1 Introduction

A fundamental topic in strategic management is understanding the relationship between the presence and management of resources and the performance of firms (Penrose, 1959; Sirmon et al., 2007). Organizational structures play an important role in understanding resource-based processes because they facilitate or hinder the dynamic allocation of resources (Lovallo et al., 2020). It appears, however, that it is not only the presence of resources but also the fact that firms have more resources than what they strictly require for day-to-day activities that influence performance (Bourgeois, 1981; Pitelis, 2007). Indeed, excess resources, or slack resources, play multiple roles for firms (Cyert & March, 1963). Slack resources contribute to the stability of organizations by providing a greater capacity to absorb shocks and enhance the adaptability of organizations to their environment by facilitating the initiation of new, riskier business activities (Bourgeois, 1981; Vanacker et al., 2017; Carnes et al., 2019).

Although research on the role of slack resources is vast, not all forms of slack have been equally considered. Most papers on slack have focused on the role of cash or financial slack (Bradley, Shepherd & Wiklund, 2011a, Bradley, Wiklund & Shepherd, 2011b; Carnes et al., 2019; Lefebvre, 2021). More recently, scholars have considered other forms of slack, such as human resources (HR) slack (Paeleman & Vanacker, 2015; Vanacker et al., 2017). HR slack can be thought of as employees’ currently underutilized skills and work time. The benefits of HR slack become apparent if one imagines a firm in which all employees are overwhelmed with work. Let us assume that one employee in the firm becomes sick. If the sick employee’s tasks cannot be performed by someone else, operations will be disrupted.

Conversely, if some employees can temporarily dedicate some time to the tasks of a missing employee because they have underutilized work time, the firm’s operations may continue smoothly. As such, HR slack plays a stabilizing and adaptative role in the firm (Vanacker et al., 2017). The fact that the slack literature only recently considered HR echoes a broader observation regarding the role of HR in strategic management in general. Indeed, it took time for strategy scholars to fully accept that among all the crucial resources for firms to survive, grow, and perform, people are central to success (Wright et al., 2001). In particular, the articulation between HR and firm performance has largely been a “black box” (Becker & Huselid, 2006), likely because of an insufficient focus on “living persons” (Powell, 2014).

The present research proposes investigating the role of firms’ organizational structures in the relationship between HR slack and performance. Over the years, the slack literature has primarily focused on the role of slack in various environments and highlighted the role of institutional contexts and industry competition on the slack-performance relationship (George, 2005; Bradley et al., 2011a; Vanacker et al., 2017). Less attention has been given to the fact that firms are not homogeneous in their internal organization. To be clear about the positioning of this study, we acknowledge that the literature has explored the relationship between financial slack and performance in the context of SMEs (e.g., Bradley et al., 2011a; La Rocca et al., 2019; Lefebvre, 2021). However, to the best of our knowledge, previous studies have not considered whether the magnitude (or nature) of the relationship between HR slack and profitability changes depending on the organization’s size or complexity, which is the purpose of this paper. Our goal is to enhance our understanding of the role of HR slack resources by taking into account the great diversity of firms’ internal organizations.

As organizations grow and change, their internal structure and management practices must evolve to fit the current stage of the organization’s development (Flamholtz & Randle, 2007). The management of HR requires both complementarity between internal management practices and the organization’s structure (internal fit) and adequacy of the organization’s current stage of development and internal structure and management (external fit) (Baird & Meshoulam, 1988; Becker & Huselid, 2006). Typically, the size of an organization appears to be an important factor to consider. Smaller firms have a lower capacity to absorb external shocks and explore business opportunities because of their liability of smallness, making the presence of HR slack a particularly valuable asset (Aldrich & Auster, 1986; Hessels & Parker, 2013; Lai et al., 2016; Kiss et al., 2018).

In small- and medium-sized enterprisesFootnote 1 (SMEs), employees are generally multitasked, while in larger firms, employees are more specialized (Russo & Perrini, 2010). The nature and magnitude of the benefits associated with the presence of HR slack may well be different between smaller and larger firms. At the same time, firms should not be considered isolated entities when considering the resources to which they have access, as legal boundaries differ from economic boundaries (Dumez & Jeunemaitre, 2010). Firms often belong to business networks and strategic alliances that broaden the scope and nature of available resources, which impact the considered firm’s performance (McKelvie & Wiklund, 2010). The benefits of having access to partners’ resources depend on the compatibility and complementarity of the firm’s resources with its partners’ resources. For instance, the degree of dependence between customers and suppliers affects slack resources management (Donada & Dostaler, 2005).

In many countries, such as those in Western Europe, as firms grow, they adopt the form of business groups (BGs), which are sets of legally independent entities held together by controlling interest and coordination mechanisms (Khanna & Yafeh, 2005; Belenzon et al., 2013). From an organizational perspective, one of the key features of BGs is the presence of internal markets that allow BG entities to share resources. Firms affiliated with a BG have access to an internal capital market and an internal labor market that can allocate the resources they need to them or, conversely, take resources from them to allocate somewhere else in the BG. As BGs grow larger, their ability to shift resources increases dramatically and raises a greatly debated question regarding the benefits of BG affiliation (Carney et al. 2011; Holmes et al., 2018; Lovallo et al., 2020).

2 Theoretical framework

2.1 HR slack and performance

The strategic role of slack resources has been a topic of continued interest in the management literature since at least the seminal work of Cyert and March (1963) set the foundations of the behavioral theory of the firm (BTF). The BTF views firms as coalitions of stakeholders that bring resources to the firm to reach organizational goals. The strength of these coalitions depends on the personal (pecuniary or non-pecuniary) satisfaction each stakeholder takes from her participation to the coalition. Should a stakeholder be unsatisfied, she would leave the organizational coalition, which would threaten the survival of the organization. In other words, there are potential internal conflicts that need to be managed between the members of the coalition (Pitelis, 2007).

The classical view in the BTF is that holding slack benefits firms’ performance through four theoretical channels because it reduces intrafirm conflicts (Bourgeois, 1981; Pitelis, 2007). First, slack resources increase a firm’s buffering capacity regarding external uncertainty. Second, slack resources enhance cohesion within an organization by ensuring that each entity of the organization receives enough resources to fulfill its own goals (Cyert & March, 1963). Third, slack resources facilitate the exploration of new, riskier business opportunities. Fourth, slack resources are an incentive for stakeholders to stay in the organization to benefit from the abundance of resources.

However, alternative theoretical frameworks stress the limitations of holding high levels of slack resources. If slack resources are not used adequately, they may represent opportunity costs or even lead to agency conflicts because slack resources increase managers’ discretional space (Kim & Bettis, 2014). For instance, the abundance of HR slack is costly because of the increased payrolls, so the potential benefits of HR slack may well be offset by the permanent costs they induce. Additionally, resource-constraint theory proposes that firms are more efficient when they have limited access to resources because they have no other choice than to make good use of what they have (Baker & Nelson, 2005). A central idea in the resource-constraint theory is that entrepreneurs and managers reject the limitations imposed by the standard resources they have access to. To make firms grow, resource-constrained firms need to find innovative ways of using what is available to them, either through optimization or through entrepreneurial bricolage (Desa & Basu, 2013). Resource-constrained firms can even scale up their bricolage and optimization process when they grow older and larger and build competitive advantages (Bush & Barkema, 2021). Building on this argument, several authors have proposed that high levels of slack negatively affect performance by reducing entrepreneurial risk-taking activities (Bradley et al., 2011b; George, 2005).

These opposing theoretical views led to an empirical prediction that proposes the existence of a quadratic, inverted-U shape relation between slack resources and performance. The rationale for this prediction is that slack is beneficial for performance but only up to a certain threshold, after which any marginal increase in slack becomes detrimental to performance (George, 2005; Vanacker et al., 2017). In other words, behavioral theory of the firm arguments about the stabilizing and adaptative roles of slack explain the relationship between slack and profitability at low levels of slack, but at high levels of slack, the inefficiencies suggested by other theories are stronger. Many papers have provided empirical insights into this question, but the results are contrasted. The majority of the empirical literature investigates the role of slack resources by focusing on financial forms of slack such as cash-in-excess (George, 2005; Kim et al., 2008; Bradley et al., 2011a, b; Kim & Park, 2012; La Rocca et al., 2019).

More recently, several papers considered other forms of slack, especially HR slack. Vanacker et al. (2013) explore the extent to which professional investors such as business angels and venture capitalists help firms exploit HR slack and financial slack. Vanacker et al. (2017) investigate the role of institutional contexts, especially labor rights protection, on the performance impact of HR slack. Last, Paeleman and Vanacker (2015) study the combination of financial and HR slack bundles on performance. Overall, the literature reports a monotonic effect of HR slack on performance, the effect being negative in some cases (Vanacker et al., 2017) and positive in others (Vanacker et al., 2013).

To more clearly develop the role of the moderating factors, we first propose a baseline hypothesis regarding the role of HR slack in profitability. Following prior works, the first hypothesis is thus that profitability increases with HR slack but only up to a given level, after which opportunity costs and entrepreneurial laziness lead to a decrease in profitability.

Hypothesis 1

There is an inverse U-shaped relationship between HR slack and profitability.

2.2 Firm size and HR slack

At this stage, it seems important to note that HR slack and other forms of slack differ substantively. For example, HR slack is less fungible, less versatile, and less easily redeployed than financial slack because it takes time to reallocate employees to different tasks and missions (Mishina et al., 2004; Nason & Wiklund, 2018). At the same time, the ease with which HR slack can be redeployed seems to depend on the size of the organization. As noted by Lai et al. (2016, p. 115), SMEs “tend to operate with less organizational slack and require the minimum numbers to be able to operate effectively thus inhibiting opportunities to reduce their workforce”. Assume, for instance, that a firm needs to make up for the temporary absence of an employee, a common situation where the presence of HR slack is useful. Indeed, if a firm has HR slack, it has an underutilized workforce that can thus be mobilized to deal with this unexpected problem (buffering capacity of slack). In the case of an SME, the redeployment of HR is facilitated because SME employees are more multitasked and thus more able to take the responsibilities of the missing employee.

By contrast, large firms’ employees are more specialized in their skills and less likely to be able to compensate for the absence of their colleagues. As documented by Lorenz and Potter (2019), the workplace organization of SMEs greatly facilitates the sharing of responsibilities and task rotations between employees. It follows that SME employees are more prepared to take on (at least some of) their colleagues’ missions should there be an expected event (such as the temporary absence of a colleague). Our argument here echoes a more general view of SMEs’ employees and managers as more multitasked and thus more versatile than large firms’ employees and managers (Russo & Perrini, 2010; Iacobucci & Rosa, 2010). Thus, SMEs with HR slack simply have a more generic resource, which can therefore be used in many more configurations, than large firms with a comparable amount of HR slack. By contrast, SMEs that operate with a low level of HR slack are more vulnerable to unexpected shocks, which makes the value of slack greater for SMEs than for large firms.

While a lack of slack is likely detrimental to firm performance, an abundance of slack is problematic for SMEs. For instance, the abundance of slack negatively affects firm growth by reducing entrepreneurial attitudes and willingness to take risks and innovate in SMEs (Bradley et al., 2011a). This idea is core to the resource-constraint theory according to which firms that face important resource constraints find innovative ways to overcome their resource-deficit (Mosakowski 2002; Baker & Nelson, 2005; Bush & Barkema, 2021). Growth is typically a problem SMEs face that large firms do not or, at least, face to a lesser extent (Aldrich & Auster, 1986; Davidsson et al., 2009; Ben-Hafaïedh & Hamelin, 2022). The issue of mobilizing slack is thus stronger for SMEs than for large firms, which can capitalize on the “inertia” that provides profits from current activities (Agarwal & Audretsch, 2001). In other words, the abundance of slack is less problematic for large firms than for SMEs, which need to “fight to survive” and extensively use available resources (Bradley et al., 2011b). Since large firms have a broader scope and more diversified activities than SMEs, they also have more business opportunities to investigate and, thus, more options to reallocate employees.

Conversely, the skills associated with the presence of high levels of HR slack in SMEs likely become obsolete relatively fast (Paeleman & Vanacker, 2015). The mobilization of substantial levels of HR slack requires professional HR management, which is often underdeveloped in SMEs (Lai et al., 2016). In other words, opportunity costs arise when firms have too much HR slack, and these costs are likely more important for SMEs. These arguments suggest that at high levels of HR slack, the costs related to the presence of HR slack negatively affect the profitability of SMEs to a greater extent than they affect the profitability of large firms. It is important to note that this argument does not necessarily contradict the argument developed for low levels of HR slack. At low levels of HR slack, the most important function of slack lies in the buffering capacity it provides. At high levels of HR slack, the buffering capacity is less of an issue, and the role of HR slack as an opportunity exploration facilitator increases (Vanacker et al., 2017). Because they are small organizations, SMEs can use HR slack to face point-in-time uncertainty, making a minimum level of HR slack a vital resource. However, they can less easily mobilize important levels of HR slack to explore new business opportunities because of their lack of professional HR management.

Taken together, these arguments show that both the lack and the abundance of HR slack strengthen the relation between HR slack and performance, leading to a flatter relationship in larger firms than in SMEs. Figure 1 helps visualize hypothesis 2.

Hypothesis 2

The inverse U-shaped relationship between HR slack and profitability is flatter in large firms than in SMEs.

Fig. 1
figure 1

Expected relationship between HR slack and firm performance in large firms versus SMEs

This figure shows the expected inverse-U shaped relationship between HR slack and firm performance and the moderating role of firm size.

Full size image

2.3 BG affiliation and internal labor markets

Our previous hypothesis contrasts the situations of SMEs with those of larger firms. It is important to recognize that firms may have access to resources outside their strict legal boundaries. A small firm that belongs to a very large business network may well have access to more resources than a large standalone firm, and this access may, in turn, increase its performance (Kim, 2014). This paper takes this issue into account and focuses on a very common form of business organization, the BG.

In most countries, firms grow as BGs, which are sets of legally independent entities bound by common controlling interests (Khanna & Yafeh, 2005). A vast body of literature documents the presence of BGs in emerging Asian and South American countries (Khanna & Yafeh, 2005) and in Western Europe and Asia (Belenzon et al., 2013; Zhang et al. 2016; Yeh & Lin 2021). The key characteristic of a BG is the existence of internal markets that allow the allocation of strategic resources between affiliated firms.

Concretely, BG affiliation means that financial and human resources can circulate across BG-affiliated firms. The allocation of these resources is supervised by the BG head firm and depends on opportunities that can be exploited by affiliated firms (Lechner & Leyronas, 2009; Manikandan & Ramachandran, 2015). The presence of internal capital and labor markets in BGs is the starting point of a broad literature that studies the benefits and limitations of BG affiliation for affiliated firms. Intuitively, the presence of internal markets is beneficial to the performance of affiliated firms because of the easier access they provide to key resources (Carney et al., 2011). However, while BG internal markets can allocate resources to affiliated firms, they can also be used to siphon them away from minority shareholders of affiliated firms. The literature on tunneling shows, for instance, that in pyramidal BGs, majority shareholders expropriate minority shareholders through intragroup transactions that concentrate cash flows in firms where majority shareholders have high cash-flow rights (Bertrand et al., 2002). Overall, whether BG affiliation is beneficial for performance remains a controversial issue (see, for example, Carney et al., 2011 and Holmes et al., 2018 for literature reviews).

Furthermore, the literature on BGs has largely focused on the benefits and potential tunneling problems related to internal capital markets, neglecting the role of internal labor markets. Internal labor markets facilitate “the recruitment, training, and job transfers among affiliates” (Holmes, et al., 2018, p.140). As BGs are larger, more complex organizations than standalone firms, they provide more internal job opportunities and enhance employees’ skills acquisition. In Western Europe, where BGs grow through a gradual diversification process, the creation of new BG entities facilitates the recruitment of external managers who bring new skills to the BG and the internal promotion of promising employees to new jobs (Iacobucci & Rosa, 2010). In countries characterized by strong employment protection laws, the internal labor markets of BGs facilitate the redeployment of employees to the geographic areas and market opportunities where they are most needed (Belenzon & Tsolmon, 2016).

Overall, a central difference between standalone firms and BG-affiliated firms is that BG-affiliated firms have access to a larger pool of HR that goes beyond the legal boundaries of the firm. Thus, a lack or abundance of HR slack at the firm level is less important for BG-affiliated firms than for standalone firms because HR transfers are possible through the internal labor market (Belenzon & Tsolmon, 2016). Let us consider an example. If a BG-affiliated firm has a low level of HR slack, it can benefit from BG resources if an unexpected shock occurs. In practical terms, this means that if a given BG subsidiary has a point-in-time need for additional HR to address an employee’s illness or the temporary absence of a manager, HR can be transferred quickly from another subsidiary. This, however, depends on the attention span of the BG and the ability of the BG headfirm to detect and allocate the required resource fast (Belenzon et al., 2019; Lovallo et al., 2020). The value of HR slack held at the firm level is thus lower than it is in the case of standalone firms. The performance of BG-affiliated firms is thus theoretically less sensitive to the presence or absence of HR slack within the strict legal boundaries of the firm than is the performance of standalone firms. Indeed, the internal labor market of BGs brings mobility to BG HR and broadens the scope of resources to which an affiliated firm has access. This leads to the following hypothesis.

Hypothesis 3

The inverse U-shaped relationship between HR slack and profitability is flatter for firms affiliated with BG than for standalone firms.

2.4 The role of BG size

Because BGs are more complex organizations than standalone firms, it is important to consider the fact that BGs themselves are not homogeneous entities (Holmes et al., 2018). BGs differ in terms of size (Khanna & Palepu, 2000), diversification of activities (Iacobucci & Rosa, 2005), and nature of ownership (Masulis et al., 2011), among other factors. Concerning our focus on HR slack, one dimension of BGs is of particular interest, namely, BG size. Larger BGs have larger internal markets that can allocate more resources and do so more efficiently (Belenzon et al., 2013; Belenzon & Tsolmon, 2016). Therefore, it is easier for a larger BG to redeploy HR slack quickly when such resources are needed in another BG entity. Larger BGs indeed have a greater capacity to identify business opportunities that require the allocation of additional resources because of their multi-entity form (Manikandan & Ramachandran, 2015). Conversely, small BGs are not very different from standalone firms in the sense that they do not have different levels of resources. Our fourth hypothesis therefores propose that the moderating impact of BG affiliation on the HR slack-profitability relation is higher in larger BGs than in smaller BGs.

Hypothesis 4

The inverse U-shaped relationship between HR slack and profitability is flatter for firms affiliated with larger BG.

3 Methods

3.1 Sample and data source

The main source of data is the Amadeus database, which was proposed by Bureau Van Dijk. Amadeus is a leading database for European privately held and publicly traded firms that has been successfully used in both BG research (Belenzon et al., 2013; Belenzon & Tsolmon, 2016) and slack research (Vanacker et al., 2017). Concretely, Amadeus provides access to accounting, financing, and ownership information for a large set of European firms of all sizes. The paper focuses on a single country, France, because this setting is particularly suited to our purpose. First, France is a country that has one of the strictest employment protection law systems (Belenzon & Tsolmon, 2016). Thus, it is interesting to observe whether the theoretical flexibility provided by BG internal labor markets is empirically confirmed in a country with strong employment protection laws. Second, BG affiliation is very common in France, as more than half of French firms are affiliated with a BG (Belenzon et al., 2013; Hamelin, 2011). French BGs are organized as pyramidal structures with a head firm at the top of the pyramid that centralizes control over a set of affiliated firms (Masulis et al., 2011). French BGs represent a decision to grow the firm as a BG because of the benefits provided by this organizational mode that facilitate the management of growth and the allocation of resources (Lechner & Leyronas, 2009).

Additionally, the growth of BGs is the outcome of a diversification process through which new activities are gradually incorporated. The BG organization facilitates geographical extension, the diversification of activities in new industries, and differentiation in different segments (Iacobucci & Rosa, 2005). Creating new subsidiaries facilitates the creation of larger managerial teams and the incorporation of new talent into the BG (Iacobucci & Rosa, 2010). It should be noted that recent works by Hernández-Trasobares and Galve-Górriz (2020) also suggest that whether BGs grow by diversification or by specialization depends on the family nature of BG ownership.

Third, using a single country ensures that institutional differences between countries in employment protection do not drive the results. As shown by Vanacker et al. (2017), the relation between HR slack and performance depends on the strength of employment protection.

The sample is limited to French privately held firms because publicly traded firms have a more developed reputation and a greater social influence that facilitates the recruitment of employees and likely reduces the need to hold HR slack (Certo et al., 2009). Amadeus has a ten-year extraction limit, so the period for our dataset is the 2009–2018 period. Consistent with the literature on BGs, financial institutions are excluded from our sample (Belenzon et al., 2013). Our sample is made of 93,244 firms, including 64,337 BG-affiliated (69.0%) firms and 28,907 standalone firms (31.0%), and 425,246 firm-year observations. The criterion used to classify firms as standalone or BG-affiliated is presented in the section ‘Moderating variable’ below.

3.2 Dependent variable

Our dependent variable is profitability, which is measured as the gross profit margin (GPM). GPM is calculated as sales less cost of goods sold divided by total assets. The literature on the performance impact of slack resources has used various measures of profitability, such as ROA, EBITDA or sales growth (Bradley et al., 2011a, b). However, GPM is more appropriate to study the impact of HR slack because this profitability measure is not influenced by salaries and wages, which themselves depend on HR slack (George, 2005; Vanacker et al., 2017).

3.3 Independent variable

HR slack is our independent variable, and it is measured as the ratio of the number of full-time equivalent employees to sales less the annual industry mean ratio of the number of full-time equivalent employees to sales, a common approach in the literature (Mishina et al., 2004; Vanacker et al., 2017). The empirical analysis uses two-digit standard industrial classification (SIC) codes to classify firms in a given industry. Given the hypothesis of a quadratic relation between HR slack and profitability, the squared value of HR slack is also included in the regressions. An alternative measure of HR slack is used in the robustness tests.

3.4 Moderating variable

Three moderating variables are used in this study. First, we distinguish between SMEs and large firms, making size a moderating variable. Because our focus is on SMEs, a dummy variable “SME” that equals one if a firm meets the criteria of the European Commission is created: the firm must have fewer than 250 employees and either total assets below 43 M€ or sales revenue below 50 M€. Furthermore, the considered firm must not be affiliated with a BG. Our second moderating variable is BG affiliation, which is also a dummy variable that equals one if a firm is affiliated with a BG and zero otherwise. It is relatively easy to identify whether a firm is affiliated with a BG in Amadeus. Indeed, for each firm, Amadeus indicates the number of firms within the BG with which it is affiliated, this number being zero if a firm is standalone.

Additionally, Amadeus indicates the number of subsidiaries a firm has. Both criteria (number of firms in the BG and number of subsidiaries) are based on a controlling interest of at least 50% either directly or ultimately between the considered firm and the BG head firm. A firm is considered to be affiliated with a BG if the number of firms within the BG is higher than 1 or if the number of subsidiaries is higher than 1. Overall, the criteria used align with those used by Belenzon et al. (2013) and Belenzon and Tsolmon (2016), who used the same database. Third, the size of BGs is a moderating variable to test Hypothesis 4. BG size is measured as the natural logarithm of the BG’s consolidated total assets. This variable is directly provided by Amadeus. Specifically, two dummy variables named “small BG affiliation” and “large BG affiliation” are created. The variable small BG affiliation equals one if a firm is affiliated with a BG with total assets below the median of our sample and zero otherwise. The variable large BG affiliation equals one if a firm is affiliated with a BG with total assets above the median (27.0 M€) of our sample.

3.5 Control variables

Following the literature, several control variables for firm, industry, and regional effects and macroeconomic conditions are included through a set of annual dummies. Concerning firm effects, we include controlling for firm size, firm age, tangibility of assets, and other forms of slack, namely, financial slack, recoverable slack, and potential slack. Firm size is measured as the natural logarithm of the number of full-time equivalent employees. Firm age is the natural logarithm of the number of years since firm creation. Asset tangibility is the ratio of tangible assets to total fixed assets. Financial slack is cash and the equivalent to total assets less the annual industry mean ratio of cash and equivalents to total assets. Potential slack is the ratio of inventories plus accounts receivable less accounts payable divided sales less the annual industry mean ratio of inventories plus accounts receivable less accounts payable divided sales. Potential slack is the ratio of total financial debt (long-term debt plus short-term debt) to total assets less the annual industry mean ratio of total financial debt to total assets. These definitions of slack are in line with previous research (Vanacker et al., 2017). Given the nonlinear relation between slack and profitability, the squared terms of financial, recoverable, and potential slack are included in the regressions (George, 2005).

Concerning industry effects, industry GPM, the number of competitors, the mean size of competitors, and the intensity of industry competition are included as control variables (Vanacker et al., 2017). Two-digit SIC codes are used to classify firms into industries. The industry GPM is the annual industry GPM. The number of competitors is the natural logarithm of the number of firms that operate in the same industry in a given year. The mean size of competitors is the annual industry mean of the size of the firms operating in this industry. Industry competition is the Herfindahl-Hirschmann index calculated as the sum of the squared market shares of each firm in a given industry.

Concerning regional effects, a set of dummy variables is used to account for regional context. Indeed, it is well known that economic activity tends to concentrate in limited geographical areas. The concentration of economic activity facilitates access to strategic resources, especially specialized workforces and finance, which influence performance (see McCann & Folta (2008) for a review). Therefore, firms that operate in more rural regions where access to a specialized workforce is likely limited may find it more desirable to maintain higher levels of HR slack. We use the official classification of territories into 101 French administrative regions called “départements”. All variables have been winsorized at the first and ninety-ninth percentiles.

4 Results

4.1 Univariate analysis

Table 1 presents descriptive statistics for our variables and highlight differences between standalone and BG-affiliated firms. First, it is important to note that the majority of our observations correspond to firms affiliated with a BG (73.2%). This observation is not surprising for the context of Western Europe and is in line with previous works on BG affiliation (Belenzon et al., 2013). SMEs account for 24.9% of the sample because firms affiliated with a BG, even if they are small, are not considered SMEs. The average GPM is 33.3%, and standalone firms are more profitable than BG-affiliated firms. However, there is no statistically significant difference between standalone and BG-affiliated firms regarding the level of HR slack. This, however, is not true for other forms of slack, as affiliated BGs have less financial slack but more potential and recoverable slack than standalone firms. The average firm is 16.8 years old, but BG-affiliated firms are older.

Table 1 Summary statistics
Full size table

Table 2 displays a correlation matrix between the variables under investigation. To ensure that multicollinearity does not affect our data, the variance inflation factors (VIFs) are calculated. The highest VIF is 2.38, and the mean VIF is 1.30. These values are well below the commonly accepted threshold of 5 above which multicollinearity becomes an issue.

Table 2 Correlation matrix
Full size table

4.2 Multivariate analysis

4.2.1 The relationship between HR slack and profitability and the role of firm size

The multivariate analysis begins with a standard industry and year fixed effects model in which GPM is regressed on HR slack and the set of control variables previously defined. The squared term of HR slack is included because hypothesis 1 predicts a quadratic, inverse U-shaped relation between HR slack and profitability. Thus, the linear term for the HR slack coefficient should be positive, and the squared term should be negative. All independent and control variables are lagged by one year with respect to GPM to help establish the direction of causality.

The results are presented in Table 3. First, we run the regression on the entire sample without interaction terms between HR slack and our other moderating variables. There is no statistically significant relation between HR slack and profitability as captured by GPM, so we find no support for Hypothesis 1. Next, the interaction terms between HR slack and the SME dummy and between HR slack squared and the SME dummy are included. The interaction term between HR slack squared and the SME dummy is negative and statistically significant, which is a necessary, but not sufficient condition to establish a quadratic, inverse U-shaped relation. It is possible that the turning point lies outside the data range or that GPM does not increase faster at lower levels of HR slack and decreases faster at higher levels of HR slack (Haans et al., 2016).

Table 3 Impact of HR slack on profitability and moderating role of BG affiliation
Full size table

Our analysis relies on the Lind and Mehlum (2010) three-step criterion to ensure the shape of the relationship between HR slack and profitability. Using the command ‘U-test’ available for Stata 15, we use Sasabuchi’s test and calculate the bounds of the 95% confidence interval for the turning point using the Fieller method. The turning point is 3.048, the minimum value for HR slack is -0.02, and the maximum is 0.02, so the turning point is well outside the data range. In other words, while the relation between HR slack and GPM is quadratic, we observe only a positive relationship between HR slack and GPM in our data since the squared term for HR slack is negative, and no support is found for Hypothesis 1 in the case of SMEs. The positive effect of HR slack diminishes at high values of HR slack for SMEs. There is only partial support for Hypothesis 2 in the sense that the profitability of SMEs is more sensitive to HR slack than is the profitability of other categories of firms.

4.2.2 The role of BG affiliation

In this section, the focus is on the role of BG affiliation as a moderating factor of the HR slack-profitability relationship. Model 3 of Table 3 includes the interaction term between BG affiliation and the linear and squared HR slack terms. For standalone firms, the linear term of HR slack is positive but not statistically significant and the squared term is negative and statistically significant. Again, this result is necessary but not sufficient to establish a quadratic, inverse U-shaped relationship. The turning point is 2.04. The minimum value for HR slack is -0.02, and the maximum is 0.02, so the turning point is well outside the data range. In other words, there is only observe a positive relationship (and not a quadratic one) between HR slack and GPM in our data since the squared term for HR slack is negative. The positive effect of HR slack diminishes at high values of HR slack. Therefore, there is no support for Hypothesis 1 for standalone firms. The interaction term between HR slack and BG affiliation is negative and statistically significant, and the interaction term between HR slack squared and BG affiliation is positive and statistically significant. These results indicate that the relationship between HR slack and GPM is flatter for BG-affiliated firms, as expected in Hypothesis 2.

More importantly, the sum of the squared HR slack coefficient and the interaction between BG affiliation and squared HR slack is negative but not significant (p = 0.48). The sum of the linear HR slack coefficient and its interaction with the BG affiliation dummy is positive but not significant (p = 0.52). We use the command ‘nlcom’ to estimate whether the sum of coefficients is statistically significant. The interaction term between HR slack and BG affiliation is negative and statistically significant but the sum of the interaction and the linear term is negative but not statistically significant (p = 0.187). These results show that there is no statistically significant effect of HR slack on performance for BG-affiliated firms. In other words, our results corroborate Hypothesis 3 about the moderating effect of BG affiliation on the HR slack-profitability relationship. Figure 2 illustrates the results.

Fig. 2
figure 2

Relationship between HR slack and profitability (GPM)

This figure shows the relationship between HR slack and profitability on subsamples made of standalone firms and BG affiliated firms. The reported values are predicted values obtained with an industry and regional-fixed effects model when holding other variables at their mean.

Full size image

In other regressions (unreported but available upon request), we explored whether the relative locations of a BG-affiliated firm and the BG mother firm matter in the relationship between HR slack and profitability. Indeed, geographical proximity between a BG affiliate and its mother firm intuitively facilitates the transfer of HR slack and the monitoring of the affiliated firm’s activities. Comparable results have been obtained when considering subsamples of BG-affiliated firms located in the same city (or not in the same city) as the BG mother firm. We also investigated whether the fact that a BG is based in France or a foreign country leads to different results but again obtained the same conclusion: for BG-affiliated firms, the relationship between HR slack and profitability is not statistically significant, so having (or lacking) HR slack in the firm’s legal boundaries does not seem to play a great role in explaining profitability.

4.2.3 The role of BG size

In this section, hypothesis 4 is examined, according to which the size of the BG with which a firm is affiliated influences the relationship between HR slack and profitability. The results of the regression that includes two dummy variables designed to capture the size of the BG with which a firm is affiliated (small BG affiliation and large BG affiliation) are displayed in model 1 of Table 4. The squared HR slack term is negative and statistically significant, indicating that the relationship between HR slack and profitability is positive because the turning point lies outside the data range. The interaction terms between the small and large BG affiliation dummies and the squared terms of HR slack are positive and statistically significant. In other words, there is a flattening effect of BG affiliation on the HR slack-profitability relationship for both categories of BG. This flattening effect fully counterbalances the positive relationship between HR slack and profitability observed for standalone firms. Although the interaction term is smaller (as expected) in the case of a small BG (0.007) than in the case of a large BG (0.008), the difference between them is not statistically significant, so there is no clear support for Hypothesis 4.

Table 4 Role of BG size
Full size table

In unreported results, we investigated whether the role of BG size is different for small and large firms affiliated with a BG, but we did not identify statistically significant differences (the results are available upon request). We also investigated whether the BG degree of diversification influences our results because larger BGs are more diversified (Iacobucci & Rosa, 2005), but at the same time, diversification of activities reduces the transferability of skills and competences across subsidiaries (Iacobucci, 2002). The size of the internal labor market should not be mistaken for the diversification of activities. Using two different proxies for diversification, namely, the number of firms in the BG and the number of distinct industries in which firms affiliated with a BG operate, no difference between firms affiliated with highly or moderately diversified BGs were observed.

4.2.4 Robustness tests

This section discusses two important causality issues and proposes adequate robustness tests to provide greater credibility to our findings. Our first concern is related to the distribution of BG-affiliated firms versus standalone firms, as our sample contains many more BG-affiliated firms than standalone firms. To improve the balance between BG-affiliated firms and standalone firms, we use a matching technique that reduces potential sample selection bias and facilitates the estimation of causal effects. Specifically, we use the coarsened exact matching (CEM) algorithm developed by Blackwell et al. (2009), which improves balancing in the considered variables between treated firms (BG affiliates) and control firms (standalone firms). The use of CEM is becoming popular in management and has been used in recent papers (Bau et al. 2018). More specifically, we use the ‘k2k’ matching approach in which the algorithm creates strata based on a set of variables measured before the investigation period and then drops observations until each stratum contains the same number of treated firms and control firms. Our matching variables are firm size, age, the industry in which a firm operates and a geographical code that corresponds to one of the 101 French administrative departments. The choice of these variables ensures that geographic and industrial differences in the repartition of BG-affiliated firms do not influence our results. It is indeed possible that BG affiliation is more common in some industries or regions characterized by easier access to HR resources.

Additionally, matching size and age is important because the results in Table 3 highlight that there are differences between smaller and larger firms with respect to the impact of HR slack on profitability. Once the CEM algorithm is run, codes are generated that can be used to retain only the observations that have been matched in any regression design. Then, the same regressions as in the previous section are estimated. The results are displayed in Table 5.

Table 5 Impact of HR slack on profitability and moderating role of BG affiliation after the CEM procedure
Full size table

As seen in Table 5, our results regarding the moderating effect of BG affiliation on the HR slack-profitability relationship hold for the matched sample. We consistently observe no statistically significant effect of HR on profitability unless a distinction is made between BG-affiliated firms and standalone firms. While there is a positive relation between HR slack and profitability for SMEs and standalone firms, generally, this relation vanishes for BG-affiliated firms that have access to BG internal labor markets.

Our second concern is related to the way in which HR slack is measured as the difference between the observed ratio of the number of employees to sales and the industry annual mean of this ratio. It is possible that the industry mean is not a good reference to measure slack, as even within a given industry, firms have their own “desired level of human resources” (Wang et al., 2016, p. 510). In other words, it is possible that our measure of HR slack is partly endogenously determined by a firm’s decision. We therefore employ an alternative measure of HR slack based on a predictive value approach. We first regress the ratio of the number of employees to sales on a firm’s size, age, and the BG affiliation dummy for each industry-year in our sample. Then, the residuals of these regressions are used as measures of HR slack in our main regression. This approach is rather common in the slack literature (Wang et al., 2016; Vanacker et al., 2017). The results are presented in Table 6 and they are largely comparable to those obtained with the standard HR slack measure.

Table 6 Impact of HR slack on profitability and moderating role of BG affiliation using a predicted value approach
Full size table

5 Conclusion

5.1 Discussion of the results

This paper investigates whether and to what extent firms’ organizational structure influences the role of slack resources in performance. Using a large sample of French privately held firms, the results show that HR slack positively impacts the profitability of standalone SMEs (hypothesis 2) but had no observable effect for BG-affiliated firms (hypotheses 3 and 4). First, it seems important to highlight that no support was found for the standard prediction of a curvilinear relationship between slack and profitability (hypothesis 1). From a theoretical perspective, the view of a curvilinear relationship between slack and profitability reflects an attempt to reconcile the view of the behavioral theory of the firm and resource-constraints arguments, according to which a lack of resources stimulates performance (Vanacker et al., 2017). The positive relation between HR slack and profitability that we identify for SMEs is thus consistent with behavioral theory of the firm arguments but does not support resource-constraint arguments.

Interestingly, it thus appears that the buffering role of HR slack that helps attenuate external shocks and deal with uncertainty outplays the laziness-inducing behavior that an abundance of slack could cause (Bradley et al., 2011a). In other words, SMEs may simply never have enough HR slack to reach a level where the opportunity costs associated with the presence of HR slack would decrease performance. This interpretation is consistent with the idea that SMEs try to operate with the minimum level of HR slack (Lai et al., 2016). SMEs that can counterbalance the disruption costs caused by external shocks have slack, highlighting the resilience-enhancing function of slack.

We also predicted that the relation between HR slack and profitability would be stronger for SMEs than for larger firms (hypothesis 2). Our results show that the positive effect of HR slack is indeed stronger for SMEs than for larger firms. For larger, more complex firms, we see no statistically significant impact of HR slack on profitability. However, we do not observe that the negative effects of HR slack abundance on profitability are greater for SMEs than for larger firms because we do not identify any negative effect of HR slack. The apparent lack of sensitivity of the profitability of larger firms and BG-affiliated firms that we document is consistent with the idea that these firms broadly adopt levels of HR slack that are well suited to their needs. The professionalization and higher formalization of HR management in larger firms likely explain these results (Storey et al., 2010). Indeed, a higher degree of HR management formalization increases productivity and profitability and ensures an efficient workforce allocation, making HR slack less valuable even at low levels and reducing the risk of inefficient HR slack accumulation at high levels (Lai et al., 2017).

However, we did not observe that BG size played a role in the relationship between HR slack and profitability. We interpret this observation under the lens of the growth of a BG. BGs are formed to facilitate the management and allocation of resources in a business entity that becomes larger, more diversified, and more complex (Lechner & Leyronas, 2009). Most BG managers use the BG organizational form because it allows them to replicate the organizational design of a single small entity in multiple entities. In other words, a BG can be seen as a collection of multiple “small” firms tied together under the BG umbrella. This growth process means that the BG owner can manage each of these entities in the same way he or she would have managed one single entity (Lechner & Leyronas, 2009; Iacobucci & Rosa, 2010). Thus, when a single business entity (for which HR slack has a positive effect on profitability) grows into a BG and a new subsidiary is created, the initial entity cannot grow any further in a way that would increase performance. In particular, an additional workforce should be allocated to a new business entity to ensure that existing business entities do not become excessively large and hard to manage. This explanation is consistent with the idea that HR slack does not affect the profitability of BG-affiliated firms (HR slack mostly aligns with the exact needs of the subsidiaries, and the internal labor market helps deal with uncertainty and point-in-time needs for HR slack). This effect likely exists even in small BGs.

5.2 Theoretical and practical implications

The present study makes several contributions to the literature on slack resources and BG affiliation. Scholarly interest in slack resources is longstanding, and the last twenty years show an increase in theoretical and empirical works on the strategic role of slack resources. However, the role of organizational structure has been largely overlooked by previous studies. Distinguishing the case of SMEs from those of larger, more complex businesses shows that the role of slack resources depends on the size and the internal organization of the firm (BG-affiliated versus standalone firms).

Our results corroborate the findings of Vanacker et al. (2013), who also identified a positive relationship between HR slack and profitability in the case of entrepreneurial firms. However, these authors also report that the presence of professional shareholders such as venture capitalists magnifies the benefits of having HR slack. We observe that BG affiliation offsets the direct benefits of slack. Interestingly, these observations confirm that the presence of specific shareholders is an important factor that explains the allocation and use of slack resources within an organization, but BG cannot be simply considered a form of “professional shareholders”. Affiliation with a BG broadens the scope of resources a firm has and makes it necessary to go beyond the legal boundaries of the firm to take into account the complexity of organizations (Dumez & Jeunemaitre, 2010; Kim, 2014).

By considering the role of BG affiliation, this paper provides new insights into the role of slack by showing that HR slack matters for standalone firms but not for BG-affiliated firms. This result has important theoretical and managerial implications. Concerning strategic management research, it is important to acknowledge that firms are not isolated and that in most countries, firms belong to larger organizations that can provide resources. Organizational structures represent an essential driver of the resource-performance relationship. Being affiliated with a BG reduces the need to hold resources (Locorotondo et al., 2014) and the impact of resources on performance (Belenzon & Tsolmon, 2016; Lefebvre, 2021). The ability of BG and other comparable organizations (like conglomerates) to dynamically allocate and reallocate resources should thus more systematically be taken into consideration in resource-based research (Lovallo et al., 2020).

Our results indicate that the BG structure is associated with greater flexibility regarding redeploying HR slack and thus provides an adequate fit for strategic HR management. The adequate use of the flexibility provided by internal labor markets is an important part of a successful HR strategy. The literature on the growth of BG has regarded the BG structure as being suited to manage and allocate key strategic resources during the growth process (Lechner & Leyronas 2009; Iacobucci & Rosa, 2010; Manikandan & Ramachandran, 2015). This study provides an important empirical illustration of the strategic value of BG seen as an organizational form that facilitates HR management.

Regarding standalone firms, it is also interesting to notice that HR slack is positively related to profitability, indicating that SMEs benefit from the presence of more employees. Given the economic weight of SMEs in the European Union and abroad (European Commission, 2020), understanding the potential weaknesses of SMEs is important to limit the consequences of economic downturns. The presence of uncommitted HR is a factor of increased performance and thus likely a source of resilience for those SMEs that have HR slack. Public policies should thus carefully consider how to help SMEs recruit and train employees in such a way that HR slack is built. The results also suggests that SME managers should consider the fact that recruiting an additional employee should correspond not only to a perceived need but also to the unpredictable profits that will result from increasing the firm’s workforce. Recruitment difficulties and the challenges associated with HR management are seen by SMEs managers as major barriers to growth (Lee 2014). Adopting a BG structure is a growth mode that SMEs managers should consider to overcome these challenges.

5.3 Limitations and directions for future research

Of course, this paper has limitations. First, the role of BG affiliation in HR slack is likely different depending on the industrial context. As shown by Bamiatzi et al. (2014), BG affiliation is more beneficial during industrial downturns during which the allocation of resources by internal markets is a substitute for the buffering capacity offered by slack resources. We did not explore whether the moderating role of BG affiliation depends on the industrial context, and this question is an interesting avenue for future research. In particular, it would be interesting to study how SMEs mobilized HR slack during the Covid-19 crisis. While the benefits of holding slack are theoretically exacerbated during a crisis, it is not clear at all that HR slack benefited firm performance in a context characterized by lockdowns and massive disruptions.

Second, our evidence on the role of BG affiliation in the HR slack-profitability relationship is relatively indirect because we do not explicitly measure intragroup employee transfers. It would be interesting to monitor whether BG concretely transfers employees from one affiliated firm to another depending on the current business opportunities it wants to exploit. The ability of organizations to dynamically re-allocate resources is indeed essential to generate a competitive advantage (Lovallo et al., 2020). In other words, simply holding slack is not enough (Carnes et al., 2019).

Third, it is possible that the allocation of HR slack in BGs is not homogeneous. Some BG firms may be more likely to hold HR slack because of their role in the BG organization. For instance, the BG head firm likely centralizes slack resources to facilitate their allocation across affiliated firms (Lefebvre 2021). Future research could investigate the factors on which the repartition of slack resources in BG depends, like the attention capability of BG head firms (Belenzon et al., 2019). Despite these limitations, we hope that this paper will stimulate research on the role of BG affiliation in the use of slack resources.