment without identifying specific factors. Hence, we do not make ex-ante assumption about the dimensions or goals that are important to define the satisfaction levels of the CEOs, and therefore avoid potential mistakes caused by failing to identify relevant dimensions. In sum, our approach seems to be particularly suited to capture the influence on satisfaction of the noneconomic dimensions that go beyond the economic performance related aspects usually captured in multi-item approaches. Independent Variables Past Performance. As in previous research (e.g. Zahra, Hayton, and Salvato, 2004; Zahra, 2005) firm s past performance is measured as the mean of the operating results (in thousands) of the last three years prior to the survey (i.e ) divided by the number of the employees. It is therefore a size adjusted measure of firm s performance that depends directly

15 on how well the firm is managing its core activities (i.e. those directly related with the production and sale of its product and services). The information was obtained from the SABI data base. Family firm (Family Firm). Consistent with previous operationalization of family firms (e.g. Barth et al., 2005; Gomez-Mejia et al., 2001; Gomez-Mejia et al., 2010) and given the small size of the companies in our sample, we consider a firm as a family firm if the family controls, directly or indirectly, more than 50 percent of the shares and at least one family member is present on the board of the directors percent of the firms in our sample meet this definition of family firms. Further, in order to test hypothesis 2 and 3, we consider within the subsample of family firms first whether they are run by a member of the family (CEOF) or by professional CEO (CEO NF). Then, we also take into account the generation and distinguish between family firms in which the founder is still present in the management of the firm (founder), and family firms in which the founder is not present as owner or manager (descendent). Control variables. We control for the respondents demographic characteristics, specifically age, education and experience. This approach accounts for the view of Upper Echelons theorists where a close relationship exists between a person s demographic characteristics, her cognitive bases and value, and in turn her strategic preferences and dispositions (Hambrick & Mason, 1984; Wiersema & Bantel, 1992): Educational level (education level), is the highest educational level that the manager has reached. It is measured by a dummy variable that takes value 1 he or she has universities studies or higher (master or doctorate) and 0 otherwise. Experience,

16 is a continuous variable that captures the number of years of labor experience in the same industry sector. CEO age (CEO age) is a variable that measures the age of the CEO of the firm. Finally, following Cooper and Artz (1995) we control for CEO s expectations (expectation) regarding firm performance, by using a dummy variable that takes value 1 when he or she thinks that the sales will increase next year and 0 otherwise. Respondents from large organizations (e.g., Gómez-Mejía et al., 2003) or within growing industries (e.g., Schulze, Lubatkin, & Dino, 2003) are likely to have different perceptions of performance. Hence, we control for the firm characteristics and industry conditions. Firm size (Size) is approached by the number of employees of the firm. Information was gathered from responses to a question in the survey instrument. Specifically, it is measured as a dummy variable taking the value 1 when the firm has more than 20 employees, and 0, when the firm has between 10 to 20 employees. (Correa et al., 2007; OECD, 1995). Service is a dummy variable that takes value one when the firm belongs to a service sector and 0 when it belongs to an industry sector. Firm age (Firm age) is also common control variable in small firm research as it may capture differences in behavior and performance due to culture and generation issues. Firm age is computed as the difference between 2010, the year the survey was administered, and the year the firm was founded. Network is the result of the factor analysis where the respondent indicate how important each of the following groups: Consultants; Lawyers; Public agencies business support; Accounting; Banks; Family; Customers; Suppliers; Employees and Political Contacts, are as a source of ideas and advice when making important decisions in their company. All items were measured in a 5-point Likert scale. Under-resourced (U_R) is another variable that measures whether or not the availability of capital has been inadequate and a major impediment to successful business

17 development. We have also control for entrepreneurial orientation (EO) as the individuals with strong entrepreneurial orientations are willing to take on high-risk projects in exchange for potentially high returns and satisfaction at individual level (Miller, 1983). To approach the entrepreneurial orientation of the firm we employed 13 items previously proposed and employed by Covin and Slevin (1989) (risk taking, innovativeness and proactiveness), Lumpkin and Dess (1996) (autonomy) and Lumpkin and Dess (2001) (aggressiveness). As originally proposed all items were measured in a 7-point Likert scale. An exploratory factor analysis was conducted on the responses to these items. Consistent with previous research the factor analysis revealed the existence of one factor that explained 64.1% of total variance. The single factor represents EO as the average value of the 13 items. This additional measure ranges between 1 and 7. The greater the value of this variable the greater the entrepreneurial orientation of the firm. While having collected data from different sources may reduce concerns about a potential common method bias problem, we run, as an additional test, a single factor analysis on the survey instrument variables (Podsakoff et al., 2003). The results discard that all the variables have significant loadings in a single factor. That is, they reject the existence of a single factor capturing a significant portion of total variance. Thus, we can confirm that our estimations are free of common method bias Methodological approach

18 When a dependent variable has more than two categories and the values of each category have a meaningful sequential order where a value is indeed higher than the previous one, and the data is following a normal distribution, ordered probit is the most appropriate model to use. Given the nature of our dependent variable, we estimate a series of ordered probit models to test our hypotheses. More specifically, we use a split sample approach to compare the impact of past performance on CEO satisfaction in family and non-family firms (hypothesis 1), founder and descendant firms (hypothesis 2) and family and non-family CEOs in family firms (hypothesis 3). The split sample method is appropriate when theory predicts independent-dependent variables relations by subgroups: (family vs. non-family) or (founder vs. descendent) or (family CEO vs. non-family CEO). It makes sense when it is expected that the relationship between the dependent and independent variables, and not only the main independent, may be different between the two groups, because each model is somehow different. Besides, it has been extensively used in previous family business studies (Gómez-Mejía et al 2003; Berrone et al., 2010). We use robust standard errors in all our multivariate estimations to avoid concerns about heteroscedasticity. Finally, variance inflation factors indicate that our estimations are free of any multicollinearity problems (Hair, Black, Babin, and Anderson, 2009). 4. Results The descriptive statistics and correlations for the variables used in this study are reported in Table percent of the firms in our sample are family firms and the remaining percent are non-family. The mean value of satisfaction is 2.96 (on a scale from 1 to 5) while

19 the mean value of the age of the firms are years. A close look at the values shows that past financial performance is positively correlated with satisfaction, but any of them have a significant relationship with our core variable, family firm. Family firm has significant relationship with the size and the age of the firms, and also with the industrial sector. INSERT TABLE 1 ABOUT HERE Multivariate analyses are necessary to provide a more nuanced test of Hypotheses 1. Those analyses are shown in Table 2. It seems that entrepreneurial orientation has positive and significant impact on satisfaction. However, firm age, firm size, firm difficulties to access to resources and manager experience have negative and significant impact on CEO s satisfaction. This indicates, consistent with Miller s (1983) view, that in firms that rate high on EO CEOs will feel more satisfied. However, in older firms, bigger firms and firm with difficulties to access to resources will lead CEOs to experience to lower satisfaction levels. Interestingly, although we could expect that CEOs with higher initial expectations would later be less satisfied (because they would have a greater expectation-performance gap), the opposite was found; those who were more optimistic initially were more satisfied later, even when controlling for performance. This result, however, is consistent with the findings reported by Cooper and Art (1995). Importantly, we run a t-test analysis in order to know if there are differences in the mean levels of satisfaction experienced by CEOs in family and nonfamily firms. Results reveal that no mean differences exist. Thus, CEOs in family firms are not more or less satisfied than their counterparts in non-family firms. INSERT TABLE 2 ABOUT HERE The first and second columns of Table 2 show the results of the ordered probit estimated to

20 test hypothesis 1. As can be seen, there is a positive and significant impact of past financial performance on CEOs satisfaction in non-family firms. However, this relationship is not significant in the sample of family firms, which supports our first hypothesis. That is, this result supports the idea that because of the relevance that noneconomic goals have in family firms, satisfaction levels, and particularly those of CEOs, in these companies are less influenced by past financial performance This effect does not consider whether the CEO is or not a member of the controlling family. We analyze this issue next, by comparing the influence of past financial performance on family and non-family CEOs satisfaction with the firm. Columns three to six in Table 2 summarize the results of ordered probit estimated to test hypotheses 2 and 3. These hypotheses advance differences in the effect of past economic performance on CEO satisfaction levels within family firms. The third and fourth columns of Table 2 show that the impact of past financial performance on satisfaction in family and non-family CEOs are not significant. Thus, hypothesis 2 is not supported. Apparently, not only family members might act as stewards (Davis, Schoorman and Donaldson, 1997) and regard firm performance as an extension of their own well-being, as stated by Anderson and Reeb (2003), non-family executives might also behave similarly. Given that this result may be at odds with previous evidence relating the presence of nonfamily CEOs with a mitigation of family s emotional considerations (Bhattacharya & Ravikumar, 2004; Poza et al., 1997; Gómez, 2004), we ran a post hoc analysis in which we divide the sample into negative and positive past performance. Results (available upon request) show that differences between family and non-family CEOs in family firms exist only when firm performance is negative. That is, when the family firm is facing negative

21 performance the impact of past financial performance on satisfaction is significant, but only for the case of non-family CEOs. This evidence adds an interesting twist as show that under positive performance non-family CEOs may play similar attention to noneconomic facets of the firm and consequently the dependence of their satisfaction on economic performance is similarly to those of family CEOs. Columns five and six in Table 2 summarize the results of the estimations ran to explore the differences in the effect of past financial performance on CEO satisfaction with the firm for CEOs in founder lead and descendant lead family firms advanced in hypothesis 3. The fifth column of Table 2 shows that the impact of past financial performance on satisfaction in founder family firms is not significant. This means that for family firms where the founder and his/her family are present the impact of past financial performance is not as important as family firms governed and managed by next generation. The last column of Table 2 shows that in family firms lead by descendants, the impact of past financial performance is positive and significant. Thus, there is support for hypothesis Conclusion and discussion The pivotal contribution of the present research stems from the effort to bring together ideas from the entrepreneurship satisfaction literature (Cooper and Artz, 1997) as well as the SEW approach (Gomez Mejia et al, 2007) to examine how family ties influence the impact of past economic performance on CEO satisfaction in the context of small firms. Although we do not provide a direct measure for the importance of noneconomic goals, being able to control for the pure economic performance of the firms our theoretical and empirical analyses provide new ways of understanding the role played by the interplay between economic and noneconomic factors in family firms.

22 Our results demonstrate that regardless of their family ties, CEOs satisfaction in family firms would be less driven by financial performance than in non-family firms. We interpret this as indirect evidence that the utility of the owners of family firms (but not of the owners of nonfamily firms) is also affected by the achievement of family-centered, noneconomic goals, which influences CEO s overall satisfaction with the firm. Given the impact of decision maker s satisfaction on firm survival and growth (Cooper and Art, 1997) and the resilience of family firms (Chrisman et al, 2011), this finding opens up for new research to investigate how non-economic goals impact the survival and performance of small family firms. If performance outcomes do not drive CEO satisfaction in family firms, this may explain why many small firms live in a state of permanent failure, which Meyer and Zucker (1989) define as a condition characterized by sustained low performance and high persistence (p.68). The lack of empirical results for our hypothesis concerning the differential effect of past performance on satisfaction depending on the family membership of the CEO also opens an interesting avenue for future research. It may indicate the high influence of family owners in imposing the pursuit of their noneconomic goals. It may also be interpreted as consistent with a vision of non-family CEOs behaving as stewards of the controlling family (Davis et al., 1997). Socialization mechanisms may be playing a role here, especially in the context of small firms in which owners and firm culture are difficult to separate. Research is needed to shed light on such issues which, as suggested by our analysis, are much more complex than they might initially appear. As indicated by our post hoc analyses, under a negative performance trend, when the non-family CEO employment and wealth seems to be really at stake (Gomez-Mejia et al., 2001), non-family CEOs posits more weight on economic indicators when assessing their satisfaction with the firm. So, efforts to socialize non-family

23 CEOs in family firms may be worthless in economically difficult times, where pressures over employment and firm survival may turn non-family CEOs attention towards the more utilitarian aspects of the employment exchange. The study of these dynamics in the case of non-family CEOs in family firms is certainly an interesting avenue for future research attempts, given the growing importance of professionalization processes in family firms and the relatively scarce literature on non-family CEOs in family firms (Huybrechts, Voordeckers and Lybaert, 2013; Blumentritt, Keyt and Astrachan, 2007). The study also makes important contribution to our understanding of how the balance between economic and noneconomic factors change as the firm transitions from one generation to the next (Mazzi 2011; Olson et al. 2003). A strong case has been made that the emphasis on preserving the family s socioemotional wealth lessens as the firm moves through generations and that financial considerations become more important as a frame of reference (Gomez-Mejia et al, 2011). Our findings contribute to this discourse by demonstrating that over generations, pure economic goals would gain relevance as determinants of CEO s satisfaction. This also represent an important addition to the SEW literature which calls for more research explaining the factors behind the various sources and degrees of SEW (Berrone et al, 2011, 270). Lastly, the paper enhances the existing literature by considering a representative sample of small firms that are characterized by more flexible non-hierarchical structures. As we have argued before our sample of small medium and high-technology firms was particularly suited to test our hypotheses given, among other things, the importance that owners and CEOs preferences have on strategic decision making. But the sample of small firms is also important as little evidence exists on the differences between family and non-family

24 controlled firms in companies of this size. Most of the evidence on differences collected to date has been obtained from samples of bigger firms. This relative absence of results using sample of smaller firms, that may be due, at least in part, to the difficulty in access to relevant information, is particularly salient if we note that small and medium sized companies are the majority in the business landscape of any country, and account for a highly significant amount of employment and value creation. Efforts to improve our understanding of the context and dynamics of these small firms are welcomed. Our research has also some practical implications. One important request from family owners seems to be the reduction of the average failure rate of non-family executives, which are often caused by dissatisfaction (Klein, 2007). Our findings suggest that to increase the chance of a satisfying and successful employment relationship, family owners and nonfamily managers need to appreciate that there will be some differences in their goals which should be communicated in a proper manner. This may influence whether some managers stay with marginal businesses, keep on going with business when the financial performance is even negative, because the satisfaction is not a pure economic measurement. Limitations and Concluding Remarks Our work is not free of limitations. At least four aspects should be highlighted. First, we use a single item measure to capture CEO s satisfaction. Despite its limitations, the use of a single-item measure of satisfaction is very common in the large literature on employee job satisfaction (e.g. Bender et al, 2005). Moreover, in considering the psychometric properties of single-item job satisfaction measures, several psychologists have found that they show high reliability, significant validity and considerable predictability (Dolbier et al., 2005;

25 Wanous et al., 1997) that alternative multi-item measures. Secondly, the data is crosssectional. Cross-sectional studies can suggest correlations but do not allow researchers to infer causal relationships or effects over time. We have been particularly cautious in the language used to avoid mentioning causal inferences. However, it is worth noting that our objective measure of performance refers to the period preceding the survey, and therefore time causation is not violated. Thirdly, our sample consists entirely of Spanish firms, thus, any inference to other countries must be made with caution. Country-specific cultural and traditional influences may reduce the generalizability of our findings. Finally, data collection also captured a unique environmental context of economic and financial crisis. This added further difficulty to the already complex environment of firms in the high tech and mediumhigh tech industries, which is characterized by high degrees of dynamicity and stiff competition. This particularly harsh context may have reduced the latitude of action of firms, as well as influenced the impact of their decisions on subsequent firm performance. We extend a call to other researchers to explore the issues analyzed here to multiple cultural contexts, as well as to time periods absent of any global economic crisis. In conclusion, the impact of past financial performance on CEOs satisfaction with the firm in family firms is lower due, at least in part, to the greater relevance of noneconomic goals in the subjective evaluation of satisfaction in these companies. The satisfaction of family and non-family CEOs is, at least in positive context, independent of past economic performance. That is not the case when CEOs determinants of satisfaction are compared in founder and descendant family firms. Different preferences over economic and noneconomic goals of owners may explain the existence of differences in the determinants of CEOs satisfaction in these firms.

26

27 6. References Anderson, R. C., & Reeb, D.M. (2003). Founding-family ownership and firm performance: evidence from the s&p 500. Journal of Finance, 58, Arregle, J., Hitt, M. A., Sirmon, D. G. & Very, P. (2007). The Development Of Organizational Social Capital: Attributes Of Family Firms. Journal Of Management Studies, 44, Astrachan, J., Klein, S., Smyrnios, K., (2002). The F-PEC scale of family influence: A proposal for solving the family business definition problem. Family Business Review, 15(1): Barth, E., Gulbrandsen, T., & Schone, P. (2005). Family ownership and productivity: The role of owner management. Journal Corporate Finance, 11(1/2): Bhattacharya, U., & Ravikumar, B. (2004). From Cronies to Professionals: The Evolution of Family Firms, Working Paper, January 2004, Indiana University/ University of Iowa, US. Bender, K.A., Donohue, S.M. & Heywood, J.S. (2005). Job satisfaction and gender segregation. Oxford Economic Papers, 57(3): 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): Berrone, P., Cruz, C., Gomez-Mejia, L.R., & Larraza-Kintana, M. (2010). Socioemotional wealth and organizational response to institutional pressures: Do family controlled firms pollute less? Administrative Science Quarterly, 55,

28 Björnberg &Nicholson (2012). Emotional ownership: The next generation s relationship with the Family Firm. Family Business Review, 25(4): Block, L. M., Claffey, C., Korow, M. K., & McCaffrey, R. (2005). The value of mentorship within nursing organizations. Nursing Forum, 40 (4): Block, Joern H., et al. (2015). Job satisfaction and wages of family employees. Entrepreneurship Theory and Practice, 39(2): Blumentritt, T. P., Keyt, A. D. & Astrachan, J. H. (2007). Creating an Environment for Successful Nonfamily CEOs: An Exploratory Study of Good Principals. Family Business Review, 20, Bourgeois LJ, III & Eisenhardt KM. (1987). Strategic decision processes in Silicon Valley: The anatomy. California Management Review, 30(1): 143. Brockmann, E.N. & Simmonds, P.G. (1997). Strategic decision making: the influence of CEO experience and use of tacit knowledge. Journal of Managerial Issues, 9(4): Carree, M.A. & Verheul, I. (2012). What makes entrepreneurs happy? Determinants of satisfaction among founders. Journal of Happiness Studies, 13(2): Casillas, J.C., Moreno, A.M., Barbero, J.L. (2010). A configurational approach to the relationship between entrepreneurial orientation and growth of family firms, Family Business Review, 23 (1): Chen, S., Chen, X., Cheng, Q., & Shevlin, T. (2010). Are family firms more tax aggressive than non-family firms? Journal of Financial Economics, 45(1): Chrisman, James J., Jess H. Chua, and Lloyd P. Steier. (2011): "Resilience of family firms: An introduction." Entrepreneurship theory and practice 35.6 (2011):