Abstract
This article critically analyses entrepreneurial orientation (EO) in family firms after a major crisis, to investigate how firms with equal initial levels of EO reach different levels over time. Based on two alternative hypotheses (stability and convergence), we analysed whether the EO of family firms remains intact, strengthens, or weakens after a crisis. Based on an examination of a database of 151 family firms collected in 2004 and 2017, our findings reveal that compared to firms with higher pre-crisis EO levels, those with lower levels saw a larger increase post crisis. Furthermore, unlike the latter group, the former was able to maintain high pre-crisis levels even after the crisis. In addition, we also we found this relationship between pre-crisis and post-crisis EO levels to be influenced by two key periodic discontinuities, namely, organisational decline and generational change contingencies. These findings advance our understanding of temporal aspects of EO and heterogeneous entrepreneurial behaviour among family firms with significant implications for both theory and practice.
Introduction
Extensive research shows the positive implications of entrepreneurial orientation (EO) on firm performance outcomes (Andersén, 2017), including sales growth (Moreno and Casillas, 2008), profitability (Anderson and Eshima, 2013) and non-financial measures (Zhai et al., 2018). These manifold advantages and the significance of family firms in the economic and social landscape (Holt et al., 2018) have increased scholarly interest in family firm EO (Arzubiaga et al., 2018; Bauweraerts and Colot, 2017; Cherchem, 2017; Yu et al., 2019). EO, which comprises of entrepreneurial top management style, organisational configuration and new entry initiatives (Wales et al., 2020), varies across organisations and is, therefore, contextual in nature (Gupta and Batra, 2016; Miller, 2011; Wales et al., 2019). Indeed, the degree of EO and its outcomes vary as a function of the endogenous and exogenous phenomena influencing a firm (Wales, 2016).
In family firms, EO is inherently interesting as there are strong theoretical reasons to believe that it significantly differs from that in non-family firms (Casillas and Moreno, 2010) owing to the idiosyncratic characteristics of family-owned businesses (Barrett and Moores, 2020; Hernández-Linares and López-Fernández, 2018; McAdam et al., 2020). It is suggested that certain characteristics of family firms, such as conservativeness (Schepers et al., 2014), risk aversion (Zahra, 2010), organisational rigidity (Calabrò et al., 2019), willingness to maintain control of the firm (Gomez-Mejía et al., 2007) and propensity to avoid external funds for entrepreneurial projects (Arzubiaga et al., 2019), constrain EO. However, other studies report that the long-term orientation of family firms (Lumpkin et al., 2010), patient capital (Danes et al., 2009) and generational involvement (Sciascia et al., 2013) might foster EO.
Although these studies identify family firm characteristics that influence their EO, research connecting them and the temporal aspects of EO requires more attention (Grühn et al., 2017; Karami and Tang, 2019; Wales et al., 2011). The temporal aspect is important as the EO construct is regarded as stable, implying a sustained pattern of behaviour that manifests as an organisational state or firm quality (Covin and Miller, 2014). In this sense, to remain competitive, firms need to foster and maintain their EO over time, allowing them to remain competitive in the long term (Hernández-Linares and López-Fernández, 2018) as well as facilitate transgenerational succession (Zellweger and Sieger, 2012). However, retaining EO levels over time is usually a major challenge, particularly in periods of uncertainty or crisis (Zachary et al., 2017). Thus, understanding how family firms overcome a crisis with EO can advance our understanding about the heterogeneous entrepreneurial behaviour among family firms when facing major downturns. To the best of our knowledge, with the exception of a few recent studies not particularly focused on crises (Llanos-Contreras et al., 2019; Zachary et al., 2017), there has been no empirical investigation into the changes in family firm EO over these difficult periods.
To address this gap, our study investigates whether, after facing major downturns, EO of family firms remains intact or strengthens/weakens from its pre-crisis levels. Wales et al. (2011) suggest that firms may alter their focus on EO during periods of development and change, which means EO pervades firms heterogeneously over time (Grühn et al., 2017). We explore a firm’s long-run commitment to EO by exploring whether past EO may act as a basis for developing new entrepreneurial behaviours in the future. To do so, we adopt two alternative theoretical approaches. We use the stability hypothesis based on the long-term resilience of family firms (Chrisman et al., 2011), which suggests that family firms persevere in their strategic decisions, and their EO remains constant in the long term. Alongside this, we consider the convergence hypothesis, which proposes that family firms have less stable EO in the long term because of their tendency to converge with the strategic posture of external actors.
We investigate whether this relationship between pre-crisis and post-crisis EO levels is influenced by the two main periodic discontinuities faced by these firms, namely, organisational decline and generational change contingencies. With regard to the former, we analyse how organisational decline may moderate the relationship as experiencing economic decline significantly affects both firm performance and strategic positioning (Trahms et al., 2013). With regard to the latter, we focus on generational change as a potential moderator of this relationship as the complexity of succession processes, characteristics of a new chief executive officer (CEO) (Van Helvert-Beugels et al., 2020) and deep transformations within the firm may significantly alter strategic decision-making in the organisation (Jaskiewicz et al., 2016). To test our model, we observe 151 Spanish family firms in two specific years, 2004 and 2017, to examine EO variation over time. We find a negative relationship between the EO levels in 2004 and 2017; this negative effect intensifies when generational change occurs and weakens in times of decline following an economic crisis.
Our study significantly contributes to the literature. First, it provides a more nuanced understanding of time by exploring EO variations in family firms (Zellweger and Sieger, 2012). This partially addresses the call by Grühn et al. (2017) for research on EO temporal change patterns in small firms, where the effects of EO are more noticeable than in larger firms (Saeed et al., 2014). Although EO in family businesses has been explored, prior research has mainly focused upon static contexts, overlooking how EO in family businesses changes over time (Zachary et al., 2017). We contribute to the understanding of the specific EO behaviour of family firms by empirically demonstrating how EO changes over time, while complementing the more conceptually static-based research conducted to date (Hernández-Linares and López-Fernández, 2018). Second, we contribute to the family business literature by reconciling the contradictory predictions from the family firm stability perspective and convergence theory, to help develop a deeper understanding of the trade-off between EO and non-financial goals in family firms in a major crisis. Using this strand, we investigate how some family firms with equal pre-crisis levels of EO reach different EO levels after a complete recovery. Our results show that neither the family firm resilience perspective, nor the homogenising effect of the convergence approach, is able to fully explain the phenomenon as they fail to account for the temporal aspect of entrepreneurial behaviour. More specifically, we find that family firms with initially lower levels of EO are able to increase their EO more than those with initially, higher EO levels. In addition, the firms with higher pre-crisis levels of EO are able to maintain those levels even after the crisis, compared to firms that exhibited lower levels of EO pre-crisis. Third, since recognising the sources and contextual factors of family firm EO is important for expanding knowledge of variations in EO over time (Grühn et al., 2017; Lee et al., 2019), we contribute by showing how organisational decline and generational change are crucial to understanding family firm EO. Organisational decline has received limited scholarly attention (Kücher and Feldbauer-Durstmüller, 2019). While managing organisational decline, decision-makers confront challenges that are unique and different from those in normative times (Trahms et al., 2013). With regard to generational change, Hernández-Linares and López-Fernández (2018: 339) show that with one exception (Nordqvist et al., 2013), succession has been largely neglected by the literature on EO and family firms and highlight how succession influences EO in family firms as a priority in current literature due to the dynamic nature of EO (which) has not yet been fully addressed. Accordingly, by focusing on the moderating effects of organisational decline and generational change, we offer new insights into the understanding of EO behaviour in particular types of organisations such as family firms.
Literature review
EO in family firms
As an essential criterion for firms to succeed and grow in today’s competitive and dynamic business environment (Moreno and Casillas, 2008; Rauch et al., 2009), EO refers to the entrepreneurial top management style, organisational configuration and new entry initiatives (Wales et al., 2020). It is defined as a firm’s inclination towards entrepreneurship (Covin and Wales, 2012), capturing its innovation, proactiveness and risk-taking (Covin and Slevin, 1989; Miller, 1983). Lumpkin and Dess (1996) propose two additional dimensions, autonomy and competitive aggressiveness, as components of EO. While these two dimensions have been widely accepted in entrepreneurship research, the full five-dimension model is rarely seen in the EO literature in comparison to the original three-dimension model (Soininen et al., 2012). It has been argued that adopting an entrepreneurially oriented posture may improve firm performance (Rauch et al., 2009). Research shows that organisations with a strong EO have higher profitability and sales growth (Rauch et al., 2009; Saeed et al., 2014).
Representing a key concept within the entrepreneurship and strategy literature (Covin and Wales, 2019; Lee et al., 2019), EO has received the attention of family business scholars (Arz, 2019; Hernández-Linares and López-Fernández, 2018; Lumpkin and Dess, 1996). Research on EO in family firms focuses on highlighting the characteristics of these firms and how they influence EO, thereby showing that the EO framework is a useful tool to study entrepreneurship in such firms. Indeed, scholars consider EO as a long-term performance indicator that reflects continuity and transgenerational wealth creation (Casillas et al., 2010; Stenholm et al., 2016). This has led to considerable research interest in examining why, how and when EO manifests in family firms (Boling et al., 2016; Stanley et al., 2019). There is an ongoing debate on how the distinctive characteristics of family firms, as well as environmental factors, foster or hinder EO (Arz, 2019; Casillas et al., 2010; Lumpkin et al., 2010; Miller et al., 2016; Naldi et al., 2007). There have been studies on structural family-related characteristics such as family involvement in ownership (Block, 2012; Zahra, 2012), governance (Arzubiaga et al., 2018; Bauweraerts and Colot, 2017), management (Alayo et al., 2019; Casillas and Moreno, 2010; D’Allura, 2019; Revilla et al., 2016) and generational involvement (Cherchem, 2017; Chirico et al., 2011; Cruz and Nordqvist, 2012; Kellermanns et al., 2008; Sciascia et al., 2012). Table 1 outlines a selection of studies on predictors of EO in family firms as compared to non-family firms. Family firms possess particular characteristics that offer a singular context for researching how environmental factors relate to their EO (Cruz and Nordqvist, 2012; Hernández-Linares and López-Fernández, 2018; Lumpkin and Dess, 2001). Contemporary research highlights the importance of the environment in the development of EO and its outcomes, as in environmental dynamism (Casillas et al., 2010), environmental jolts (Zachary et al., 2017) and environmental hostility (Casillas et al., 2011).
Selected studies regarding predictors of EO in family firms (compared to non-family firms).
EO: entrepreneurial orientation; FBs: family businesses; CEO: chief executive officer.
The findings are inconclusive on how these internal and external factors influence EO in family firms (Arz, 2019). This may be because research on EO and family firms has usually been based on cross-sectional studies, neglecting the causality within EO relationships (Hernández-Linares and López-Fernández, 2018). Moreover, ‘the problem with observing EO at a single point in time is it assumes that EO is stable and that, if measured at another point in time, no differences would exist’ (Zachary et al., 2017: 61–62). EO is likely to change, as firms evaluate their strategy in a continuously changing environment (Zachary et al., 2017). Despite its significance, there is little research on how EO changes over time, especially in the long term (Grühn et al., 2017; Wales et al., 2011). This is an important research gap, given that firms may alter their focus on EO during certain periods (Wales et al., 2011). Thus, although the relevance of the temporal dimension of EO has been acknowledged in both general management and entrepreneurship studies (Grühn et al., 2017), especially in uncertain times or major crises (Zachary et al., 2017), to the best of our knowledge, how EO changes in family firms over time during major crises has rarely been empirically investigated. By examining whether, when facing major downturns, family firm EO remains unchanged, or strengthens or weakens from its pre-crisis levels, we provide new insights into the heterogeneity of EO in family firms.
Hypothesis development
Regarding long-term changes in EO, two alternative hypotheses can be considered: a stability and a convergence hypothesis. The stability hypothesis is based on the long-term resilience of family firms (Chrisman et al., 2011). Defined as the ability of organisations to avoid, absorb, respond to and recover from situations that could threaten their existence (Lengnick-Hall and Beck, 2005), organisational resilience explains how firms adjust, adapt and reinvent their business models in a changing environment (Sharma and Salvato, 2013). Because the main goal of family business owners is long-term survival (Lumpkin and Brigham, 2011), resilience is especially important to ensure the adaptability of business models over time (Linnenluecke, 2017) as it facilitates the successful ownership and management of the firm by the next generation (Chrisman et al., 2011). The essence of resilience is the intrinsic ability of an organisation to maintain a dynamically stable state (Hollnagel, 2006), which forces family firms into the pursuit of longevity to manage the trade-off between continuity and adaptability (Campopiano et al., 2019). Resilience allows family firms to remain flexible and balance the core essence of both the business and the family (Sharma and Salvato, 2013). Consequently, family firms are prone to maintaining their strategic orientation through stable or evolutionary (instead of revolutionary) changes by promoting continuity in or a gradual expansion of existing strategies and EO (Zachary et al., 2017).
Family firms tend to emphasise long-term behaviour, favouring stable strategies aimed at long-term survival so, for example, family CEOs tend to remain longer in their roles (Boling et al., 2016) and show greater resilience and stability in decision-making about the business and the family (Sharma and Salvato, 2013). A key differentiator of family firms is their long-term perspective (Brigham et al., 2014; Chandler et al., 2016; Diaz-Moriana et al., 2020). It is commonly acknowledged that the longer horizon may stabilise strategic behaviour of family firms because it links current decisions to their long-term effects (Le Breton-Miller and Miller, 2006). This proclivity towards the long term may promote long-term strategies and perseverance with strategic decisions (Lumpkin and Brigham, 2011). Thus, the comparative human, social and knowledge advantages of family firms, over non-family businesses, may help them identify and exploit entrepreneurial opportunities that require a long-term orientation towards strategic decisions (Sirmon and Hitt, 2003; Zachary et al., 2017). This stability in decision-making results in greater permanence of family managers, especially long family CEO tenures (Huybrechts et al., 2013). As such, we expect that family-owned businesses with previous low EO levels would continue to behave conservatively, whereas family businesses with previous high EO levels would grow their EO over time. Therefore, we propose the following:
H1a (stability). In family firms, pre-crisis EO levels are not associated with EO growth over time, that is, pre-crisis EO levels are positively related to post-crisis EO levels.
However, the convergence hypothesis implies that firms tend to balance their EO over time, in a way that firms with lower levels of EO at the beginning of a period tend to increase EO over time, while companies with higher levels of prior EO reduce it. This convergence process reduces between-firm variance and increases within-firm variance over time (Certo et al., 2017). Firms change their strategic posture in the long term to fit that of the external actors (Greenwood and Hinings, 1996; Stinchcombe, 1997). This behaviour is related to isomorphism, defined as a ‘constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions’ (DiMaggio and Powell, 1983: 149). On similar lines, Tang et al. (2008) find, in the Chinese context, an inverted U-shaped relationship between EO and performance due to institutional processes. Relatedly, Wang et al. (2017) posit that legitimation, jointly with EO, helps new ventures overcome potential resource and capability constraints.
Research indicates that family firms adapt their behaviour to institutional context (Leaptrott, 2005). For instance, Berrone et al. (2010) underline that family firms are more sensitive to institutional pressures due to their non-financial goals (Gomez-Mejía et al., 2007). As Berrone et al. (2010: 84) state, ‘family owners tend to place greater value on social legitimacy for its own sake independent of financial considerations’. One of the main non-financial goals is to maintain social relationships (Berrone et al., 2012), defined as closeness to the environment that provides the family firm with higher legitimacy in its social context that transcends the family and reaches its members. Institutionalisation promotes a convergence process in long-term decision-making in family firms; more specifically, this convergence with external actors might lead to less stability of EO in family firms because the firm must adapt its strategic posture to external actors. According to this view, family firms will try to avoid extreme levels of EO (too low or too high), preferring to follow the average EO level of most firms, reducing the between-firm variability and increasing the within-firm variability over time (Certo et al., 2017). This dynamic allows family firms to follow strategies similar to those of other firms in their environment to achieve similar performance and protect their non-financial goals from the potential risk from extreme EO levels. Our convergence hypothesis suggests that family firms with lower EO in the past would try to increase their EO in the long term, while those exhibiting higher EO in the past would reduce it (negative EO growth). Thus, we propose the following:
H1b (convergence). In family firms, pre-crisis EO levels are negatively associated with EO growth over time, that is, pre-crisis EO levels are negatively related to post-crisis EO levels.
Several factors suggest that changes in EO do not manifest uniformly over time, but rather show a distinct temporal change pattern (Grühn et al., 2017). We extend our understanding of how pre-crisis levels of EO might influence long-term changes by exploring the moderating effect of the two main discontinuities in family firms. The first, organisational economic decline, is often related to an environmental influence on firm performance (Gomez-Mejía et al., 2018b), while the second, generational change, is purely internal (Cruz and Nordqvist, 2012).
Organisational decline
Contingencies affecting the relationship between socio-emotional and economic factors are important for understanding the heterogeneity in family firm decision-making (Martin and Gomez-Mejía, 2016). The volatile global economy, driving a continuous threat of organisational decline, has become a significant concern for managers worldwide (Trahms et al., 2013). However, the concerns related to strategy redefinition differ from those related to improving performance in a non-declining situation (Gomez-Mejía et al., 2018b). Organisational decline occurs when a firm’s performance or resource base weakens over a sustained period (Bruton et al., 1994; Weitzel and Jonsson, 1989), mainly because of gradual variations in market niches, sudden environmental jolts (Short et al., 2007), technological change (Dowell and Swaminathan, 2006), industry decline and changes in competitive dynamics (Van Witteloostuijn, 1998). In organisations in decline, management decisions and strategic actions vary from those in non-declining situations (Casillas et al., 2019). For instance, a decline in sales reflects that the firm is performing poorer than its historical standards and faces a higher probability of failure or severe restrictions on growth. (Gomez-Mejía et al., 2018b).
In the case of family firms, organisational decline may alter a family’s priorities and its subsequent firm strategies (Basco, 2013; Ruiz-Jimenez et al., 2015). These firms face organisational decline by taking actions at the firm and family levels, balancing their commercial and family interests that are usually in conflict with each other (Jaskiewicz et al., 2016). Hence, family firms that may face, or have faced, severe losses (e.g. a decline in sales) to their overall wealth usually alter their risk perception and management (Gomez-Mejía et al., 2018a). This is in line with family firms going above and beyond to maintain the continuity of the firm to protect their socio-emotional wealth (SEW) (Gomez-Mejía et al., 2007). Gomez-Mejía et al. (2018a) propose that ‘controlling families show a tendency to shift their attention from their primary SEW reference point toward an economic reference point’, emulating the preferences and strategic behaviours of their non-family counterparts during economic turmoil (p. 996).
From the perspective of the stability hypothesis, we contend that the tendency for family firms to protect their social status and identity determines how they perceive an unexpected environmental change in a way that reinforces their resilience (Campopiano et al., 2019). Resilience provides family firms with the capacity to reduce the consequences of environmental shocks that cause an economic decline as well as with the ability to benefit from unexpected environmental change (Campopiano et al., 2019). Particularly, resilience may help to develop creativity and trust in adapting to a new setting as well as assist the family firm in understanding new changes, such as business opportunities in a new competitive setting (Brewton et al., 2010). Indeed, the capacity to absorb and react to environmental jolts based on their resilience promotes firms to bear higher levels of risk (Smith, 2016) and adopt innovation strategies (Kotlar et al., 2014). Therefore, relaxing the constraining effects of non-financial goal protection preferences (Kotlar et al., 2018) results in substantial changes in firm EO levels (Zachary et al., 2017). So, we posit that family firms moderate their EO levels during organisational decline, thereby reducing EO stability over time. In other words, we expect that a period of organisational decline would attenuate the positive relationship between initial and final EO levels.
Under the convergence hypothesis, following past poor performance, family managers search outside the firm to ascertain what peers are doing and potentially imitate them (DiMaggio and Powell, 1983; Frattini et al., 2014). In this way, family managers balance the need to change EO to improve performance and the aversion to make decisions that could jeopardise survival. After a period of decline, firms are driven to behave like competitors to minimise risks and avoid outlier behaviours. Thus, from a convergence point of view, we posit that a period of organisational decline in family firms is positively related to EO growth over time. Therefore, we propose the following alternative hypotheses:
H2a (stability). In family firms, a period of decline in firm performance is positively associated with EO growth over time – a period of decline in firm’s performance attenuates the positive relationship between pre-crisis and post-crisis EO levels.
H2b (convergence). In family firms, the negative relationship between pre-crisis EO levels and its growth over time is more intense when the firm has experienced a period of decline – a period of decline in firm performance intensifies the positive relationship between pre-crisis and post-crisis EO levels.
Generational change
EO has a strong impact on risk perception which includes both non-financial and financial wealth (Alessandri et al., 2018). Furthermore, family managers tend to change their entrepreneurial posture towards preserving non-financial goals and family interests (Schmid et al., 2015). As such, a change in leadership in general, and generational change in particular, is regarded as an internal shock to family firms, which may influence their EO (Sciascia et al., 2013).
Family firms owned and run by the founding family achieve higher levels of EO, which decreases when the firms pass on to subsequent generations (Fang et al., 2018). This is closely related to the greater amount of time, energy and money devoted to founding the firm, enhancing the levels of commitment and identification between the family and firm (Bauweraerts et al., 2020). In addition, family firms owned and run by the founding generation may emphasise protection of non-financial goal endowment by favouring risk-averse strategies (Berrone et al., 2012). In later generations, when ownership passes from the founding generation, non-financial goals become less salient in decision-making (Bauweraerts et al., 2020), and these newer generations prefer to improve the economic welfare of their family over that of the business (Ensley and Pearson, 2005). As such, later generations could be more inclined towards reaping private benefits so, depriving the family firm of essential resources (Miller et al., 2011), through a modification of the EO level in relation to prior behaviour.
In addition, preserving EO over generations is considered one of the most important challenges for family businesses (Jaskiewicz et al., 2016). EO tends to be closely tied to the founder (Schein, 1983), and EO drivers change in different generations, with environmental drivers more important in predicting EO in second generation onward (Cruz and Nordqvist, 2012). Furthermore, higher formalisation and more developed control systems belonging to second- and third-generation family firms are known to affect their long-term EO (Cherchem, 2017). Multigenerational family firms involve complex family relationships (conflicts, resilience levels and divergent interests), which could counteract a shared vision and agreement on entrepreneurial decision-making (Eddleston and Kellermanns, 2007). Indeed, family conflicts in later generations could restrict learning capacity by impeding knowledge integration (Chirico and Salvato, 2008), affecting organisational resilience. Therefore, we posit that family firms, aiming to reduce conflicts, avoid extreme strategic decisions (high and low EO levels), search for shared frameworks and find external reference points in their environment.
We have suggested that under both hypotheses (stability and convergence), successors bring with them new ideas to develop the family firm and identify opportunities to change strategies and firm behaviour (Grühn et al., 2017) in search of long-term legitimacy. Under the stability hypothesis, a generational change affects EO, which leads the new-generation managers to search for external reference points to justify new strategic decisions to the rest of the family (Erdogan et al., 2018). New-generation managers would aim to maintain a dynamically stable state in the quest for longevity. They would search for flexibility with resilience and would be more open to external influences (e.g. other family businesses, competitors, non-family managers or directors), potentially changing previous levels of EO. For the same reasons, under the convergence hypothesis, new-generation family managers observe their environment and imitate the strategies of other firms to minimise potential mistakes that could jeopardise non-financial goals. At least in the initial years, the new generations would try to adapt the company’s previous strategies to those of the competition to minimise the potential risks from too singular behaviour in the market. Hence, generational change should alter the levels of EO displayed when the family firm was managed by the same generation for many years. Therefore, we hypothesise the following:
H3a (stability). In family firms, generational change is positively associated with EO growth over time – generational change attenuates the positive relationship between pre-crisis and post-crisis EO levels.
H3b (convergence). In family firms, the negative relationship between pre-crisis EO levels and EO growth over time is more intense when generational change occurs – generational change attenuates the positive relationship between pre-crisis and post-crisis EO levels.
Figure 1 summarises our two proposed models.

Theoretical models.
Methodology
Sample
Analysing the evolution of EO over time requires information from a sample at two time intervals. This study is based on a survey of 430 Spanish companies with initial data obtained in 2004. In 2017, these companies were recontacted to repeat the same questionnaire with additional information about potential changes in corporate governance. The aim of our research required a sufficiently long study period to be able to consider the influence of the long economic crisis in Spain (2008–2013) and real possibility of generational change. Family manager tenure is typically lengthy; Le Breton-Miller and Miller (2006) find that tenures in family-controlled firms usually exceed 15 years. Similarly, Cruz et al. (2010) find that CEO tenure in family firms is about 14 years, and Huybrechts et al. (2013) show that family CEOs have an average tenure of about 16 years, higher than non-family CEOs (8.5 years). Thus, our time span (13 years) is sufficient to capture both the effect of the economic crisis on performance and possible generational succession.
The Sabi database indicates that only 12 firms closed during this period. Sabi includes financial information on all firms in Spain (excluding individual entrepreneurs). We received 179 completed surveys, 151 from family firms. We tested potential biases in relation to firms that answered and did not answer the second survey, finding no significant differences in terms of size, age, sector, profitability or geographical scope. The Sabi database was used to obtain financial information for the whole sample period. During these 13 years, 56% of the firms experienced a period of decline and 45% experienced generational change. The sample comprised of small- and medium-sized firms (average size: 27 employees; 130 small-sized and 21 medium-sized businesses), with an average age of 42 years in 2017.
Variables
Independent variable
Our model assumes that prior levels of EO influence its future evolution. Therefore, our exogenous variable was the initial level of EO (EOi ). In both surveys (2004 and 2017), the EO of the firm was measured using the nine-item, five-point Likert-type scale proposed by Covin and Slevin (1989), taking the first measure of EO (2004) as the independent variable.
Dependent variable
We used two complementary dependent variables. The first dependent variable was EO growth in the long term. To measure this EO growth, we calculated the difference between EO 17 and EO 04 (EOgrowth). Positive values of EOgrowth represent increases in EO over time, while negative values represent decreases. The second dependent variable was EO 17, following Gibrat’s growth model in which the dependent variable is the predictor variable with time lag (Sorenson and Søensen, 2001). Using a lagged dependent variable in the regression model implies the existence of a growth trend and allows us to analyse how other predictors determine a potential deviation from this trend (Osiyevskyy et al., 2020).
Moderating variables
Two moderating variables were considered according to our theoretical model. Decline was the first dichotomous variable representing whether the company experienced a decline during the 2008–2013 crisis in Spain (Casillas et al., 2019). This phenomenon was measured following the literature on decline and corporate restructuring (Barker et al., 2001; Lim et al., 2013), which suggests that a company is in decline if it experiences at least two consecutive years of decline in operating profit, with at least one of them showing a loss. We coded this variable 1 if firms experienced a decline between 2008 and 2013 (35.6% of the sample) and 0 if they did not. Generational change was the second dichotomous variable representing whether there was a change in the generation managing the company between 2004 and 2017, coded 1 for generational change (45% of the sample) and 0 for no generational change.
Control variables
A wide set of control variables was considered. First, we considered firm size, measured as the log form (to reduce heteroscedasticity problems due to the skewed distribution) of the number of employees in 2017. Firm size is a proxy for the potential availability of the resources and capabilities (Barney, 1991) that may affect the strategic decisions of family businesses (Sirmon and Hitt, 2003). Second, we considered firm age, measured as the number of years between the company foundation and 2017, also transformed by the logarithm to avoid potential heteroscedasticity. Firm age is especially relevant in family businesses because it is usually connected to the generational level and potential complexity of corporate governance structure and family involvement (Anderson and Reeb, 2003). Third, we controlled for sector using three dummies representing Manufacturing, Services and Construction, with Agriculture as the reference sector. In addition, four financial variables were used as control variables based on the Sabi database: (a) Liquidity, measured by the mean value (2004–2017) of the firm’s current assets divided by current liabilities (Gentry et al., 2016); (b) Leverage, the mean value (2004–2017) of total liabilities divided by total liabilities plus net equity (Arteaga and Menéndez-Requejo, 2017); (c) ROA, the mean value (2004–2017) of return on assets (Tangpong et al., 2015); and (d) Volatility (McConaughy et al., 2001), the deviation between ROAs each year (2004–2017) and mean ROA of the period. Finally, we controlled for three family governance variables: (a) Family CEO, a dummy variable coded 1 if the CEO is a member of the controlling family; (b) CEO age (log form to avoid heteroscedasticity); and (c) Founder generation, a dummy variable representing whether the firm is still owned and managed by the founder (coded 0) or has been passed onto successive generations (coded 1).
Multiple linear regression models with interaction variables, previously standardised, were applied. As mentioned above, we modified some of the variables to solve distribution problems and verified that no multicollinearity problems arose.
Results
Table 2 summarises the descriptive and correlation matrices, and Table 3 presents the estimated models for both dependent variables (EO growth, model 1; EO 17, model 2). Models 1a and 2a (baseline model) include the control variables. Models 1b and 2b use the initial EO as an independent variable to test the independent effect of the two moderating variables (economic decline and generational change). Models 1c and 2c consist of the interaction effects of these two moderating variables. Finally, Models 1d and 2d include the possible triple interaction between the independent variable and two moderating variables. The computed variance inflation factor values do not exceed 3.23, far below the critical threshold of nine (Cohen et al., 2003).
Means, standard deviations and correlations.
EO: entrepreneurial orientation; CEO: chief executive officer; ROA: return on assets.*Correlation is significant at the 0.05 level (two-tailed). **Correlation is significant at the 0.01 level (two-tailed).
Regression analyses.
EO: entrepreneurial orientation; CEO: chief executive officer; ROA: return on assets.
p < 0.1. *p < 0.05. **p < 0.01. ***p < 0.001.
For H1, we find a negative relationship between the initial EO and its subsequent growth as well as a positive relationship between pre-crisis EO and post-crisis EO levels (2004–2017). This effect is significant in all the estimated models. Although these results could seem contradictory, they show a convergence process between firms over time. Our findings suggest that firms with lower pre-crisis EO levels increase their EO more than those that begin with higher pre-crisis EO levels although those firms that showed higher pre-crisis EO levels in 2004 are still able to maintain higher post-crisis EO levels in 2017, compared to those that started the period with lower pre-crisis EO levels. These results are also consistent with the fact that the mean EO values at both moments (EO 04–EO 17) for the entire sample are almost the same (0.631 and 0.648, respectively). This implies while globally, firms have not altered their EO for a long period of time (involving a major economic crisis), individually they have modified their levels of EO (in-firm variability), reducing variability between firms. Hence, the convergence hypothesis (H1b) is supported and the stability hypothesis (H1a) is rejected.
Regarding the moderating effects, there are no significant direct effects between generational change and economic decline in EO growth over time. However, the interaction effects show significant results. Thus, in model 1c, a positive moderating effect of economic decline (β = 0.149; p < 0.05) and a negative moderating effect of generational change (β = −0.168; p < 0.05) are observed. To help interpret the results related to the interactions, Figures 2 and 3 present the interaction effects. The results are similar when we consider EO 17 (post-crisis EO level) as the dependent variable. Model 2c shows a positive interaction effect of declining performance (β = 0.189; p < 0.05) and a negative moderation effect of generational change (β = −0.181; p < 0.05)

Moderation effect of generational change. (a) Dependent variable EO growth. (b) Dependent variable EO 17.

Moderation effect of decline. (a) Dependent variable EO growth. (b) Dependent variable EO 17.
Figures 2 and 3 show support for the convergence hypothesis, with a declining relationship between pre-crisis EO level and EO growth (Figures 2a and 3a) and a convergent evolution of EO from 2004 to 2017 (Figures 2b and 3b). Regarding the moderating role of generational change, the convergence hypothesis is more intense when the firm has experienced a generational change (greater slope in Figure 2a and less in 2b). This shows that generational change is a strategic change; if the predecessor generation had high levels of EO, the succeeding generation tends to reduce the EO and vice versa. Therefore, H2b is confirmed. Conversely, for the moderating effect of economic decline, the results in Figure 3 show that the convergence hypothesis is less intense when the company has experienced a period of decline in 2008–2013. However, if there has been economic turmoil, the line shows a lower slope in Figure 3a and a higher slope in Figure 3b; that is, firms tend to keep their EO more stable and maintain their initial strategy. This effect is the opposite to that proposed in H3b. Model 4 incorporates the triple interaction to verify the joint effect of economic decline and generational change. The results show that this effect is significant in both model 1d (β = −0.128; p < 0.05, see Figure 4a) and model 2d (β = −0.138; p < 0.05, see Figure 4b).

Moderation effect of decline and generational change. (a) Dependent variable EO growth. (b) Dependent variable EO 17.
The convergence hypothesis is observed in three of the four cases, with greater intensity in case of generational change. However, when there is no generational change, the convergence process is lower, and EO is more stable over time. Even when there is no generational change, there is a certain process of convergence when the company does not experience any issues with the results. Nevertheless, if the company faces a decline and there is no generational change, there is an emphasis on maintaining the initial strategy, showing higher levels of organisational inertia (Van Witteloostuijn, 1998). In this case, these firms are represented not only by a flatter (horizontal) line in relation to EO growth and a steeper line (direct relation between pre-crisis and post-crisis EO levels) but also by higher growth in EO. In other words, when faced with problems, family firms with no generational change for a long time are able to increase their EO more than those undergoing generational change.
Discussion
This article explored whether EO is a stable attribute of family businesses over time, especially over an extended period of time that includes an economic crisis. We also analysed this potential EO change when family firms experienced some internal changes during this period (generational transition and declining performance). We evaluated whether pre-crisis EO levels act as a basis for developing new entrepreneurial behaviour in the future (post-crisis EO levels) and how family firms engage in resilient behaviours to survive and adapt to external and internal contingencies. We adopted two theoretical approaches. The stability hypothesis suggested that family firms are resilient in adapting their strategic decisions to changing environments to maintain their EO in the long term, while the convergence hypothesis proposed that family firms tend to moderate their EO in the long term, following the behaviour of other firms. In addition, we analysed whether the relationship between the pre-crisis and post-crisis EO levels is influenced by the periodic discontinuities faced by these firms, namely, economic decline and generational change. Our findings showed a negative relationship between the pre-crisis EO level and its later trend, demonstrating that this negative effect arises more when generational change occurs and is hampered when family firms suffer a decline in performance.
Overall, our results supported the convergence approach. Family firms, oriented to surviving in turbulent and challenging environments, avoid taking individual risks and develop mimetic behaviour in the long run. To complete our analysis, we divided the sample into two groups: (a) family businesses that increased their EO between 2004 and 2017, that is, those that reacted to a crisis by improving their post-crisis EO levels (51% of the sample) and (b) family businesses that weakened their EO over this period, that is, those in which the crisis damaged their EO over time (49% of the sample). We ran a logistic regression analysis using these two groups as dichotomous dependent variables (dependent variable = ‘1’ if firm has increased EO from 2004 to 2017, and ‘0’ otherwise). Table 4 summarises our main results. Consistent with our previous results, we found that pre-crisis EO levels have a significant and negative effect on the probability of EO increasing, according to the convergence approach.
Logistic regression analyses (EO increasing).
EO: entrepreneurial orientation; CEO: chief executive officer; ROA: return on assets. *Correlation is significant at the 0.05 level (two-tailed). **Correlation is significant at the 0.01 level (two-tailed).***Correlation is significant at the 0.001 level (two-tailed).
Our findings showed that convergence behaviour is more intense when family firms face generational transition, suggesting that such businesses tend to catch up with others as new generations take over. This is in line with the idea that family firms tend to diminish their EO as they pass from the first generation to later ones (Martin and Lumpkin, 2003) and that environmental drivers are more important in predicting EO in family businesses in the second and subsequent generations (Cruz and Nordqvist, 2012). However, this convergent behaviour is less relevant for firms with a declining performance during economic crisis.
In addition, our results showed that family firms tend to be resilient in the long run, aiming to stabilise EO even in the face of a decline in financial outcomes. In other words, in contrast to non-family firms, changes in short-run financial performance do not promote substantial changes in family firm EO, thereby showing the stability and long-term orientation of this type of firms (Lumpkin and Brigham, 2011; Lumpkin et al., 2010). The results of the logistic analysis show that declining performance negatively moderates the relationship between pre-crisis EO level and the probability to increase it over time so that firms that experienced a declining performance during the crisis period tended to maintain a more stable EO in the future.
We also found that when facing declining performance in an economic crisis, family firms are less prone to changing their long-term EO showing resilience that is, the intrinsic ability of an organisation to maintain a dynamically stable state (Hollnagel, 2006). Both results – more stable EO under declining performance but more convergence in EO under generational change – show that family firms behave differently depending on the type of change they encounter. Such firms adapt their EO to contingent changes, shaping their behaviour in a resilient manner, for long-term survival (Chrisman et al., 2011). In addition, the logistic regression analysis found that only when family firms experience declining performance without any generational change, they maintain stable levels of EO (positive effect of EO pre-crisis levels (EO 04) on the probability of increasing EO over time; see Figure 5).

Moderation effect in logistic regression (probability of EO increasing).
This article makes several important contributions to the literature. First, while prior research on family firms has mainly focused on static contexts therefore, ignoring EO changes over time (Zachary et al., 2017), our study empirically examines EO behaviour over time. Following the call for research to investigate EO temporal patterns (Grühn et al., 2017), our research provides a more nuanced understanding of EO temporal changes in family firms (cf. Zellweger and Sieger, 2012). We provide evidence that the initial levels of EO in family firms are associated with their future levels. Accordingly, an important contribution is the proposition that initial EO levels not only help us understand future EO levels in family firms but also may be an important predictor of the EO behaviour of the firm over time. Consequently, we emphasise the importance of treating family firms as heterogeneous (e.g. Li and Daspit, 2016) by understanding their different approaches to EO evolution. We contend that EO in family firms is not as stable as often assumed in the literature; rather, it depends on the nature of the changes that the firm faces to survive in the long term.
Second, the study contributes to the literature by reconciling the contradictory predictions of family firm stability and convergence perspectives to understand the trade-off between EO and non-financial goals that family firms face after a crisis. We study how some family firms with equal pre-crisis EO levels reach different EO levels after recovery. Our results show that neither the family firm resilience perspective, nor the homogenising effect of the convergence approach, can fully explain the phenomenon as they fail to account for the temporal context of entrepreneurial behaviour. In particular, our results reveal that firms with initially lower levels of EO increase their EO more than those with higher initial levels of EO, and those firms that showed higher initial pre-crisis levels of EO are able to maintain those levels after the crisis compared to those that exhibited lower levels of EO, pre-crisis. In particular, we use this logic to investigate how family firms with previously equal EO levels reach different EO levels over time. Our research suggests that over time, family firms usually follow a convergence process in decision-making by avoiding extreme levels of EO and preferring EO levels similar to those of other firms.
Third, we expand the literature on family business and EO by providing a more nuanced understanding of how organisational decline and generational change in family firms are particularly important in understanding their EO. Recognising the sources and contextual factors of family firm EO is important to expand our knowledge of the evolution of EO (Grühn et al., 2017). Hence, our empirical results serve to develop a theory of EO and reinforce previous findings on the influence of environmental factors (Sharma and Salvato, 2013) and endogenous shocks (De Massis et al., 2013) on EO. In so doing, we respond to calls for research to examine the differences among family firm EO (Wales et al., 2011; Wales, 2016) in organisations with recent succession processes (Hernández-Linares and López-Fernández, 2018) as well as shed light on how these firms respond to the challenges emerging in an organisational decline context (Trahms et al., 2013). Another interesting feature is that EO evolution is tested on a sample of 151 manufacturing family firms in Spain, making the study a specific review of the temporal effect of EO and its evolution in privately owned family firms based on data that are not available from archival sources and are extremely difficult to obtain.
Finally, by extending our understanding of how family firm EO evolves during a major crisis and how a CEO succession might affect it, we contribute to the understanding of its impact for management and family business research and provide relevant insights for management practice, especially for family firm owners, managers and advisors. These insights are particularly useful in major environmental jolts such as the current COVID-19 pandemic, especially after the sudden death or departure of quite a few senior family business leaders (De Massis and Rondi, 2020). Through our insights, practitioners in these firms can develop a more coherent understanding of their EO behaviour over time when a generational change occurs by design or default.
Limitations and future research
Our study has some limitations that offer opportunities for further research. First, while relying on a single informant is relatively common in research surveying key informants (Simon and Shrader, 2012), this could be a limitation to inferring firm-level behaviour based on CEO responses (Cruz and Nordqvist, 2012). In both years (2004 and 2017), the informant was the CEO of the firm; however, whether they were the same person was unclear, even when no generational change was stated. We also relied on subjective data on our key construct, EO, which may have biased the results. However, CEO in-depth knowledge of decision-making and firm behaviour in private firms, as well as the scarcity of objective data on these firms, seem to legitimise the use of subjective data obtained from a single informant (Arzubiaga et al., 2018). In addition, there was no cause for concern in testing for common method bias (Kellermanns and Eddleston, 2006).
Second, we relied on data from a single country, Spain, which affects the generalisability of our results. Although we do not expect significant variation in the results between Spain and other Western countries since EO is contextual in nature, cultural (Martin and Javalgi, 2016) and institutional (Chow, 2006) factors may influence our arguments, especially in countries with cultures that significantly differ from that of Spain (Hofstede, 2001). Thus, we invite researchers to replicate our study in other geographic contexts to determine the generalisability of our findings.
Third, to keep our research model parsimonious, we excluded several aspects from our analyses. More research is thus required to augment the robustness of the model regarding commitment to EO over time (Grühn et al., 2017). Although our sample of 151 family businesses represents a solid database, survivorship bias might exist in our data because any firms that went bankrupt or merged before June 2017 were excluded from our sample. Moreover, we did not include other antecedents or moderators that could influence EO and possibly change EO patterns. Future studies could benefit by distinguishing between the incorporation of family and non-family CEOs in succession when examining the temporal change patterns in EO, since family CEOs are usually more dependent on family dynamics and temporal patterns than non-family CEOs (Zona, 2016). Moreover, future research could investigate the temporal aspects and patterns of changes in EO by exploring other internal and external triggers such as changes in ownership and board composition, political stability and tax incentives.
Fourth, an in-depth study of how EO changes over time is advised, by examining the potential conditions in past EO levels that predict EO continuity, while controlling for extant levels of this variable. These potential conditions could be related to resilience or even the tendency to embrace efficiency at all costs. Generational change and economic decline experience could moderate the relationship with increased/decreased EO after crises. As such, the moderators might be closely connected to new potential EO drivers such as willingness to change (Kellermanns and Eddleston, 2006), perceived technological opportunities (Cruz and Nordqvist, 2012) and access to financial resources (Arzubiaga et al., 2019).
Fifth, future research could further investigate the effect of generational change and economic decline when transforming the changes in EO into firm performance. The direct and indirect effects of EO on firm performance were outside the scope of this study, but future studies could examine how generational change and economic decline moderate the relationship between increased/decreased EO and firm performance. This research avenue may shed light on how changes in important strategic behaviour can help achieve reasonable financial performance and firm growth.
Conclusion
Research on EO has recently moved towards addressing the temporal aspects of entrepreneurial behaviour. The specificities of family firms provide valuable opportunities to explore how and why EO changes in broader organisational contexts. By leveraging the effect of past EO levels on the changes in EO (strengthens, weakens or remains intact) through economic crises, this study goes beyond traditional research on EO drivers illustrating how family businesses modify their EO over time. Our data indicates that previous levels of EO are negatively associated with changes in EO over time in economic crises; these variations are stronger when generational change occurs and weaker when the firm has experienced a period of decline. These findings have significant implications for both theory and practice.
Footnotes
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
Author biographies
