Mobilisation of survivability capital – family firm response to the coronavirus crisis

Family firms' responses to economic crises

The researchers analysing family firms' response to the 2008–2009 global financial crisis focused on several factors, such as growth (Arrondo-Garcia et al., 2016; Amato et al., 2020), capital structure (Ramalho et al., 2018), employment (Lee, 2006; Bjuggren, 2015; Van Essen et al., 2015), risk-taking (Arrondo-Garcia et al., 2016) and governance (Braun and Latham, 2009). Studies also investigated the performance of family firms in comparison with non-family peers (Amann and Jaussaud, 2012; Kachaner et al., 2012; Siakas et al., 2014; Van Essen et al., 2015; Minichilli et al., 2016; Arrondo-Garcia et al., 2016).

Most studies have confirmed that during economically difficult times, family firms outperform non-family businesses (Amann and Jaussaud, 2012; Kachaner et al., 2012; Van Essen et al., 2015; Minichilli et al., 2016; Arrondo-Garcia et al., 2016; Hansen et al., 2020). Dyer (2018), however, wonders whether all family firms outperform their non-family peers during times of crisis. For instance, Hansen et al. (2020) did not find such a cyclical effect when examining family firms operating in Organisation for Economic Co-operation and Development (OECD) countries. In contrast, the family firms' relative outperformance during economically difficult times has been demonstrated for Asian and non-OECD countries (Hansen et al., 2020).

Better performance of family firms during crises has been shown for publicly listed companies (Minichilli et al., 2016; Saleh et al., 2017) as well as private family firms (Bauweraerts, 2013; Arrondo-Garcia et al., 2016). Lins et al. (2013), however, found the opposite based on a dataset of nearly 8,500 family and non-family firms from 35 countries. The authors demonstrated that family firms underperformed, relative to other firms, during the 2008–2009 global financial crisis.

The ability of family firms to mobilise resources during and after an economic downturn might be an explanation of their stronger performance and ability to recover (Saleh et al., 2017; Amato et al., 2020). As a consequence of their greater ability to mobilise resources, family firms adopt long-term orientation – even during a crisis – investing more, conducting more research and development, and controlling costs better (Amann and Jaussaud, 2012; Kachaner et al., 2012). Explaining the better performance of family firms, Kachaner et al. (2012) claimed that family firms focus more on resilience than on performance and tend to manage their downside more than their upside, as opposed to their counterparts. Testing the results of 219 private companies from 2002 to 2011, Bauweraerts (2013) showed that family firms performed better in the Belgian market. The author explained that the positive impact of family involvement on firm performance during crises results from a less formalistic view of the organisation and a concern for the preservation of socioemotional wealth. Regarding the emotional attachment to the firm examined during the crisis in Spain over the 2006–2011 period, Arrondo-Garcia et al. (2016) argued that first-generation private family firms performed worse, investing and borrowing more than multigenerational family firms. This finding results from the first generation's special interest in non-financial goals, particularly the firm's survival, so it can be passed to the next generation and control can be maintained (Arrondo-Garcia et al., 2016).

Based on an S&P 500 dataset of non-financial firms covering 2006–2010, Zhou et al. (2017) obtained opposite results. The authors found that family firms outperform non-family firms during a crisis, where the firms which contributed to the outperformance were those where the founder was still present. The authors suggested that the superior performance of the founder firms may have been caused by the founders' lesser incentive to overinvest in order to boost short-term earnings during a crisis (Zhou et al., 2017). Revilla et al. (2016) found that greater family involvement in management was negatively related to the likelihood of business failure in the context of an economic crisis. The authors explained the finding as resulting from the higher socioemotional and financial costs incurred by family members actively engaged in daily business activities.

Family firms also tend to prioritise family goals in times of crisis. This result is related to the fact that in the event of the company's collapse, family businesses incur unique costs, in terms of both financial and socioemotional wealth (Gomez-Mejia et al., 2007). This conclusion may underpin the results of the survey carried out by Hirigoyen and Basly (2019) among small and medium enterprises (SMEs) in France, which demonstrated that a financial and economic crisis does not seem to be a factor in the decision to sell a given family firm, even if its financial results have deteriorated.

Summing up, family firms appear to manage exogenous economic shocks with agility. In general, lesser reactivity to economic cycles is explained by the family firms' unique characteristics: a long-term view and the specific resources resulting from familiness (Amann and Jaussaud, 2012; Kachaner et al., 2012; Bjuggren, 2015; Van Essen et al., 2015). In family firms, the owning family can serve as a resource enhancer and therefore become a stress buffer (Danes et al., 2009). The question is whether the familiness-related qualities of a firm contributed to the employment of private resources (survivability capital) during the pandemic, which differs from previous economic crises.

Family business behaviour after natural disasters

Opposite results may occur in cases of natural disasters because they can simultaneously constrain both the family and the business (Haynes et al., 2019). Natural disasters have the potential to disrupt firm owners' routines or standard operating procedures. The demand in family and business systems tends to be high and thus the usual patterns of work/family balance may no longer be effective. Business owners may feel torn between the needs of the family or those of the business (Brewton et al., 2010; Danes et al., 2009) [1].

The results of Brewton et al. (2010) show that rural family firms' resilience can be reduced at the time of a crisis caused by a natural disaster. These findings only pertain to family farm businesses, however, which may differ from other family firms. In the case of urban family firms, Brewton et al. (2010) concluded that when a given business was considered a way of life, as opposed to a way of earning income, the level of resilience increased. Cater and Beal (2014) studied the experiences of the family firms affected by the biggest oil spill in Louisiana. Based on a qualitative case study approach, they listed strong network relationships, trust-based relationships, extensive local idiosyncratic knowledge, flexibility and the ability for immediate decision-making as unique characteristics of family firms, which increase their competitive advantage and enhance performance during an externally induced crisis. The results of Haynes et al. (2019) suggest that the small family firms affected by Hurricane Katrina, in which private issues were more frequently in conflict with work demands, were more likely to survive. The family firms that experienced greater stress and elevated family-business conflicts, however, had less chance of success.

Salvato et al. (2020) found that, after the earthquake in central Italy in 2009, family firms outperformed their non-family counterparts. The authors suggested that the family firms' better performance resulted from the family-related social resources and the support from political constituencies. In their opinion, the long-term perspective and the willingness to pass the company onto subsequent generations provided these firms with the social and emotional capital required to cope with the crises caused by natural disasters. Salvato et al. (2020) further underlined the fact that exogenous shocks, such as natural disasters, might threaten family involvement in ownership and management. Shocks may force a family-owned business to shut down or relinquish control to other firms. Thus, given the pandemic's similarity to natural disasters, the question is whether the owning families are ready to mobilise survivability capital.

Family business reactions to the COVID-19 pandemic

A growing body of research explores the family effect on the businesses' responses to pandemic threats (e.g. Amore et al., 2020; Kraus et al., 2020; Zainal, 2020; Zajkowski and Żukowska, 2020; Marjański and Sułkowski, 2021; Rivo-López et al., 2021; Soluk et al., 2021).

Based on the daily stock return data for listed companies in Italy, Amore et al. (2020) argued that family firms have outperformed their non-family peers during the pandemic. The study focuses on the firms in which a family is the controlling shareholder and holds the CEO position. As surveyed by Kraus et al. (2020), liquidity is vital for family firms, though not all of them suffer from severe liquidity problems. Family firms' liquidity needs depend on their ability to mobilise the financial capital of the family business owners when needed (Kraus et al., 2020). Such owning family's support is meant to secure investments and employment. These results are consistent with the previous research confirming the family entrepreneurs' support of their employees during a crisis (Bjuggren, 2015; Lee, 2006; Van Essen et al., 2015).

Similar results obtained by Zajkowski and Żukowska (2020) and Marjański and Sułkowski (2021), indicate that most Polish family firms' owners avoid employment redundancy. Yet, as pointed out by Kraus et al. (2020), mobilisation of family resources to support family business does not necessarily occur in all family firms. It should be noted that Kraus et al.'s study was conducted at the beginning of the pandemic, on a sample of only 27 family firms from five countries. The results obtained by Zajkowski and Żukowska (2020) showed that family firm owners in Poland chose neutral or persevering strategies to minimise the consequences of the COVID-19 pandemic in the first days of the downturn.

Family members and survivability capital

Concerning the family firms' resilience, the extensive literature underlines their unique resources and specific strategic capabilities, which differentiates them from non-family firms (Lee, 2006; Brewton et al., 2010; Chrisman et al., 2011; Bauweraerts, 2013; Minichilli et al., 2016; Arrondo-Garcia et al., 2016; Saleh et al., 2017; Amato et al., 2020). To gain better insights into the resources contributing to the resilience of family firms, this study is based on RBV and SFBT.

RBV states that better firm performance that is not attributable to an industry or the economic conditions is commonly based on the competitive advantage of a company. Furthermore, RBV indicates how resources can contribute to the competitive advantage of organisations. The theory asserts that opportunities for competitive advantage and superior performance result from the idiosyncratic, the immobile, the inimitable and sometimes the intangible resources controlled by a firm (Barney, 1991; Peteraf, 1993). Barney (1991) recognised all assets, capabilities, organisational processes, firm attributes, information, knowledge and other elements controlled by a firm as resources. They include a broad range of organizational, social and individual phenomena within firms (Habbershon et al., 2003). In family firms, these resources are associated with three independent but overlapping subsystems: business, ownership, and family (Tagiuri and Davis, 1996; Gersick et al., 1997). The family subsystem differentiates family firms from their non-family peers (Chua et al., 1999; Chrisman et al., 2003). The idiosyncratic bundle of resources and capabilities resulting from the interaction between the family, its members, and the business are referred to as the “familiness” (Habbershon and Williams, 1999). An individual family member in a family firm system may be placed within one of the sectors formed by the three overlapping circles (Tagiuri and Davis, 1996; Gersick et al., 1997). Family members who belong to the three subsystems are family-owner employees (founder-owner who is active in the firm's operation). Family members who occupy the subsystem ownership and the family are family owners (e.g.: co-owner – spouse of the founder who is not active in the firm's operations). Family members representing the family and the business subsystem are family employees (e.g. a child or relative who is active in the firm's operation without holding any shares) (Tagiuri and Davis, 1996; Gersick et al., 1997). These three groups of family members are directly involved in the family firm activities (in the study referred to as directly involved family members). They are not only linked to the family firm economically, through employment and ownership, but they are also emotionally attached, with a strong desire to contribute positively to the organizational outcomes. They are the family members who can derive both financial and nonfinancial benefits from the firm.

A group of family members who are neither the owners nor employees also exists in family firms (Tagiuri and Davis, 1996; Gersick et al., 1997; Danes et al., 2008). This group can be represented by the children not yet engaged in family firm operations, older generations who have already resigned, as well as close and distant relatives not engaged in the business. They are not directly involved family members but more or less emotionally attached to the business. Although they do not represent the formal family link to the business, either through employment or ownership, nevertheless they constitute an important part of the family firm system (Anderson et al., 2005).

All these characteristics and the coalitions dominant inside a family firm may affect the family's willingness to activate private resources to support the business. Each of the interest groups identified has its own goals, concerns and dynamics. They also differ in terms of the importance they place on SEW. Each group of family members has a different basis for its commitment to the family firm, which might not always be aligned (Dunn, 1999; Villalonga et al., 2015; Revilla et al., 2016). When functioning in a family subsystem, people value family ties and relationships, whereas for the owners the goal entails high returns and the firms' success (von Schlippe and Frank, 2013). Some family firm resources held by the family and its members may be mobilised at a different level, regardless of the group to which they belong. Haynes et al. (1999), for example, pointed out that the use of family resources is more probable in sole proprietorships by older owners without children in the household or if a given business owes money to financial institutions.

Habbershon et al. (2003) suggested that “family companies may have numerous intuitive based resources not accounted for in the everyday assessment of their competitive advantage” (p. 12). These causally ambiguous resources only appear during periods of change (Habbershon et al., 2003). This could mean that some of the resources residing in one of the subsystems (business, ownership and family) appear to be “dormant resources” similar to “dormant stakeholders” (Mitchell et al., 1997). “Dormant resources”, just as “dormant stakeholders”, have little or no interaction with the firm in day-to-day operations; nevertheless, the dynamic nature of their relationship with the family firm can make them more salient if they become critical to the family firm's functioning (Mitchell et al., 1997; Craig and Moores, 2010). It is believed that “dormant resources” can only be mobilised in hard times or to make strategic decisions. This assumption seems to be in line with SFBT. The unique contribution of SFBT is that it recognises the fact that the standard operating procedures used in normal, stable times need to be adapted during times of change. In stable times, the family and the company are managed within their boundaries. In times of disruption, resources flow as needed between the family and the business systems (Stafford et al., 1999; Danes and Brewton, 2012; Danes et al., 2016). It may therefore be assumed that some of these resources were “dormant” in stable times until disruption, similar to certain family network ties recognized by Jack (2005). Resources arise from the family's environment and from within the family itself (Stafford et al., 1999; Danes and Brewton, 2012; Danes et al., 2016). They help family firms to sustain the business, especially under adverse or unusual circumstances. Under “normal” conditions of business functioning, these resources constitute a sort of invisible and virtual asset. They are employed occasionally, as a tool supporting the business processes, only engaged to protect sustainable growth and the probability of business survival (Yilmazer and Schrank, 2006; Glover, 2010; McDonald and Marshall, 2018).

The unique resources of family firms have been referred to as “familiness”, while survivability capital is one of these resources (Sirmon and Hitt, 2003). As suggested by Sirmon and Hitt (2003), to produce value, resources must be managed appropriately, because not all firm resources provide a competitive advantage (Barney, 1991). Weismeier-Sammer et al. (2013) maintain that, if groups of resources are combined with strong family cohesion and high communicative capabilities among the employees and family members, familiness can develop in a positive manner. If trust, norms, obligations, and identity are strong and oriented at the collective, family members are more committed to each other and more likely to form an entrepreneurial team (Cruz et al., 2013), thus more likely to mobilise the “dormant resources”. Familiness is not without a dark side, however. Some families are characterized by competition, nepotism, and destructive behaviour (Schulze et al., 2003).

This study focuses on survivability capital, which is defined by Sirmon and Hitt (2003) as “the personal resources that family members are willing to loan, contribute or share for the benefit of the family business” (p. 343). In their assumptions, however, Sirmon and Hitt (2003) have not distinguished between family members directly and not directly involved in the family firm. Following the assumptions of SFBT (Danes et al., 2008), in the research presented, family members have been divided into two groups.

Survivability capital seems not to be attributed to company size, location, industry or economic conditions. It may take the form of unpaid family labour, a cut in pay during times of difficulties, additional equity investments, monetary loans, and more (Haynes et al., 1999; Olson et al., 2003; Mzid, 2017). Survivability capital is unique for any given enterprise and family, therefore impossible to imitate. It represents the family members' ability to strengthen the financial and non-financial resources of the businesses facing threats to their economic existence (Haynes et al., 1999; Olson et al., 2003; Sirmon and Hitt, 2003; Van Auken and Werbel, 2006; Saleh et al., 2017; Neubaum and Voordeckers, 2018; Amato et al., 2020). The method of survivability capital management for the benefit of an enterprise is also characteristic of a given family firm and contributes to the firm's competitive advantage.

Although researchers have mentioned the existence of survivability capital (Carney, 2005; Sciascia and Mazzola, 2008; Brewton et al., 2010; Memili et al., 2013; Wilson et al., 2013; Arrondo-Garcia et al., 2016), they have rarely investigated the circumstances under which it is mobilised, nor have they measured its influence on firm performance. Examination of the circumstances in which survivability capital is mobilised seems to be of importance since the identification of its activation impulses can constitute the basis for explaining the greater likelihood of survival and the better financial performance of family firms compared to non-family firms after external shocks (Amann and Jaussaud, 2012; Kachaner et al., 2012; Wilson et al., 2013; Van Essen et al., 2015; Minichilli et al., 2016; Arrondo-Garcia et al., 2016; Hansen et al., 2020). Of course, not all family businesses possess survivability capital (Sirmon and Hitt, 2003) and the degree of family commitment to a business varies across families (Van Auken and Werbel, 2006).

To conclude, it is proposed that survivability capital constitutes a type of a “dormant resource” within RBV. In line with SFBT, family firms can access external, privately owned family resources from the family system when needed, since family and business interact by exchanging resources across their boundaries during times of change or shock (Stafford et al., 1999; Danes and Brewton, 2012; Danes et al., 2016). If a given family firm does not require “dormant resource” reactivation and manifestation, these resources remain latent and dormant as certain family network ties (Jack, 2005). Survivability capital might also be labelled as a type of “effectual resource”, with the ability to turn an unexpected situation into a profitable one. This effectual resource also seems to be considered by family members as an affordable loss (acceptable risk) incurred to the benefit of the family firm (Sarasvathy, 2001; Barrett and Moores, 2012).

It could therefore be assumed that the importance of this source of “family business support” gains momentum in a crisis period. Shocks that affect family firms and their economic stability trigger the need to “engage” their survivability capital. These shocks can be family- or business-related and may be caused by declining performance or emergence of problems with survival. As Casillas et al. (2019) suggested, deterioration of economic situations and survival problems are two different concepts. Both situations are correlated in the long run but not necessarily in the short term. Crisis-induced deterioration of the economic situation is associated with a decline in turnover and profits or with cash flow problems, whereas problems with survival are associated with both poor financial results and the possibility of losing SEW. Correspondingly, it seems worth distinguishing the two manners of reacting to both phenomena, depending on whether or not a firm's survival is threatened, given that the firms' long-term orientation and the owners' desire to transfer the wealth to future generations constitute fundamental objectives of family firms (Chua et al., 1999; Lumpkin and Brigham, 2011; Basco, 2017), whereas any damage to SEW is a critical and major loss for a family firm (Gomez-Mejia et al., 2007).

The COVID-19 crisis has had an impact on both the condition of the family (health) and the firm (demand and supply chains). Considering the related financial needs of both these systems, the authors are aware that the amount of survivability capital activated during the pandemic may be lower than in typical financial crises, such as the 2008–2009 global financial crisis. What is more, the different components of survivability capital will probably be influenced by other determinants, which in turn may reduce or enlarge the potential level of the survivability capital mobilised. For example, an epidemiological situation in the family or the firm, or the psychological attitudes towards a threatening situation (such as fear of disease) may impose or limit additional unpaid working hours on the part of family members. The same impact on this component of survivability capital may affect the family life cycle (e.g. the need to care for children or the elderly). The family members' financial wealth can strengthen or shatter the survivability capital available in the form of additional equity investments, monetary loans or pay cuts for family members. Considering family firms' characteristics such as a long-term perspective and emotional attachment (Gómez-Mejía et al., 2007; Lumpkin and Brigham, 2011; Basco, 2017), it seems that families will attempt to mobilise as much survivability capital as they can, even during the time of the COVID-19 crisis. It thus seems that when family members predict a crisis-induced deterioration of the economic situation to be normalized in the short term, resources that are readily available and can be mobilised immediately will be involved. Furthermore, Olson et al. (2003) determined that family business' success depends on family processes and the response to disruptions, which were observed to have a greater impact on family business revenues (20% of variance) than on family resources (2% of variance).

In line with the above concepts, it is assumed that some facets of survivability capital (e.g. unpaid work, use of private assets, securing a new loan with private assets, or providing monetary support) may be provided by directly involved family members. Further in the paper, we describe it as internal survivability capital.

The authors posit that:

SFBT recognizes that family and business exchange resources and capital, including the resources of family members not directly involved in the business (Danes et al., 2008). Although the aid provided by family members who are not directly involved in the business is especially important at the start-up stage, indirectly involved family members often provide a range of assistance to entrepreneurs in subsequent years. The rapidity of their help and its low or non-existent cost is vital (Anderson et al., 2005). As Anderson et al. (2005) concluded, more than two-thirds of the family members to whom entrepreneurs turned for help were not involved in a given business. They can offer a range of important resources to entrepreneurs, both professional and effective. Glover's (2010) findings suggested that family networks constitute an important source of support during difficult times. Olson et al. (2003) also pointed out that the external family resources mobilised due to disruptions impacted the business revenue. Mzid (2017) concluded that family firms can overcome crisis-induced financial constraints through family support.

Summing up, mobilisation of the survivability capital held by family members not involved in the business can constitute the means of extending the family firm and familiness boundaries as well as provide benefits from a broad range of family networks (Wilson and Tonner, 2020). The survivability capital derived from this group of family members is defined further in the paper as external survivability capital.

Accordingly, the authors posit that:

As mentioned above, Kachaner et al. (2012) argued that family firms focus more on resilience rather than on performance. Thus, family firms can react differently in crisis-induced deterioration of the economic situation and when they expect problems with their survival. When the decline of financial performance is strong enough to threaten a firm's survival, firm longevity is under threat. A decreasing probability of survival prevents the family firm from being passed on to the next generations, which is why family members will likely take actions to ensure firm survival (Wilson et al., 2013). Problems with survival may result in unique costs for families and family firms, associated with both financial wealth and SEW (Gomez-Mejia et al., 2007; Revilla et al., 2016). “Business failure is more than a professional failure for family businesses; it is, to a certain extent, a family failure which affects its internal and external reputation” (Casillas et al., 2019, p. 63).

Responses to problems with survival may depend on the nature of the family members' participation in the business. When they actively participate in management and ownership, their reaction can be stronger and quicker than that of the family members who are not involved in the family firm. This behaviour is explained by the higher socioemotional and financial costs incurred by directly involved family members (Revilla et al., 2016). For directly involved family members, the financial condition of the household and the business are closely related, meaning that achievement of a given family's welfare is impossible without achievement of security and prosperity in the business (Danes and Brewton, 2012). It appears that during difficult times families are pressed to mobilise private assets and external financial resources, providing the company with survivability capital. Directly involved family members can additionally ask older generations and close or distant relatives who are not directly involved in the family firm for some sort of support. Due to the blood ties and family relationships, the above-mentioned relatives may get involved and help a given family firm through a loan, unpaid work, and so on. Problems with a family firm's survival can have adverse implications on its reputation, which affects the entire family, not only the directly involved family members (DeTienne and Chirico, 2013). The not involved family members' willingness to support a family firm can result from the desire to preserve its good reputation. Very often a given family firm's brand is linked with the family name and any reputational damage can blemish the entire family, its social status, and interests (Mzid, 2017; Alrubaishi, 2020). When a given family's name is linked to the firm's name, family members exhibit a stronger commitment to the family firm, because all family members, both those directly involved and not directly involved, are aware that it is not possible to change the family if the family name is so to speak soiled (Dyer and Whetten, 2006; Sageder et al., 2018). Different family members can own and run multiple family firms simultaneously. In such a situation, one family firm's problems with survival can damage not only the reputation of that family firm but of the entire portfolio of firms within a given family system (Jenkins and McKelvie, 2016). Moreover, the relatives not directly involved in the family firm can sometimes act as informal investors. In case of a given family firm's failure, its financial problems can extend to those members (Jenkins and McKelvie, 2016). In this context, the authors claim that if family members suspect problems with survival, they try to use all the resources available, as per the “all hands on deck!” rule.

Based on this discussion, the authors hypothesise that:

Data collection

The authors conducted a cross-sectional study on a sample of Polish private family firms. The Polish economy is in many areas comparable to countries with stable market economies. The enterprise sector generates nearly 72.7% of the Polish GDP, and the SME sector generates nearly half (49.1%) of that, creating 67.4% of total employment in the enterprise sector (PARP, 2021). According to World Bank data, during the years 1990–2019, Poland achieved one of the highest rates of growth of GDP per capita out of all OECD and European countries (https://data.worldbank.org). During that period, GDP per capita increased by almost 807%, from $1,731 (1990) to $15,695 (2019) (https://data.worldbank.org). Nevertheless, the GDP per capita in Poland is still much lower on average than in the European Union or the United States.

Most private businesses in Poland were established in 1990 or later. This means that most family firms have a rather short market experience compared to, for example, the Anglo-Saxon countries. The average age of a Polish family firm is only 21 years (Pernsteiner and Węcławski, 2016), whereas in the United States an average private family firm has been in operation for 49 years (Schulze et al., 2003). The consequence is that most family firms in Poland are still managed and owned by their founders (Kowalewski et al., 2010; Pernsteiner and Węcławski, 2016; Dick et al., 2021). Węcławski and Żukowska (2019) describe them as beginners “who seek good succession patterns and good practices in transferring family business value” (p. 134).

Primary data were collected from 14 April to 6 May 2020 via mail surveys. In Poland, this period represents the business restrictions and closures implemented to minimise the risk of COVID-19 transmission. The Great Lockdown in Poland started on 24 March 2020. From that point, the most severe restrictions were announced: nonessential travel was prohibited and many businesses were closed. To alleviate the adverse effects of the lockdown, the Polish government implemented an economic support programme in the form of the Anti-Crisis Shield. The scope of the support changed dynamically, thus two versions of the Anti-Crisis Shield were implemented during the research period. In addition, during the research process, an initial defrosting plan was announced (Figure 1).

These rapidly changing circumstances were challenging for the Polish family firms. They had to stay current with unstable policies and quickly adapt their businesses to new standards. The authors consider this period as a suitable setting for testing whether families use their private resources to support their businesses and for examining the way family firm survivability capital is mobilised.

The research was addressed to owners and managers of private family firms in Poland. A list of 8,428 potential respondents, with their contact details, was developed by searching the media, national registers, and checking the family firms forums, foundations and websites (there is no official dataset of family firms in Poland). A similar approach was used by Machek et al. (2015) and Madison et al. (2018). After initial and follow-up mailings, 272 surveys were completed, which resulted in an initial response rate of 3.2%.

Because of the difficulty with precise identification of family firms a priori, they were examined ex post. The subject literature contains numerous definitions of family firm (Hernández-Linares et al., 2018; Hernández-Perlines et al., 2020). The definition chosen for this research is based on self-defining. In the authors' opinion, this approach is the most suitable for examining the family firms' unique resources and their idiosyncratic crisis responses. This approach is based on the well-known Thomas theorem, which is defined as follows: “If men define situations as real, they are real in their consequences” (Bornmann and Marx, 2020, p. 554). It can be assumed that if an owner or manager declares that his/her firm is a family firm, then all the consequences (including the creation of and access to unique resources) are the real output of this situation. This criterion has been used in previous studies (e.g. Gallo et al., 2004; Zellweger et al., 2012). It should be highlighted that the use of this sample limits the extent to which the results can be generalized (e.g. they cannot be applied to family firms that do not consider themselves family firms). Nevertheless, in this case, the authors believe that it makes the results more conceptually aligned with the aim of the paper.

In total, 167 completed online questionnaires were received from family firm managers and owners. This low response rate partially resulted from the ongoing COVID-19 crisis (the authors received 100 automatic responses with information that a given firm was temporarily closed, or the manager/owner was absent due to childcare leave). Due to the low response rate, analysis was conducted to check for potential non-response bias (Hudson et al., 2004). The data was split into two groups: early and late respondents. Independent sample t-tests were conducted to compare the means of the variables used in the research. No significant differences between the variables analysed were found, therefore, non-response bias did not occur in the research presented.

The sample's common method bias was checked twofold. The first was Harman's one-factor test. The total variance extracted was 40.78%, which is lower than the maximum threshold of 50% (Riley et al., 2018). The second followed Knock (2015). If all the VIFs resulting from a full collinearity test are equal to or lower than 3.3, the model can be considered free of common method bias.

Additionally, due to the small sample size, potential sample bias was examined (Madison et al., 2018). The characteristics of the research sample were compared with a population of all small and medium enterprises from Statistics Poland and with a sample of 396 family firms used by Pernsteiner and Węcławski (2016). Because different characteristics were presented in these studies, a complete comparison could not be provided. Based on the data presented in Table 1, however, it can be assumed that the research sample is similar to the population of businesses used by Pernsteiner and Węcławski (2016). Hence, sample bias did not occur in the research presented (Table 1).

The size of the data set is adequate considering the methods used in the paper. As per recommendations, the sample size is greater than (1) ten times the largest number of the survey items subjected to factor analysis (Jung and Marron, 2009); (2) a reasonable sample size for CFA analysis, which is 150 (Muthén and Muthén, 2002); and (3) ten times the largest number of the structural model paths directed at a particular construct in the structural model (recommendation for PLS-SEM analysis) (Hair et al., 2014). The sample was dominated by small and medium enterprises, with an average employment of 44 people and average revenues of 37.8 million PLN. The average age of the firms was 23.64 years. More than a half (54.1%) were managed by the second generation, only 5.9% by the third, and 0.5% by the fourth. However, 47.3% of examined businesses were owned by founders, with 46.7% in the hands of the second and only 6% of the third generation. These characteristics are typical of the Polish family firms sector, which has a relatively short market experience (as mentioned above).

Dependent variables – measuring the survivability capital construct

As the literature shows, the family business field requires its own tools to recognise its uniqueness. For mainstream family business research to advance, proper measurement of the constructs is essential, hence the development of specific models and scales dedicated to measuring the idiosyncratic constructs in family firms (Astrachan et al., 2002; Habbershon et al., 2003; Klein et al., 2005; Berrone et al., 2012; Debicki et al., 2016; Frank et al., 2017). The review of the existing literature seems not to reveal a scale for examining survivability capital as a separate construct. Sustainability capital is also not included in the family influence familiness scale (FIFS) (Frank et al., 2017), although it is considered as a resource based on familiness. Due to the unique opportunities resulting from the ongoing crisis, the authors decided to develop their own measurement scale called the “mobilisation of survivability capital construct”. This scale attempts to complement the concept of familiness and improve the general understanding of family firms as unique entities. The set of items examined was generated following a deductive approach (applying Sirmon and Hitt's (2003) definition and fragmentary studies).

Independent variables

To verify Hypothesis 1, a single independent variable (“crisis-induced deterioration of the economic situation”) was operationalised. This independent variable is measured using a four-item self-perceived scale with a Cronbach's alpha of 0.66 (which is adequate according to Taber (2018)), which captures the current and the predicted (next 2–3 months) drop in revenues, as well as the current and the predicted (next 2–3 months) drop in employment. The revenue and employment drops were coded using a binary variable (1 – drop in …; 0 – without changes or growth of …). To verify Hypothesis 2, self-estimation of survival probability was used. Respondents were asked regarding their predicted chances of survival, expressed on a 1 to 10 scale with 1 being the lowest probability of survival and 10 the highest (see Table 4).

Measurement model evaluation

Appendix 1 provides detailed information on the reliability of the measurement model. First, collinearity assessment involves the computation of each item's variance inflation factor (VIF). As a rule, VIF values above 5 indicate strong collinearity among the indicators (Sarstedt et al., 2017). In the model, all the VIFs calculated were below this threshold. In the next step, bootstrapping analyses of weight confidence were carried out. Weights inform how each formative indicator contributes to a given composite construct (Chin, 1998). In two cases, the weight confidences included a zero, which could suggest the removal of these indicators. Nonetheless, the authors decided to keep these in the model, taking three considerations into account. First, the removal of a formative indicator would imply the elimination of a part of the composite latent construct. As theory suggests, the elimination of formative indicators from a model should be deeply considered, where such practice is an exception (Sarstedt et al., 2014, 2017). Second, as suggested in Hair et al. (2016), indicators with nonsignificant weight should be eliminated if the loading is also nonsignificant. In the model, the results of all single indicators' loading confidences were significant. In other cases, only indicator deletion should be considered; hence, a decision was made to implement them in further calculations. Finally, lower loadings are frequently observed in empirical research, particularly when newly developed scales are used (Hulland, 1999). This observation is in line with the study and is partly connected with the evaluation of the manner of survivability capital mobilisation.

Structural model evaluation

Construct reliability shows whether the indicators express (measure) the constructs adequately. To test the construct reliability in the model, composite reliability (CR) and average variance extracted (AVE) were calculated (see Table 5). In the case of CR, a value of 0.7–0.95 can be assessed as “satisfactory to good” (Sarstedt et al., 2014). In the model, all ratios fell within the interval recommended. An acceptable AVE is 0.5 or higher, which indicates that the construct explains at least 50% of its items' variance (Hair et al., 2019). The model presented in this paper met this requirement. Additionally, Cronbach's alpha coefficients are presented in Table 5, which depict the unidimensional, multi-item scale's internal consistency. The common threshold for sufficient values of Cronbach's alpha is 0.6 (Hair et al., 2006). All the latent constructs in the model exceeded this threshold as well.

The heterotrait–monotrait ratio of correlations (HTMT), which is an estimate of the factor correlation, was also calculated. To discriminate between the two factors, HTMT should be significantly less than one (Henseler et al., 2016). Adequate ratios associated with the model are presented in Table 6. All the HTMT in the model are statistically significant, while the variables analysed reached discriminant validity, as per the HTMT criterion.

Model assessment

Reliability of the structural model was assessed by the R² value, the f² effect size, the Q² predictive relevance, and SRMR, which reflect approximate model fit (Henseler et al., 2016). Additionally, to check collinearity, the VIF values of the latent constructs were examined. The model's SRMR is 0.08, which indicates an acceptable fit. Other ratios are shown in Table 7.

Considering the figures in Table 7, VIFs are below the threshold of 5 (Sarstedt et al., 2017), hence the reliability standards have been met. R-squares are relatively low. Following Hair et al. (2016), a low level of these ratios is acceptable by management research standards. In terms of the effect sizes, the f² for the structural model relationship can be assessed as weak. To calculate the cross-validated redundancy index (Q²), blindfolding was used, with a suggestion that if Q² exceeds level 0, the model has predictive relevance (Chin and Dibbern, 2010; Henseler et al., 2009).

Finally, the structural model was calculated, which was used for hypothesis testing (Table 8). Calculations were conducted using 10,000 bootstrap resamples. An interval that does not contain zero means that the structural path coefficient is significantly different from zero, at a confidence level of 95%. As such, the path coefficient is regarded as significant. The relationships observed are visualised in Figure 2.

First, the results for Hypothesis 1a, which examines the relationship between the crisis-induced deterioration of the economic situation and the internal survivability capital mobilisation was statistically significant. Hypothesis 1a has therefore been confirmed (p < 0.01). The path coefficient between the crisis-induced deterioration of the economic situation and the external survivability capital mobilisation was not significant. Hypothesis 1a has therefore not been confirmed.

The decrease in survival probability acts in favour of internal survivability capital mobilisation, at a statistically significant level. Thus, Hypothesis 2a is supported (p < 0.001). Assessment of survival probability as lower also motivates family members to mobilise external survivability capital. Hypothesis 2a has therefore been confirmed (p < 0.001). The above findings are discussed in the following section.