Staying Alive! Entrepreneurial Human Capital and Resilience of New Ventures

1. Introduction

Survival can be argued to be the ultimate test of a firm’s resilience capacity. Over a firm’s lifespan, it will encounter a multitude of adverse and unexpected events, some of which might threaten its survival. These may stem from a wide range of events, for example, natural catastrophes, terrorism, pandemics, economic shocks at the micro- or macro-level, or internally generated problems at the firm level, that is, organizational fallacies, mismanagement, and fraudulent behavior. When such situations arise, firms mobilize their resources to mitigate the deleterious effects and return to normal, or at least survive (Malik et al. 2018). The outcome, however, differs substantially despite firms seemingly having similar prerequisites to counteract economic fluctuations or other disturbances. Hence, deepening our theoretical and practical understanding of the factors that promote firm survival should be of core value to policymakers, society at large, and organizations.

We theorize and provide evidence that having former entrepreneurs as employees provides abilities and skills that strengthen new ventures’ resilience, that is, their survival probabilities. Such entrepreneurial human capital (EHC) is defined as a set of abilities and skills that individuals acquire when starting and managing a new venture. As indicated by research on founders’ backgrounds, previous entrepreneurial experience appears to have a strong effect on firm outcomes, including growth, venture capital attraction, innovation, and productivity (Wright et al. 1997, Parker 2013, Cao and Posen 2023). Increased access to EHC has also been shown to improve firm-level performance, for example, productivity and innovativeness (Braunerhjelm and Lappi 2023, Lindbjerg and Vladasel 2025). Engaging in entrepreneurial endeavors is thus likely to provide individuals with skills and abilities that deviate from traditional human capital acquired through education or previous work experience, that is, specific entrepreneurial human capital is accumulated. We argue that EHC should enhance firm-level resilience.

Theoretically, we adopt an eclectic approach. First, we draw on contributions originating in the resource-based theory of the firm (Penrose 1959; Barney 1991, 1996), defined as a combination of differentiated knowledge and competencies of firms’ employees bundled with specific firm-level attributes. We argue that the liability of newness (Stinchcombe 1965) is associated with resource constraints for new ventures, thereby influencing their resilience capacity. Second, we bridge organizational resilience models (Annarelli and Nonino 2016, Branicki et al. 2018) with the literature arguing that small firms are better positioned to equip, and benefit from, employees with entrepreneurial competencies. Smaller firms have been claimed to adopt flatter and more informal organizational structures, which are argued to amplify entrepreneurial learning (Sorensen 2007, Elfenbein et al. 2010). Acknowledging that the “small firm effect” has been nuanced to account for organizational structure and employee heterogeneity (Kacperczyk and Marx 2016, Tåg et al. 2016, King et al. 2024), we take steps in the analysis to address similar concerns as we examine the relationship between EHC and survival of new ventures. Third, we also draw on research examining entrepreneurial traits, entrepreneurial learning, and individual labor market consequences associated with prior entrepreneurial experience (Manso 2016, Boudreaux et al. 2019, Feng et al. 2022, Kacperczyk and Younkin 2022, Mahieu et al. 2022). Combining these strands of theoretical constructs enables us to derive and test five interconnected hypotheses regarding the relationship between EHC and survival.

To empirically assess our hypotheses, we use individual- and firm-level data for Sweden to construct firm-level measures of each employee’s previous entrepreneurial experience (EHC) for the period 1997–2016. The analysis is restricted to new ventures, defined as being less than five years old. Our identifying assumptions rely on our ability to control for a wide range of observable characteristics in a population of new ventures across both the manufacturing and service sectors. To avoid our results being driven by observable differences between new ventures that hire former entrepreneurs and those that do not, our preferred empirical model uses the weights obtained from coarsened exact matching (CEM) in the estimations. The main finding implies that having former entrepreneurs as employees is strongly associated with a reduced probability of exit for new ventures. The results are robust to alternative estimation techniques, such as logit and hazard models, as well as to modifications in variable definitions, cut-off points, and alternative definitions of entrepreneurship.

The analysis contributes several new theoretical and empirical insights regarding firm resilience. Theoretically, we claim that a positive effect of EHC on new ventures’ survival resides in (i) a widening of the resource base to benefit from entrepreneurial abilities, magnified by learning from longer spells in entrepreneurship, (ii) an organizational fit related to size where smaller new ventures facilitate absorption and utilization of EHC, and (iii) a human capital fit where employees’ prior entrepreneurial experience have been acquired in knowledge domains that aligns with the competence requirement of the new venture. Empirically, we first complement traditional human capital variables at the firm level (i.e., education and prior work experience) by including EHC, defined in a way that enables its implementation in longitudinal analyses. The results are consistent with the hypothesis that access to EHC reduces the probability of exit for new ventures. As we narrow the analysis to two exogenous macroeconomic crises, we fail to detect a strengthening of the relationship between EHC and survival. This contrasts with our expectations but might be associated with specific crisis phenomena, that is, the origination of crises (e.g., in financial markets or technology), or because of other options becoming more attractive to the most able employees with EHC (Kahle and Stulz 2013, Cao and Im 2018). An extended analysis reveals that EHC appears to be correlated differently with survival rates across the two crises.

Second, the analysis provides suggestive evidence that learning from engaging in entrepreneurship, measured by actual years of entrepreneurial experience, is significant and positively associated with survival. Third, we argue that size, albeit imperfectly, can be used to approximate organizational fit and the extent to which EHC is further assimilated and utilized. Hence, smaller new ventures are shown to benefit most from leveraging abilities stemming from employees’ prior entrepreneurial experience. Based on these inherent organizational differences, we present a new boundary condition for examining EHC’s contribution to organizational outcomes, thereby further motivating the focus on the relationship between EHC and the survival of new ventures. Fourth, we define the correspondence between an employee’s acquired knowledge and the knowledge requirements of their current employer as human capital fit. We examine similarities regarding technological level (high- or low-tech) and industry experiences (related or unrelated) between employees who were formerly entrepreneurs and their current employers. We also implement measures of their education levels. These mechanisms are important for understanding how EHC is associated with the survival of new ventures. The results are partly in line with our hypothesis. Overall, the results imply the need to refine current theoretical constructs to enable further unbundling and understanding of firms’ resource bases and organizational resilience.

2. Theoretical Framework and Hypotheses

2.1. Organizational Resilience: Definition and Previous Contributions

Increased economic volatility, exogenous shocks, and perceived increases in high-risk events have sparked a surge of research across several disciplines aimed at refining, developing, and integrating resilience in various contexts (Lee et al. 2020). However, as pointed out by Linnenluecke (2017), ambiguities persist, and there is no consensus on how to exactly define or model organizational resilience. Despite varying definitions, a unifying thread refers to the objective of resilience, that is, to enhance the capacity to handle situations of uncertainty, discontinuity, and emergencies.

Our definition of resilience originates in Holling’s (1973) seminal work on ecological systems, that is, the capacity to overcome adverse events and persist. Holling extends his reasoning to the managerial level and claims that resilience “…does not require a precise capacity to predict the future, but only a qualitative capacity to devise systems that can absorb and accommodate future events in whatever unexpected form they may take” (p. 21, Holling 1973). More recent contributions have emphasized an organization’s ability to maintain functions, identify opportunities, build up and utilize resources during and after crises, and respond to sudden adversities, ultimately surviving (Williams et al. 2017, Doern et al. 2019). In the current analysis, we adhere to this more elaborate definition of resilience and relate it to previous entrepreneurial experience.¹

Thus, resilience models strive to capture how organizations can build the capacity to manage emergent internal and external adversities, thereby further strengthening their ability to mitigate potential crises (Sutcliffe and Vogus 2003, Annarelli and Nonino 2016). Redundancy and flexibility are key ingredients, with the former associated with slack and backup systems (Sheffi 2007), while the latter refers to a more general capacity for adjustment. Furthermore, resilience is often modeled as a multilevel structure involving an organization’s resources, routines, and processes, and the ways different levels align and interact (Gibson and Tarrant 2010, Bundy et al. 2017).

Organizational resilience models emphasize the role of management in introducing awareness and preparedness to maneuver crises (Ruiz-Martin et al. 2018). One critical task of managers is to mobilize, utilize, and enhance the abilities embodied in individuals within an organization (Malik et al. 2018). Hence, employees are (implicitly) allotted a considerable role in building resilience, albeit not highlighted to the same extent as organizational structure and management strategies.² Below, we embark on that observation as we discuss the links between organizational structure, “the small firm effect”, employees’ entrepreneurial abilities, and resilience. Hence, adopting an eclectic approach, we draw on different but interrelated fields of prior research to delve deeper into the intersection among employees’ knowledge base, the size and type of the employer organization, and the effects of EHC on firms’ resilience. This constitutes the basis for the five interconnected hypotheses that we present below.

2.2. Entrepreneurial Experience and Survival of New Ventures: Hypotheses

2.2.1. Heterogeneous Resource Bases, Entrepreneurial Traits, and Survival.

According to the resource-based view, identifying and exploiting a firm’s unique resources is key to strengthening its competitive position, which embraces its capability to build resilience to withstand and handle unexpected and adverse events (Penrose 1959; Barney 1991, 1996; Storey 1994). We stress that firms’ resource bases are heterogeneous, where new ventures are particularly resource-constrained due to the liability of newness (Stinchcombe 1965). Hence, as new ventures are confronted with a severe threat to their continuation, they cannot simply mobilize underutilized resources, making them more vulnerable to exogenous shocks and survival in general (Hamel and Valikangas 2003). An emergent literature analyzing how firms respond to large-scale exogenous shocks supports these theoretical inferences (Esteve-Pérez and Mañez-Castillejo 2008, Ambulkar et al. 2015, Williams and Shepherd 2016).

Overall, despite their alleged advantages, including greater flexibility and less bureaucratic structure (Dean et al. 1998), younger firms are disproportionately vulnerable to outside threats due to limited internal resources (Smallbone et al. 2012, Branicki et al. 2018). However, prior analyses have not considered the composition of firms’ resource bases in detail, for instance, the endowment of entrepreneurial human capital among employees, which has been shown to positively affect other organizational performance variables, such as productivity and innovation (Braunerhjelm and Lappi 2023, Lindbjerg and Vladasel 2025).

We combine insights from the resource-based view of the firm with the literature on entrepreneurs′ socio-cognitive traits. This vein of research emphasizes that features such as self-efficacy, risk-taking, handling uncertainty, adaptability, opportunity recognition, high creativity, and desire to perform set entrepreneurs apart from employees (McClelland 1961, Levine and Rubinstein 2017). According to the socio-cognitive theory (Bandura 1986, Lent and Brown 1996, Mauer et al. 2017), self-efficacy is a key predictor of entrepreneurial action, individual motivation, and persistence. It is claimed to be a trait of particular importance in dynamic, complex, and highly uncertain situations (Rauch and Frese 2007, Boudreaux et al. 2019), that is, typical abilities needed when firms encounter adversity. Employees endowed with such cognitive characteristics can be expected to be particularly important for resource-constrained new ventures. Others have shown that entrepreneurially oriented individuals self-select into new ventures, possibly adding to the firm’s pool of EHC (Roach and Sauermann 2015, Tåg et al. 2016). Hence, we argue that such traits, claimed to be embodied in entrepreneurs in previous research, largely coincide with those vital to building resilience.³ This constitutes the basis for our first hypothesis:

Recent entrepreneurship research has shown that exit strategies vary, and selection also occurs on that end (Arora et al. 2021). High opportunity costs of entrepreneurship could accelerate exit, which may be perceived as failure (Arora and Nandkumar 2011). The notion that all entrepreneurs who exit are failures has also been called into question, as preferences, individual threshold levels for performance, or proposals to acquire the firm can trigger an exit (Wennberg et al. 2010, Wennberg and DeTienne 2014). That would typically not be characterized as failures or a lack of resilience.

One way to address such ambiguities is to examine the effects of EHC during severe and exogenous crises. More precisely, the reasons for firms to exit during crises are expected to differ from “business-as-usual” times, when other motives may lead to exits. However, as firms struggle to survive during crises, entrepreneurial traits such as stress tolerance, the ability to handle uncertainty, risk-taking, and creativity should be particularly adequate for resilience.

To examine the relationship between EHC and new ventures’ survival during crises, we apply the definition of crisis suggested by Pearson and Clair (1998, p. 66), that is, “…a low probability, high-impact situation that is perceived by critical stakeholders to threaten the viability of the organization”. This definition does not distinguish between internally or externally generated shocks. Here, we adopt a crisis-as-event perspective (Williams et al. 2017), which is incorporated in our second hypothesis:

2.2.2. Entrepreneurial Learning and Survival.

The source of entrepreneurial abilities has long been debated as to whether they are innate, acquired through prior entrepreneurial endeavors, or developed through other learning mechanisms. Even though specific entrepreneurial traits can be expected, at least in part, to be innate, research also finds that learning occurs, for example, through the length of entrepreneurial experience, which further augments entrepreneurial abilities (Minniti and Bygrave 2001). Simply remaining in business for longer indicates the ability to handle various business situations that have arisen over the years.

An example of learning effects is the rich literature showing that serial or portfolio entrepreneurs are more successful than those who start a single venture (Politis 2005), suggesting that learning from experience matters when starting and running a firm. For instance, Shaw and Sørensen (2019) provide evidence that serial entrepreneurs have 98% higher sales and 49% higher productivity than novice firms. Also, Kerr et al. (2014) and Manso (2016) stress that experimentation and learning are vital parts of the entrepreneurial process, while Braunerhjelm and Lappi (2023) show that more experienced former entrepreneurs have a larger impact on firm productivity. These insights into entrepreneurial learning lead us to our third hypothesis:

2.2.3. EHC, Organizational Fit, and Survival.

Comprehending why EHC is expected to be particularly important for new ventures’ resilience requires closer scrutiny of the mechanisms that determine how entrepreneurial abilities are acquired, nurtured, and diffused. One strand of literature argues that differences in firms’ resource bases generate heterogeneous organizational structures with implications for employees’ knowledge accumulation, knowledge dissemination, and entrepreneurial learning (Foss et al. 2013, Pereira et al. 2020, Sorenson et al. 2021). More precisely, previous research suggests that organizational size is positively associated with the degree of specialization, routinization, and bureaucratization in firms (Sorensen and Fassiotto 2011, Cao and Im 2018, Sorenson et al. 2021). That contrasts with smaller and younger firms, where employees are more likely to be empowered with decision-making authority and delegated responsibilities. Such less narrowly defined tasks are claimed to foster networking, enhance knowledge absorption and opportunity recognition, that is, a broader set of “jack of all trades” abilities are acquired that supposedly facilitate the transition into entrepreneurship (Lazear 2005, Ireland et al. 2009, Elfenbein et al. 2010, King et al. 2024).⁴

Furthermore, it has been asserted that employees with a broader set of abilities are better equipped to learn from colleagues, some of whom may be former entrepreneurs, and to integrate this knowledge with their innate or previously acquired competencies (Nanda and Sorensen 2010, Sorenson et al. 2021). To the extent that skills are tacit, direct collaboration and cooperation may be necessary to absorb and exploit such abilities (Foss et al. 2013). To summarize, this vein of the literature aims to shed light on how organizational size influences its structure and entrepreneurial learning, which is likely to impact the effect of EHC.

However, this “small firm effect” has been challenged. For instance, Tåg et al. (2016) argue that the degree of hierarchy within firms, rather than firm size, matters for employees’ knowledge acquisition and entrepreneurial success. The effect is particularly pronounced for those in high-ranking positions. They attribute these findings to preference- and ability-sorting into less bureaucratic firms. Similarly, Kacperczyk and Marx (2016), although they recognize a small-firm effect, contend that larger firms offer superior career opportunities and better breeding grounds for employees to implement their innovative ideas, thereby fostering greater intrapreneurship. They provide evidence by examining the success of a subset of employees (tech-workers) in entrepreneurial endeavors after firms have dissolved, that is, when there are no internal opportunities left to exploit.

These partially opposing effects regarding organizational size, learning, and spawning of entrepreneurs are convincingly reconciled by King et al. (2024). They find strong support for a small-firm effect when all types of workers are included, which they argue stems from employees being exposed to and learning a broader set of skills. This is indicative of small firms having less bureaucratic and routinized organizations (Sorensen 2007). Simultaneously, due to lower expected profitability and more severe resource constraints, high-tech (STEM) workers in small firms are more inclined to start their own ventures than those in larger firms.⁵ Hence, two reasons appear relevant to the higher spawning of entrepreneurs from smaller firms: the learning and absorption of specific abilities by all workers and resource-based constraints on technical workers.

These prior contributions on size, organizational structure, and opportunities suggest that the effect of EHC should be most pronounced among resource-constrained and smaller new ventures. We refer to this size effect as an organizational fit that facilitates a firm’s absorption and integration of EHC.⁶ This leads to our fourth hypothesis:

2.2.4. EHC, Human Capital Fit, and Survival.

Even though former entrepreneurs who transition to employment may benefit firms by being more innovative, productive, and solution-oriented, such capabilities are likely difficult for potential employers to evaluate. In addition, entrepreneurs may be associated with certain stereotypes that hinder their chances of securing a job offer. We argue that new ventures are better positioned to assess the general capabilities of previous entrepreneurs and their specific knowledge bases, thereby yielding a superior human capital fit and a stronger impact of EHC.

Based on role congruity theory, Feng et al. (2022) and Kacperczyk and Younkin (2022) note that entrepreneurs are perceived as more difficult to manage, less team-oriented, risk-takers, and as challenging authority.⁷ These unobservable but perceived characteristics seem to influence former entrepreneurs’ job options and distort recruiters’ views of entrepreneurs’ fitness as employees, albeit sometimes moderated by counter stereotypes. One such moderating factor is the recruiter’s own entrepreneurial experience (Feng et al. 2022). Hence, the founders of new ventures share, at least in part, backgrounds similar to those of employees with prior entrepreneurial experience, which supposedly makes them better suited to evaluate such applicants’ skills and abilities. Founders are also less likely to hold a negative perception of entrepreneurial experiences. In addition, since former entrepreneurs can be expected to possess a diversified knowledge base shaped by the characteristics of young organizations, this should increase the likelihood of a good match and the exploitation of EHC the employee brings to the new venture.

Furthermore, to grasp the effect of EHC and its relationship to other individual characteristics, it is important to include additional facets of human capital fit and their bearing on firm survival. Higher education is expected to develop communication skills, increase aspirations, and strengthen complementary networks. Moreover, it is supposed to raise the threshold for entering entrepreneurship by providing a broader set of occupational opportunities (Gimeno et al. 1997). In addition, higher education signals a capacity for knowledge assimilation (Fox and Smeets 2011). These alleged effects of higher education are expected to positively affect the quality of EHC and firm survival. This is corroborated by evidence that employees who engage in more advanced tasks are more successful when transitioning into entrepreneurship (Elfenbein et al. 2010, Tåg et al. 2016).

Beyond education, human capital theory posits that individuals learn and accrue knowledge through experience across different fields, which influences their human capital fit as they shift between occupations (Campbell 2013). One such distinguishing feature of employees with entrepreneurial experience is the industry or sector from which their experience stems. According to Agarwal and Shah (2014) and Agarwal et al. (2020), individuals acquire knowledge through different occupations, referred to as knowledge corridors, which can be exploited in new ventures. Moreover, Dahl and Sorenson (2014) and Dahl and Klepper (2015) claim that entrepreneurs’ pre-entry knowledge and experience of an industry are decisive for firm performance. Similarly, Klepper (2010) and Bennett and Chatterji (2023) present findings that organizational heredity has dominated the evolution of firms, entrepreneurial endeavors, and agglomeration of economic activities. This suggests that it is important whether individuals’ entrepreneurial endeavors originate in related or unrelated industries and the degree of technological sophistication, for example, whether previous entrepreneurship occurred in high- or low-tech industries (Cao and Im 2018). Consequently, we expect the origins of employees’ entrepreneurial experience and prior knowledge sourcing to impact the individual human capital fit and organizational resilience (i.e., survival) in the fifth hypothesis:

3. Data, Empirical Model, and Descriptives

3.1. Data and Background

To test our hypotheses, we use register-based matched employer-employee data for Sweden provided by Statistics Sweden (SCB). The data combine several population registers on individuals and firms, available at the yearly level. We can link individuals over time to their employers or to the firms they own using anonymized, unique identifiers for individuals and firms. The firm information comes from the Statistical Business Register (FDB) and the Dynamics of Enterprises and Establishments Register (FAD), which contain accounting information on all firms registered in Sweden. For individuals, we have detailed information about all residents who are over 16 years old from the Longitudinal Integrated Database for Health Insurance and Labour Market Studies (LISA) database. The data allow tracking of individual characteristics, such as gender, age, marital status, residence, and educational attainment, as well as labor market information, including labor market status.

We include only new ventures in our analysis, defined as organizations younger than five years. The five-year cut-off is frequently used to categorize new firms, see for example, Calvino et al. (2015). In the Online Appendix, we elaborate on alternative cut-off points. Our sample spans 1997 to 2016 and comprises the population of privately owned single-establishment firms active in the manufacturing and service sectors. Hence, our sample spans 20 years, including two severe crises: the IT crisis of 2001–2003 and the Great Recession (GR) of 2008–2009.

3.2. Empirical Model and Descriptive Statistics

Firms in our sample are allowed to enter and exit the market freely, which means our empirical model uses an unbalanced panel for new venture i at year t. We estimate the following equation:

Our dependent variable ${\mathit{\text{Exit}}}_{it+1}$ is a dichotomous variable that takes the value 1 if the firm discontinues in t + 1 and 0 otherwise. The exit variable encompasses both true failures and those who choose to discontinue, even if not strictly pushed out of the market through liquidation. Notably, young high-technology firms seem to adopt exit strategies to sell their firm relatively soon after inception (Arora et al. 2021). We acknowledge the prevalence of such exit strategies but argue that the incentives to remain in the market dominate for most firms. Moreover, our data allows us to exclude exits through mergers and acquisitions. When estimating exit probabilities, a negative coefficient indicates a decrease in the probability that the firm will exit in the next period.

Our main variable of interest is $E{HC}_{it}$, representing the share of employees with prior entrepreneurial experience, that is, those endowed with entrepreneurial human capital. We calculate each employee’s entrepreneurial experience from labor market records dating back to 1993. An individual is defined as an entrepreneur in the labor market records if they have owned an incorporated business. We exclude sole proprietors as these are less likely to be growth-oriented (Levine and Rubinstein 2017).⁸ For each firm, we then aggregate the number of current employees who have previously been entrepreneurs to obtain the firm-level EHC share.⁹ The estimated $\phi$term captures the relationship between EHC and exit probability.

Following prior literature, we include a set of control variables, that is, the $\mathit{X}$ vector, that explain firm survival (Audretsch and Mahmood 1995, Stearns et al. 1995).¹⁰ These include variables from firms’ balance and income statements, as well as the context in which they operate. The first set of control variables included in the estimations is the natural logarithm of capital stocks (LnK) and the natural logarithm of total employees (lnL). Firms with larger capital investments can be expected to have greater resource leverage in the event of financial distress. Additionally, larger firms are less likely to exit, as they generally have superior access to internal and external resources, which is expected to be reflected in greater organizational slack. To account for the firm’s actual operating/financial performance, we also include the natural logarithm of the firm’s net sales (LnSales), since better-performing firms should be associated with a more affluent resource base and are less likely to exit. The stock of capital and the sales variables are calculated at 2016 constant prices in Swedish Krona.

Since human capital is an important determinant of firm performance, we include the Share of Highly Educated Employees, defined as the share of employees with at least three years of tertiary education. Furthermore, some studies find that employee composition matters for firm performance (Parrotta et al. 2014), which is controlled for by inserting the Share of Male Employees and the Share of Foreign-Born Employees.¹¹ Another frequently used control variable is labor turnover within the firm, which is supposed to capture the overall composition of employee tenure in the organization. To account for such differences in labor composition, we include the share of new employees between years t and t − 1 (Share of New Hires) and the share of employees who left the firm between years t and t − 1 (Share of Leavers). All employee composition variables are defined relative to the total employees in the firm. Controlling for these employee characteristics is important because it allows us to further isolate the relationship between former entrepreneurs and organizational survival.¹²

Market variables frequently include involvement in international trade. Penetrating several markets could mean risk diversification and better access to resources, but also increased vulnerability to international disturbances. Hence, we control for the imports and exports of intermediate or final products (Imports and/or Exports). Likewise, there are differences in firms’ survival probabilities across locations, as firms in more resource-abundant urban regions have better access to competencies and capital but may face higher competition. Thus, we control whether a firm is located in the largest and most dynamic Swedish regions, that is, the metropolitan areas of Stockholm, Malmö, or Gothenburg (Metropolitan). Even though we focus on new ventures (defined as less than five years old), baseline differences in survival probabilities persist as firms mature. Hence, our empirical models also include the firm’s age measured in years (Firm Age) as a control variable.

Finally, we include a set of fixed effects in our empirical model. Firms have different survival probabilities depending on product-market competition and other industry-specific factors. To control for industry-specific confounders $({\gamma }_k)$, we include industry (k) fixed effects at the two-digit industry level, following the Swedish Standard Industrial Classification (SNI). We also account for year-specific variation $({\gamma }_t)$ by including year (t) fixed effects.¹³

We first estimate a linear probability model as our baseline model, following Equation 1 (Audretsch et al. 2000). Firms that hire former entrepreneurs can differ in various observable and unobservable dimensions. To further improve covariate balance and address functional-form concerns, we implement Coarsened Exact Matching (CEM) to control for confounding bias arising from observable differences between firms (Iacus et al. 2012). More precisely, firms are matched on age, firm size, value added per employee, whether they are active in a high-technology or knowledge-intensive service sector located in a metropolitan region, and, finally, if the firm is part of a multinational corporation. We then include the obtained matching weights in a linear probability model, which is our preferred empirical model.¹⁴ The estimated coefficients in the models can be interpreted as the relationship of a particular covariate on the probability of a firm’s exit.

Matching on observables at the firm level does not eliminate all our self-selection problems, implying that our coefficients should be interpreted as correlational evidence. For causal inference, we would need exogenous variation in the share of employees who are former entrepreneurs, independent of the firm’s survival prospects. As we have longitudinal data spanning two decades, obtaining such variation becomes even more difficult. Hence, our identification relies on controlling for an extensive set of confounders using detailed register data and on CEM matching to account for any observable differences between firms that employ former entrepreneurs and those that do not. For additional robustness, we include firm fixed effects in our estimations to control for time-invariant firm-level confounders, which can at least partly account for unobservable characteristics that could drive our results and are correlated with EHC.

Our empirical models use firm-year observations to estimate exit probabilities. Table 1 provides descriptive statistics for our sample period. Note that, on average, 5.1% of employees in a new venture have entrepreneurial backgrounds, and around 19% of firms in the sample eventually exit. As we focus on new ventures, most such firms are also small, that is, they have, on average, seven employees. However, a few new ventures in the sample have experienced rapid growth within the first five years of operation, as evidenced by the extreme value for one firm having over 2,000 employees.

Table 1. Descriptive Statistics of Estimation Sample

3.3. Entrepreneurial Human Capital and Sorting

Even though we control for a large set of firm-level confounders in the estimations, there might be sorting of employees with prior entrepreneurial experience into specific types of firms that the matching or firm-fixed effects do not account for. We focus specifically on sorting at two ends. First, at the individual level, we want to evaluate the selection into and out of entrepreneurship. We are interested in how individuals who start firms differ from those who do not, and how entrepreneurs who exit differ from those who continue their ventures. There may be bias regarding the individual characteristics of those who return to employment, particularly in how these characteristics then relate to organizational survival. Second, we also want to evaluate the self-selection of former entrepreneurs into specific types of firms. For instance, former entrepreneurs may be more attracted by better-performing firms, which might then impact the analysis. This could introduce bias into our estimates, since employees and firms are endogenously matched.

Regarding the first self-selection, that is, the entry and exit of entrepreneurs, we provide individual-level descriptive statistics in Table 2. We differentiate individuals by whether they entered entrepreneurship between two consecutive years over the period 1997 to 2016 (Columns 1 and 2). Furthermore, for individuals who are entrepreneurs during the same period, we divide them into those who continue and those who exit in the next year (Columns 3 and 4). We also include information from conscription data that we can link to our labor market records, providing further evidence of differences across these dimensions and in individuals’ cognitive and noncognitive abilities. For the entrepreneurs, we also include firm-level information about their ventures.

Table 2. Descriptive Statistics for Selection into and Out From Entrepreneurship

Those who enter entrepreneurship earn more on average, are more often male, and are older and more educated with higher cognitive and noncognitive scores. This positive selection to entrepreneurship based on skills has been shown in the past (e.g., Levine and Rubinstein 2017). On the other hand, those who exit entrepreneurship earn slightly less, are older, and have more employment experience. Judging from observable individual characteristics, the differences between those who continue and those who exit are quite modest. When evaluating the success of new ventures, we see that those that exit perform worse in some respects, although the differences are relatively small. This suggests that exits are not entirely synonymous with failures because if that were the case, we would see even worse financial performance, for example, negative profits.

However, the selection of individuals into and out of entrepreneurship is unlikely to pose a significant problem in our empirical setting, provided individual characteristics are observable. In our empirical models, we examine different skill levels by education among former entrepreneurs and assess whether selection in these domains affects our estimates. However, entrepreneurs may also differ in many other hard-to-observe ways, which we cannot fully account for. If the selection of these traits completely biases our results, we should not find any evidence of learning, that is, differences in outcomes based on the length of entrepreneurial experience.

The second conceivable selection issue mentioned above concerns endogeneity in matching former entrepreneurs to firms. To evaluate this matching process descriptively, we present individual-level information for employees in new ventures with and without entrepreneurial backgrounds.

As shown in Table 3, employees with prior entrepreneurial experience are, on average, older, earn slightly more than the average employee, and have a similar level of education. The representative employee with entrepreneurial experience has spent about four years in entrepreneurship. There is evidence pointing toward sorting of former entrepreneurs into smaller firms, corroborating previous findings (Elfenbein et al. 2010, Sorenson et al. 2021). Otherwise, there is little evidence of sorting based on the firm’s productivity or other firm characteristics. There are additional observable individual-level differences between employees with and without entrepreneurial experience, for example, former entrepreneurs being older and more often male.

Table 3. Descriptive Statistics for Employees with and Without EHC in New ventures, Individual Level

The descriptive statistics indicate that former entrepreneurs are not randomly assigned to firms, consistent with our expectations. However, because former entrepreneurs are sorted into worse-performing firms, we tend to underestimate the EHC coefficient. Empirically, we include firm fixed effects as an alternative estimator to control for firm-specific time-invariant and unobservable characteristics, even though this does not resolve the sorting issue.

4. Results

4.1. Main Results – EHC and Survival

Based on our theoretical arguments, we empirically examine the extent to which employees with entrepreneurial backgrounds (EHC) are associated with organizational survival. To accomplish this, we first estimate Equation 1 using a linear probability model (OLS), adding industry fixed effects and the control variables stepwise, and then use our preferred estimator, a linear probability model with matching weights obtained from Coarsened Exact Matching (CEM). We report the marginal effects of our main variables in Table 4, where a negative coefficient on EHC means a decrease in the exit probability.¹⁵

Table 4. Main Results

Our preferred model (Column 4) shows that employing former entrepreneurs (EHC) is associated with a significantly higher survival rate (b = −0.043, p = 0.000). Furthermore, note that including the matching weights makes a small difference, showing a downward bias in the estimates in the absence of controlling for observable differences across firms that employ former entrepreneurs (Column 3 versus Column 4). Hence, the results conform with our first hypothesis, which posits a positive relationship between new venture resilience and the employment of prior entrepreneurs. These findings align with previous research showing that employees’ entrepreneurial experience has long-term positive outcomes for organizations (Braunerhjelm and Lappi 2023, Lindbjerg and Vladasel 2025). We extend this growing field by showing that these positive organizational outcomes also apply to resilience in new ventures.

4.2. Crisis, EHC, and Survival

Next, we analyze whether former entrepreneurs are particularly important for survival during crises. Using our longitudinal data, we focus on two specific large-scale macroeconomic shocks: the IT crisis of 2001 to 2003 and the Great Recession of 2008 to 2009. Sweden has been at the forefront of information technology for a long time, primarily due to the formation of clusters around the large anchor firms Ericsson and Telia. This meant that the Swedish economy was severely hit after the IT bubble burst in 2000. The downturn was even deeper during the GR due to Sweden’s disproportionately large banking sector. In 2009, GDP contracted by five percent, but recovered quickly in subsequent years (Berg et al. 2018).

To account for potential crisis effects, we use dummy variables that equal 1 in the crisis years. Furthermore, to test whether the effect of EHC differs across crises, we interact the crisis dummies with EHC. The results are presented in Table 5, where Columns 1 and 2 include a uniform dummy variable for both crisis periods, whereas Columns 3 and 4 distinguish between them.

Table 5. Results: Crisis Periods

As expected, the exit probability for new ventures is shown to increase during crises (about 1.2 percentage points). However, the interaction between EHC and the crisis dummy shows a positive relationship, indicating that firms with higher shares of EHC have a lower survival rate. Thus, we find no evidence that employees with entrepreneurial experience increase the probability of new ventures’ survival during crises, that is, the findings are not consistent with Hypothesis 2. The estimated total relationship of EHC on survival remains positive (−0.047 + 0.016 = −0.031), but overall, the results suggest that the relationship between resilience and EHC is weakened during crises. When we differentiate between the two crises in Columns 3 and 4, the negative effect on survival seems to be driven by the Great Recession. Hence, crisis-specific factors are likely to influence the impact of EHC.

Instead of estimating coefficients across the crisis and noncrisis periods, an alternative approach is to estimate the yearly effect of EHC on exits. As compared with Equation 1, all the variables are unchanged except for the yearly (t) variations (${\partial }_t$) in the entrepreneurial human capital ($E{HC}_{it}$) variable in Equation 2:

The point estimates of the EHC coefficients, that is, the ${\partial }_t$ terms, are derived from a linear probability model and presented in Figure 1.

The results corroborate our baseline results (Table 4) but reveal substantial annual variation in the estimated coefficients. Note that EHC is positively associated with firm survival during the GR crisis in 2009, but not in 2008.¹⁶ There are also a few post-Great Recession years where the marginal effects of EHC are not statistically significant, even though the coefficient has the expected sign. Similarly, during the IT crisis, EHC is positively associated with new ventures’ survival, but the estimated coefficient is insignificant in 2002. However, overall, the results align with our baseline estimations when we look at yearly variations. Overall, we find no evidence for the hypothesis that EHC is more strongly associated with survival during crises.

4.3. Learning from Entrepreneurial Experience

Access to longitudinal data allows us to examine whether the effect of EHC increases with the length of the entrepreneurial spell. If the longevity of entrepreneurship reduces the probability of exit, it suggests that learning occurs through experience and subsequently benefits survival.

In the individual records, we trace entrepreneurial experience back to 1993 and then aggregate employees’ total years of entrepreneurship for each firm to account for learning effects. We implement several versions of this variable to estimate potential learning effects. First, the share of former entrepreneurs with more than five years of experience are separated from those with fewer than five years. We include these two modified shares of firms’ entrepreneurial human capital in Table 6 (Column 1). Alternatively, we use the accumulated number of years of experience of former entrepreneurs as a direct measure of EHC (Column 2). Finally, the total accumulated entrepreneurial years are divided by the number of former entrepreneurs to assess whether the average experience of employees with entrepreneurial backgrounds matters (Column 3). These two latter measures provide the marginal effect of each additional year of entrepreneurial experience on resilience. If our estimated relationship in Table 4 is driven solely by entrepreneurs’ innate traits, we expect only minor changes when we include length of experience in the estimation.

Table 6. Results: Learning

We find that the estimated coefficients increase with the length of experience among former entrepreneurs, thereby amplifying survival for new firms (b = −0.079, p = 0.000) relative to those with shorter experience (b = −0.030, p = 0.000). The magnitude of the coefficient is more than twice as large for those with longer experience (Column 1), although less-experienced former entrepreneurs also have a positive effect on survival. Even though we cannot fully separate traits from learning, as correlational evidence cannot disentangle trait selection from experience-based learning, the results suggest that longer experience is consistent with upgrading entrepreneurial knowledge, in accordance with human capital theory. This relationship is further corroborated by the estimates of the marginal effects of each additional year of entrepreneurial experience in Columns 2 and 3. Hence, survival in new ventures is positively related to the length of entrepreneurial experience, presumably because entrepreneurs upgrade and refine their skills over time. Nevertheless, we cannot entirely dismiss that this is at least partly driven by selection.¹⁷ Taken together with previous findings on entrepreneurial learning, we provide suggestive evidence consistent with Hypothesis 3. Hence, it is unlikely that entrepreneurs’ innate traits entirely drive our results.

4.4. Organizational Fit

We have hypothesized that smaller new ventures offer a better organizational fit for former entrepreneurs, given their organizational characteristics and how knowledge is developed and implemented. This “small firm effect” implies that new ventures are expected to be associated with enhanced absorptive capacity and improved exploitation of EHC, which should be positively related to survival rates. To test this proposition, we examine how the marginal effect of having former entrepreneurs varies with the size distributions of the new ventures. We plot the marginal effects of EHC across different organizational sizes in Figure 2 below.

The results highlight that having former entrepreneurs as employees benefits organizational survival, however, only for new ventures with fewer than 50 employees. The importance of EHC appears to be inversely related to size: the smallest new ventures gain the most from access to it. We interpret the result as being consistent with Hypothesis 4, which posits that smaller new ventures provide an organizational fit that facilitates the utilization and incorporation of former entrepreneurs’ abilities. The result thus suggests that the double liability of being young and small makes EHC important for survival.

4.5. Human Capital Fit

We continue by examining how additional human capital attributes of employees with prior entrepreneurial experience contribute to organizational resilience, that is, whether their human capital fit is influenced by the type of prior experience or by level of education. Since skill demands and firms’ aspirations may differ considerably across different types of firms, we first categorize new ventures into two groups: one composed of firms belonging to the high-tech (manufacturing) or knowledge-intensive service industries according to the Eurostat definition, while the other group contains all remaining firms. We denote the former group as high-tech and the latter as low-tech.¹⁸

Next, we investigate whether the benefits of having former entrepreneurs as employees stem from their different technological backgrounds, that is, if their entrepreneurial endeavors occurred in high- or low-technology industries. Such differences in technological background are associated with different competencies that could impact the survival of new ventures in the high- and low-technology sectors, respectively. Admittedly, drawing on prior experience in the high- and low-tech industries means applying broad measures to capture differences in human capital. Therefore, we further distinguish the industry-specificity of former entrepreneurs’ knowledge base by separating experience obtained within the same related two-digit industry from that obtained in unrelated sectors. Finally, we consider the educational level of former entrepreneurs. We use information from the individual-level register data to categorize former entrepreneurs into those having three or more years of tertiary education (highly educated) and the rest (less educated).¹⁹ The results for these different educational categories are presented in Table 7 below.

Table 7. Results: Human Capital Fit

The results, as we classify new ventures into high-and low-tech sectors in Table 7, reveal some interesting sectoral differences. As regards the main relationship between EHC and survival, it is of similar magnitude irrespective of the sector to which new ventures belong (Columns 1 and 5: −0.049 versus −0.042). Hence, there is a general positive relationship between having former entrepreneurs as employees and resilience, which does not vary with the venture’s knowledge and technology requirements. However, when we disentangle the knowledge domains of the new venture and the former entrepreneurs, we find distinct differences.

First, for new ventures classified as high-tech, the estimated relationship for employees with entrepreneurial experience in this sector is twice that for those from low-tech sectors (−0.048 versus −0.024, Column 2). Conversely, for new ventures in the low-tech sector, only employees with entrepreneurial experience in the same sector are positively associated with survival (Column 6). Second, in the high-technology sector, employees with prior entrepreneurial experience in both related and unrelated industries are positively associated with survival, albeit there appears to be a sizable premium for experience obtained in unrelated industries, indicating significant industry spillovers (Column 3). In low-technology ventures, we find no significant difference in industry experience (Column 7), with both coefficients sizable and significant.

Lastly, the educational background of employees with EHC matters less. For new ventures in the high-tech sector, being less educated is more strongly correlated with survival than being highly educated. This could reflect that employees with higher education are more specialized, whereas the survival of a new venture might require a diverse set of skills. Or simply because of the combination of presumably low cash flow for new ventures and higher wages for employees with a tertiary education. Interestingly, the opposite pattern prevails for new ventures in the low-tech sector. In this case, the knowledge of highly educated employees may be particularly valuable, assuming this type of human capital is scarce in low-tech new ventures. However, these differences should not be exaggerated, given that all relationships are significant except for previous entrepreneurs with a background in high-tech sectors who are currently employed in low-tech sectors. Taken together, we interpret these findings as being partially consistent with Hypothesis 5.

4.6. Robustness

We present a series of robustness tests in the Online Appendix.²⁰ In Table A.3, we include multiple alternative estimation methods. To control for any firm-specific time-invariant confounders, we use a linear probability model with firm fixed effects. However, relying on firm fixed effects, which in our case includes a sample with many early exits among new firms, implies that identifying any positive relationship between EHC and survival relies on yearly variation in hiring former entrepreneurs, which is extremely restrictive. As an alternative to the matching weights, we also employ a k-2-k matching technique that pairs each firm with a single control firm. In addition, we include a logit regression since our dependent variable is dichotomous. Finally, we model survival with the Cox proportional hazards model. When adding firm fixed effects, the estimated coefficient becomes closer to zero and loses significance. Given that we are especially interested in new ventures, which, by construction, have short panels in the sample, these results should be interpreted with caution. Nevertheless, irrespective of the estimation technique used, the same general pattern appears as in our baseline results (Table 4).

In our main empirical models, we control for industry differences at the two-digit level. However, there can be systematic differences across industries that are not captured by industry dummies. To account for such effects, we re-estimate our empirical models separately for each one-digit industry (Table A.5 in the Online Appendix). The industry classification is based on the Swedish Standard Industrial Classification, which aligns with the NACE Rev. 2 classification system. Hence, at the one-digit level, we can evaluate broad sectors. The limited variation in the magnitude of EHC indicates that industry differences are relatively modest.

Our definition of a new venture is less than five years old. Although this measure is frequently used in previous literature, we investigate the sensitivity of our findings to alternative measures. As the cutoff point is changed to three and five years, respectively (Online Appendix, Table A.6), the results remain qualitatively consistent with our main findings.

Next, we evaluate the sensitivity of our results to the timing of the entrepreneurial experience (Table A.7 in the Online Appendix). In our baseline estimate, the entrepreneurial experience could have occurred at any time, dating back to 1993. However, entrepreneurial skills might depreciate over time. Hence, we consider only recent entrepreneurial experiences, defined as those that occurred within the eight years preceding the current employment (Column 1). Furthermore, a potentially larger problem concerns the definition of an entrepreneur, as it is both a complex concept and an occupation. We tackle such definitional problems in three complementary ways. First, we differentiate between employees with entrepreneurial experiences who were either employers or solo entrepreneurs (Column 2). Second, the longitudinal nature of the data allows us to differentiate among individuals who transitioned from employment to entrepreneurship (pull effects), those who transitioned from unemployment (push effects), and a residual group for whom we have no labor market information (Other). The latter group includes, for example, recent graduates or individuals who have been working abroad (Column 3). We also account for exit modes by former entrepreneurs (Column 4). Exit reasons could be firm closure (Closed), a merger or acquisition (Success), or the entrepreneur leaving the firm while the business continues (Nonclosed). The overall findings from including these different measures show some variation in the magnitude of the estimated coefficient, highlighting relevant boundary conditions, but overall, the results are not significantly affected.

5. Conclusion and Discussion

5.1. Summarizing the Results

Using detailed longitudinal data covering the population of private-sector nonagricultural new ventures over 20 years, our results show that having former entrepreneurs as employees (EHC) is positively associated with new venture survival. We argue that this relationship is due to the following overarching mechanisms: (i) entrepreneurial traits but also entrepreneurial learning that is accrued in entrepreneurial human capital, (ii) size matters and smaller new ventures provide a better organizational fit for EHC since their resource constraints tend to foster entrepreneurial, that is, less routinized and bureaucratic, organizations, and, (iii) an increased likelihood of a better human capital fit due to competencies possessed and acquired through previous entrepreneurial endeavors.

Altogether, the results suggest that entrepreneurial human capital, partly acquired through entrepreneurship and partly innate, can be utilized and diffused as entrepreneurs shift between occupations. Note that the positive resilience effects conferred to EHC do not seem to be driven by traditional human capital measures such as levels of education, work experience in related or unrelated industries, or whether entrepreneurial endeavor took place in sectors of different technological sophistication (Table 7). These human capital fit measures influence the magnitude of the coefficients, albeit not always in the expected way. Hence, this indicates that the effect cannot be reduced to specific knowledge requirements of an employing organization, even if we find evidence of large differences within and between technology sectors and knowledge domains.

The empirical analysis provides evidence in line with Hypothesis 1, which posits that having former entrepreneurs as employees increases organizational resilience. However, contrary to our expectations, the results do not align with Hypothesis 2, which posits a stronger relationship between EHC and survival during crises. Even though having former entrepreneurs remains beneficial for new ventures, the magnitude of this benefit is smaller than in noncrisis periods. As we further examine this relationship, we find that the results diverge between the IT-crisis 2001–2003 and the GR-crisis 2008–2009. For the former crisis, the estimates indicate a positive relationship with survival, whereas the opposite effect is observed for the GR crisis. As we examine the point estimates for each year, these differences across crises are confirmed. One explanation could be that encompassing demand shocks, followed by severe credit constraints and uncertainty, as was the case during the Great Recession, exert stronger negative effects on entrepreneurial activity (Fort et al. 2013, Clementi and Palazzo 2016) than more idiosyncratic and sectoral shocks (Archibugi et al. 2013, Amore 2015, Klimek et al. 2019). Previous contributions emphasize innovation as key to abating crises, which might have been easier during the more technology-intensive IT shock. Our interpretation is that crisis-specific features influence the impact of EHC, which warrants further scrutiny.

Our results are consistent with both Hypotheses 3 and 4, that is, longer spells of entrepreneurial experience are more strongly associated with survival, and the relationship between EHC and survival is more pronounced in smaller new ventures. Finally, the findings are partially consistent with Hypothesis 5, which posits that previous entrepreneurs with experience from knowledge domains that match their current employer’s, render a better human capital fit and a magnified effect of EHC.

5.2. Theoretical, Practical, and Policy Implications

Theoretically, a first contribution refers to the underlying mechanism that affects performance according to the resource-based view of the firm. By starting, managing, and running new ventures, individuals develop competencies that complement other forms of human capital and carry them across occupations. Aggregating such competencies into entrepreneurial human capital for each individual and firm, we enable a more fine-grained decomposition of the firm’s resource base. This should enhance the model’s accuracy, better illustrate real-world phenomena, and more clearly define which resources are crucial for new ventures’ resilience.

Second, current organizational resilience models address various dimensions of resilience, yet employees’ capabilities remain opaque. Instead, organizational structure, management practices, and similar firm-level characteristics have been stressed (Annarelli and Nonino 2016, Bundy et al. 2017). For instance, Ruiz-Martin et al. (2018) pinpoint three perspectives (attribution, outcome, tolerance) to capture organizational resilience and the implications for building resilient organizations. All three can be argued to be associated with human capital in general and EHC in particular. This follows from the abilities argued to be important for resilience and the similarities with characteristics claimed to be typical for entrepreneurs (Levine and Rubinstein 2017, Boudreaux et al. 2019). Gibson and Tarrant (2010) present a similar conceptual model of resilience based on three stages referred to as reactive, proactive, and adaptive. These concern the ability and preparations to handle unexpected crises. In each of these stages, EHC could either complement or substitute for new ventures’ (scarce) resource base, depending on their specific constraints and access to internal and external assets.

As we focus on organizational size and its conduciveness for generating and exploiting entrepreneurial abilities, we find that reciprocity prevails regarding organizational and human capital fit. In smaller, presumably less hierarchical new ventures, where a broader set of “jack-of-all-trades” skills among employees can be expected to prevail, the perception of former entrepreneurs is less likely to be negative. Hence, hurdles to employing former entrepreneurs should be considerably lower in new ventures. In addition, assuming that former entrepreneurs have managed their firm from its inception, we argue that they have acquired a diversified knowledge base stemming from the characteristics of smaller, nonroutinized, and decentralized organizations. Such knowledge bases should increase the likelihood of a good match between employer and employee, particularly for smaller organizations, as confirmed by our empirical analysis.

Thus, we add to existing theoretical constructs by suggesting that the relationship between firm size and entrepreneurship is not unidirectional. Rather, the small-firm effect not only implies that competencies are acquired that might spur entrepreneurial activities within the firm or lead to spinoffs, but also that small and new ventures provide a better organizational fit for job seekers with prior entrepreneurial experience. That should facilitate the absorption of new and complementary external knowledge, embodied by entrepreneurs who become employees.

Furthermore, as we extend the analysis to include more specific forms of entrepreneurial human capital, as suggested by their previous entrepreneurial experiences (high- versus low-tech, related versus unrelated), the positive effect on survival remains (with one exception). However, correspondence in technology level or degree of relatedness does not unambiguously strengthen the relationship between EHC and survival. The results relate to Kacperczyk and Marx (2016) analysis, even though their analysis addressed the opposite situation, that is, spin-offs from firms of varied sizes. They found a deficient fit between employees engaged in technical tasks and small firms, which they attribute to a constrained resource base and lack of internal opportunities in smaller firms. Our analysis offers a nuanced explanation, that is, such spawning of firms might stem from a mismatch between individual- and firm-level characteristics, as evident from the insignificant effects of previous high-tech entrepreneurs being employed by low-tech firms.

If smaller organizations tend to promote abilities and competencies often associated with entrepreneurs, whether innate or acquired, one might question whether employees with entrepreneurial backgrounds can enhance resilience. Yet, our empirical analyses provide robust evidence of a significant relationship between EHC and resilience. The interpretation is that there is a marked distinction between the skills acquired from starting and managing a firm, that is, having full responsibility for its viability, as compared with being employed in a context that enables the acquisition of skills typically associated with entrepreneurs. The latter captures experience and training that might augment the prospects for a successful launch of a start-up, but differ substantially from de facto setting up and running a new venture. This also aligns with the arguments presented by Roach and Sauermann (2015) regarding the distribution of skills between entrepreneurs and employees with entrepreneurial intentions.

Our analysis also suggests that organizations of varied sizes require different resources to build resilience. We present a new boundary condition for the impact of the previous entrepreneur’s contribution on organizational outcomes, which motivated us to restrict the analysis to new ventures. For those firms, our research has clear applied and practical implications. Hiring decisions should consider potential employees’ prior entrepreneurial experience, bearing in mind that the effects of EHC may vary with individual backgrounds and experience. Proper recruitment strategies should enhance firms’ resource base and strengthen their resilience (Malik et al. 2018). Firms that already have employees with entrepreneurial backgrounds should take steps to harness these competencies, that is, organize to integrate, understand, and absorb the capabilities stemming from prior entrepreneurial experience (Campbell et al. 2012). As observed in previous research, employers often seem reluctant to hire former entrepreneurs because of unobserved employee abilities and their presumed inability to adapt to organizational cultures (Kacperczyk and Younkin 2022, Mahieu et al. 2022).

Hence, the positive relationship between EHC and resilience in new ventures indicates that EHC can serve as a mechanism to substitute for internal resource constraints in the early stages of a firm’s life cycle. It can be viewed as a learning–by–hiring mechanism to foster resilience (Song et al. 2003, Braunerhjelm et al. 2020). As noted by Roach and Sauermann (2015), employees with traits similar to those of entrepreneurs, whom they refer to as joiners, complement entrepreneurs’ competencies. A similar mechanism might be at work for former entrepreneurs, providing a partial explanation for why its effect appears more pronounced in new, smaller ventures. The alleged greater flexibility and agility of new firms is another factor that could facilitate the absorption and integration of entrepreneurial skills contained in EHC. We conclude that EHC offers new ventures an opportunity to strengthen and complement their resource base, an opportunity that has not been adequately recognized previously.

Our results also provide useful information for policymakers, as employees endowed with EHC appear to reduce the probability of firms exiting the market and strengthen the resilience of new ventures. It suggests apparent positive macroeconomic effects, for example, reduced societal costs associated with unemployment and the deployment of productive resources, with viable firms more easily fending off crises. Since accumulating EHC requires individuals to start new ventures, policies should promote entrepreneurship by keeping barriers to market entry low, avoiding excessive regulatory burdens, and designing tax policies that encourage risk-taking and market experimentation. Moreover, mobility between occupations should be facilitated to enable individuals and firms to learn and acquire skills from different occupations.

Finally, an obvious shortcoming of the present analysis is that our results are correlational and based on observational data. More in-depth, qualitative analyses are required to understand exactly how EHC relates to other competencies and resources within firms. That should also embrace the interdependence between the impact of EHC on survival and the contextual and institutional differences across regions and countries in which firms operate. Deepening our understanding of potential alternative mechanisms would clarify how entrepreneurial experience leads to better organizational outcomes. In addition, since crises are likely to be heterogeneous, more comprehensive analyses that encompass multiple crises over longer periods would be valuable for better understanding the relationship between entrepreneurial human capital and organizational resilience during extreme events.