1 Introduction
Dismissal costs cause small and medium-sized enterprises (SMEs) to keep underutilized labour (Caballero et al.,
2013; Mallett et al.,
2019; Van Landuyt et al.,
2017). However, it is largely ignored that many SMEs can circumvent these costs by reallocating excess personnel through a business group’s internal labour market (ILM).
A business group can be defined as a set of legally independent firms linked together via ownership ties (Cainelli & Iacobucci,
2011). The literature on business groups has mostly focused on how business groups can add value to their members by functioning as internal capital markets (e.g., Belenzon et al.,
2013; Boutin et al.,
2013; Kabbach-de-Castro et al.,
2022). Little is known though about the benefits of articulating ILMs. Furthermore, the focus of the few existing studies is large corporations (Faccio & O’Brien,
2021; Jung et al.,
2019), thus neglecting that many SMEs may benefit from ILM access. Therefore, it is relevant to investigate the value-added of ILM access for SMEs given that (1) they are particularly constrained in adjusting labour on changing opportunities (Williamson,
2000), and because (2) business groups are also ubiquitous in the small business sector (Iacobucci & Rosa,
2005,
2010; Lechner & Leyronas,
2009).
Moreover, previous research does not consider the role of the bargaining power of labour in the feasibility and value-added of ILM access. Anecdotal evidence suggests that reallocation of redundant workers across firms of the same business group often involves substantial changes in employment conditions: employees might learn new skills, lose power within the organization, and see their employment, effort, and pay levels on less favourable terms (Capron & Guillén,
2009; MacKenzie & McLachlan,
2022). This suggests that employee bargaining power might be key in the value-added potential of ILMs.
The aim of this paper is twofold. First, we analyse whether SMEs with ILM access enjoy a labour productivity premium. Second, we investigate whether any potential productivity premium associated with ILM access depends on employee bargaining power. To empirically test these ideas, we collect a sample of 119,801 SMEs (639,675 firm-year observations) across 208 European regions (NUTS 2) covering 21 European countries with available data. Our research context offers an ideal environment for our objectives for two reasons. First, European countries are characterized by an unrivalled accuracy, reliability, and comparability of cross-country SMEs data. Second, it allows us to observe a high level of heterogeneity in the bargaining power of workers, varying across countries, subnational regions, and industries.
Our findings reveal that firms with ILM access have relative higher labour productivity than firms without ILM access. For example, when the average SME is embedded in a network (same subnational region-industry) composed of 10 (20) group firms, its labour productivity increases by 1.5% (3.5%) compared with SMEs without ILM access. In addition, we find that this productivity premium is higher in contexts associated with less employee bargaining power. Specifically, (1) in countries with more permissive labour laws, lower trade union density and collective bargaining power coverage; (2) in subnational regions with lower labour market tightness; and (3) in industries with lower job vacancy rates. For example, when regional labour market tightness is at the mean - 1 S.D. of its distribution—which is associated with lower employee bargaining power—, SMEs embedded in a network composed of 10 (20) group firms see their productivity increase by 3.5% (7%) (again, compared to SMEs without ILM access).
These findings contribute to the literatures on small business and business groups. In the small business literature, we document that ILM access is key to understand labour productivity at the firm level. Few studies consider that many SMEs are embedded in inter-firm networks, which allow them to reduce unproductive labour without using the external labour market. Therefore, future research that investigates SME productivity cannot neglect SME affiliation to these business networks at the industrial-regional level, namely, ILM access. With respect to the business group literature, we provide two main contributions that help to advance our knowledge about the value-added potential of group membership. First, our results suggest that group affiliation per se does not provide the advantages associated with ILM access. In contrast to other affiliation benefits that may apply to all companies in the group (Carney et al.,
2011; Holmes et al.,
2018; Locorotondo et al.,
2012), our findings indicate the need of a certain concentration of companies in the same geographic-industrial area for SMEs to reap the benefits of ILMs. Second, we advance our understanding of how the value-added of being affiliated to these interfirm networks is moderated by the environment. Specifically, we show that contexts in which the bargaining power of workers is higher, either institutionally or through scarcity, erode the ability of ILMs to redeploy unproductive labor and, thus, its associated productivity premium.
This article is organized as follows. Section 2 reviews the literature on business groups’ ILMs and formulates our hypotheses. Section 3 describes our sample and econometric methodology. Sections 4 and 5 present our main results and robustness checks, respectively. Section 6 discusses the empirical evidence and concludes.
2 Theory and hypotheses
Whether business groups create or destroy value for group-affiliated firms has been a central topic in the strategic management and industrial organization literature (e.g., Cainelli et al.,
2022; Carney et al.,
2011; Chu,
2004). One dominant explanation of their value-added potential is that business groups can circumvent market imperfections by establishing internal markets through which affiliates share resources with other sister companies in the group (Khanna & Palepu,
1997,
2000; Leff,
1978). Internal markets in business groups can be especially beneficial to SMEs because they are more resource-constrained than larger firms (Bongini et al.,
2021; Motta,
2020; Owalla et al.,
2022). For example, the extra resources provided by groups can help SMEs to innovate (Belenzon & Berkovitz,
2010; Guzzini & Iacobucci,
2014), increase exports (Eduardsen et al.,
2022; Tajeddin & Carney,
2019), attract financial resources (Cainelli et al.,
2020; Iacobucci & Rosa,
2010; Lechner & Leyronas,
2009) and grow (Bamiatzi et al.,
2014; Iacobucci,
2002; Iacobucci & Rosa,
2005).
Specifically, ILM access—i.e., being embedded in a network of sister companies that can reallocate labour due to their geographic and industrial proximity—endows SMEs with extra flexibility by offering them an additional avenue to reduce the pool of unproductive labour (Belenzon et al.,
2019a; Faccio & O’Brien,
2021; Huneeus et al.,
2021). While SMEs without ILM access can only adjust their workforce by hiring and firing workers in the external labour market, SMEs with ILM access can reallocate labour to other sister companies in the group network. This second option induces more flexibility to the extent that employment regulation applies only to labour readjustments that take place through external labour markets, not the ones inside business groups (Belenzon & Tsolmon,
2016; Cestone et al.,
2016). Specifically, the EU Directive 96/71/EC
1 allows affiliated companies to transfer employees without having to bear the costs of dismissal, such as meeting procedural requirements or taking specific actions when issuing the dismissal to the worker. This includes the length of the notice period, the amount of severance pay, and monetary compensation to the worker following an unfair dismissal, as well as the possibility of reinstatement after an unfair dismissal (OECD,
2020).
Consider for example two companies “A” and “B”. On the one hand, A is a company affiliated with a business group and with access to ILM, that is, with the possibility of reallocating workers to other companies in the same region-industry within the group if these workers are not being used in the most productive way in company A. On the other hand, B is a firm that does not enjoy the flexibility derived from ILM access. Now imagine that both companies receive an equivalent negative demand shock. This decrease in firm sales generates a situation of overstaffing, where both companies need to reduce their workforce to realign their labour cost structure with their revenues. Facing this sales decline, firm A enjoys more flexibility to reduce the excess labour generated (and therefore increase labour productivity or a productivity premium) if it can redeploy these workers towards other sister companies through ILM, where they can be fully utilized. The more companies that make up this ILM, the more opportunities firm A will have to reduce unproductive labour without incurring dismissal costs and reputational penalties linked to layoffs (Flanagan & O’Shaughnessy,
2005; Love & Kraatz,
2009; Zyglidopoulos,
2005). In contrast, it is more likely that company B, which does not have the flexibility granted by access to ILM, will retain (some) unproductive workers for longer because the only way it has to trim its workforce is to bear the full costs of layoffs (Autor et al.,
2007; Bassanini et al.,
2009; Caballero et al.,
2013), which include both layoff-related adverse organizational and human effects as well as dismissal costs (e.g., severance payments). If firm A benefits from an optimal ILM allocation of labour across group affiliates when facing changing economic conditions (Cestone et al.,
2016), we expect that firm B should exhibit more slack human resources and thus less labour productivity, ceteris paribus.
Redeploying employees from units where they are no longer needed to units where they are needed may also boost productivity by facilitating knowledge spillovers (Chang & Hong,
2000; Lee et al.,
2016). When employees move between different companies within a business group, they bring with them the knowledge and expertise they have gained in other sister firms. This can help to transfer knowledge and best practices across the group’s companies. Moreover, this efficiency-based mobility of employees can create opportunities for developing a group-wide shared knowledge base, where employees can gain exposure to different group units and develop a broader understanding of the group’s operations.
Interestingly, intra-group transfers are undertaken in a context of lower information asymmetry concerning workers’ quality. By contrast, external labor markets are fraught with information problems, making it very difficult for hiring firms to predict human capital returns (Chiang & Chiang,
1990; Greenwald,
1986; Jovanovic,
1979). This information problem is, in essence, a version of the classic “market for lemons” phenomenon (Akerlof,
1970). Human capital information asymmetries lead employers to offer lower wages as if they were hiring “lemons” to minimize adverse selection costs. In turn, wage reductions drive high-quality workers (“peaches”) to avoid this job-switching penalty and push them out of the labour market, perpetuating the lemons problem (Chadwick,
2017; Coff,
1997). As a result, firms have a harder time searching for “peaches” in the external labour market, which imposes additional hiring (search/training) costs. In this context, ILM access may enhance labour productivity by giving the option of selecting potential employees (from whom skills and abilities are known) that will better match the resources of the firm (Cestone et al.,
2016; Faccio & O’Brien,
2021).
Thus, our baseline hypothesis is that SMEs with the true option to reallocate unproductive labour to other companies in the group should exhibit higher labour productivity than firms without this flexibility:
However, the previous arguments to some extent ignore the inherent “stickiness” of human resources (Mishina et al.,
2004; Penrose,
1959; Voss et al.,
2008). Research results indicate that unused labour is difficult to redeploy, especially when employees’ bargaining power constrains managers’ decision-making and discretion. In terms of employment relationships, bargaining power is the capacity of a party to produce an agreement on its own terms (Dencker,
2009; Pfeffer & Salancik,
1978; Phillips & Sørensen,
2003). The relative power of employees varies with labour market institutions (as manifested in employment laws and collective bargaining agreements) and labour market conditions (tightness/slackness) (Dencker,
2009; Hansen,
1970; Vanacker et al.,
2017). For example, in a tight labour market—i.e., when unemployment is low—, employees have greater bargaining power as they face lower mobility costs outside organizations (search, bargaining, and switching costs) (Campbell et al.,
2012; Hansen,
1970). It therefore follows that business groups will face more difficulties to conduct productivity-enhancing reallocations when labour market conditions are tight. Through labour legislation, the bargaining power of workers is determined by the procedural requirements that the firm must follow before or when dismissing the worker, regulations regarding notice and severance pay, the framework for defining unfair dismissal, and the enforcement of unfair dismissal regulations (OECD,
2020). For example, employees have greater bargaining power when the amount of severance pay is increased, the possibility of reinstatement following unfair dismissal is higher, and when they are released from the burden of proof when filing a complaint for unfair dismissal.
Scholars have used relative bargaining power theory to advance our understanding on how differences in the distribution of power among key stakeholders influence restructuring decisions. For instance, Van Essen et al. (
2013) find that stronger employee rights restrict the ability of blockholders to pursue value-enhancing strategies such as reducing the pool of redundant employees. Capron and Guillén (
2009) find that stricter employment protection laws (EPL) regimes significantly constrain an acquirers’ ability to redeploy employees from the target. Another example is Vanacker et al. (
2017), who find that stronger EPL restrict managers’ discretion to reduce underutilized human resources.
In our context, power differentials determine if (and at what cost) business groups can reallocate redundant workers across group-affiliated firms. Intra-group reallocations of redundant personnel denote a clear shift in the employment relationship, as firms exercise their power not only to break the current employment relationship but also to change its terms considerably. When redundant personnel are transferred across group affiliates, they must often learn new skills, may lose power within the organization, and see their employment, effort and pay levels on less favourable terms (Capron & Guillén,
2009; MacKenzie & McLachlan,
2022). Consequently, if employee bargaining power is high, target employees are likely to oppose internal reallocation or demand a more aggressive compensation for their mobility—which limit ILMs’ ability to minimize unproductive labour.
In sum, if the productivity premium associated with ILM access depends on group’s discretion to reshape their pool of unproductive labour via intra-group reallocations, we expect to see this productivity premium attenuated when employees have strong bargaining power vis-à-vis the business group. Thus, we propose that:
4 Results
Table
2 contains the regressions that allow us to test H1 and H2. Model 1 includes only control variables. In Model 2, in addition to our controls, we add ILM access as an independent variable to test H1. In Models 3–7, we include two-way interaction terms between ILM access and the variables representing employee bargaining power to test H2.
Firm size | 0.059*** | 0.059*** | 0.060*** | 0.066*** | 0.059*** | 0.061*** | 0.058*** |
(0.003) | (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | (0.003) |
Firm age | 0.051*** | 0.051*** | 0.050*** | 0.047*** | 0.052*** | 0.051*** | 0.050*** |
(0.002) | (0.002) | (0.002) | (0.002) | (0.002) | (0.002) | (0.002) |
Sales growth | 0.215*** | 0.215*** | 0.213*** | 0.209*** | 0.208*** | 0.216*** | 0.215*** |
(0.003) | (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | (0.003) |
Capital intensity | 0.372*** | 0.372*** | 0.375*** | 0.387*** | 0.381*** | 0.373*** | 0.372*** |
(0.002) | (0.002) | (0.002) | (0.002) | (0.002) | (0.002) | (0.002) |
Num. group firms outside the region-industry | 0.008*** | 0.006*** | 0.006*** | 0.006*** | 0.007*** | 0.007*** | 0.006*** |
(0.000) | (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | (0.000) |
Leverage | -0.005*** | -0.005*** | -0.005*** | -0.006*** | -0.005*** | -0.006*** | -0.005*** |
(0.000) | (0.000) | (0.000) | (0.001) | (0.001) | (0.000) | (0.000) |
Industry labour productivity | 0.243*** | 0.243*** | 0.240*** | 0.248*** | 0.231*** | 0.244*** | 0.244*** |
(0.005) | (0.005) | (0.006) | (0.006) | (0.006) | (0.006) | (0.005) |
Regional GDP | 0.015*** | 0.014*** | 0.014*** | 0.012*** | 0.022*** | 0.020*** | 0.011*** |
(0.002) | (0.002) | (0.002) | (0.002) | (0.003) | (0.002) | (0.002) |
Regional competitiveness index | 0.132*** | 0.132*** | 0.131*** | 0.146*** | 0.136*** | 0.100*** | 0.092*** |
(0.005) | (0.005) | (0.005) | (0.005) | (0.005) | (0.005) | (0.006) |
ILM access | | 0.009*** | 0.009*** | 0.011*** | 0.012*** | 0.009*** | 0.009*** |
| (0.001) | (0.001) | (0.001) | (0.001) | (0.001) | (0.001) |
Bargaining power | | | 0.031*** | | | 0.018*** | 0.692*** |
| | (0.003) | | | (0.001) | (0.034) |
ILM access × Bargaining power | | | -0.006*** | -0.024*** | -0.040*** | -0.003*** | -0.101*** |
| | (0.002) | (0.004) | (0.006) | (0.001) | (0.011) |
Constant | 1.711*** | 1.717*** | 1.702*** | 1.783*** | 1.787*** | 1.665*** | 1.726*** |
(0.255) | (0.255) | (0.254) | (0.242) | (0.258) | (0.257) | (0.257) |
Observations Firms | 639,675 | 639,675 | 625,919 | 532,269 | 527,996 | 609,962 | 639,548 |
119,801 | 119,801 | 119,313 | 109,694 | 101,853 | 113,150 | 119,776 |
Wald chi square | 287,346.276 | 287,787.696 | 284,657.374 | 263,498.899 | 244,185.147 | 277,232.692 | 291,514.908 |
p-value | 0.000 | 0.000 | 0.000 | 0.000 | 0.000 | 0.000 | 0.000 |
R2 | 0.719 | 0.719 | 0.718 | 0.724 | 0.713 | 0.722 | 0.720 |
The results for the control variables provide some interesting insights. Firm size is positively related with labour productivity (b = 0.059, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in firm size increases labour productivity by about 1.6% for the average firm. Firm age is also positively related with labour productivity (b = 0.051, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in firm age increases labour productivity by about 1.8% for the average firm. Sales growth is positively related with labour productivity (b = 0.215, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in sales growth increases labour productivity by about 3.22% for the average firm. Capital intensity is positively related and has an important economic effect on labour productivity (b = 0.372, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in capital intensity increases labour productivity by about 20.4% for the average firm. An increasing number of sister companies outside the region-industry is positively related with labour productivity (b = 0.008, p < 0.001). If this variable increases from the mean - 1 S.D. to the mean + 1 S.D., labour productivity increases by about 2% for the average firm. Firm leverage is negatively related with labour productivity (b = -0.005, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in firm leverage decreases labour productivity by about -0.4% for the average firm.
With respect to the industry-level controls, industry labour productivity is positively related with labour productivity (b = 0.243, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in industry labour productivity increases labour productivity by about 8.3% for the average firm. Regarding the regional-level controls, regional GDP is positively related with labour productivity (b = 0.015, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in regional GDP increases labour productivity by about 0.4% for the average firm. The regional competitiveness index is also positively related with labour productivity (b = 0.132, p < 0.001). An increase from the mean - 1 S.D. to the mean + 1 S.D. in this index increases labour productivity by about 2.6% for the average firm.
As shown in Model 2 (Table
2), ILM access is positively related and has an economically significant impact on labour productivity (
b = 0.009,
p < 0.001). When all other variables are held at their means, our model predicts that SMEs embedded in a network (same subnational region-industry) composed of 10 (20) group firms, each of them sees their labour productivity increase by 1.5% (3.5%) compared with SMEs without ILM access. These findings support H1, which proposes that ILM access is associated with a labour productivity premium.
In Models 3–7 in Table
2, we test H2, which suggests that the labour productivity expected in H1 is higher when employee bargaining power is lower. In Models 3–7, we find that the coefficients of the interactions between ILM access and our proxies for labour bargaining power are negative and statistically significant. These findings suggest that the productivity premium associated with ILM access is higher (1) in countries with less stringent labour laws, lower trade union density and collective bargaining power coverage; (2) in subnational regions with lower labour market tightness; and (3) in industries with lower job vacancy rates. For example, as shown in Model 3 of Table
2, the coefficient of ILM access × EPL strictness (
b = -0.006,
p < 0.001) means that, when EPL strictness is at the mean + 1 S.D. of its distribution, the average SME embedded in a network of 10 (20) group firms sees its productivity increase by 1.4% (2.8%) with respect to its non-ILM access counterpart. In contrast, when EPL strictness is at the mean - 1 S.D. of its distribution, the average SME embedded in a network of 10 (20) group firms experiences a productivity increase of 2.2% (4.2%) with respect to its non-ILM access counterpart. Put differently, the lower the EPL strictness, the higher the productivity premium granted by ILM access.
Table 3
Percentage increase in labour productivity of the average SME with ILM access with respect to its non-ILM access counterpart, by level of employee bargaining power
EPL strictness | µ - 2σ | 2.4 | 4.8 |
µ - σ | 2.2 | 4.2 |
µ + σ | 1.4 | 2.8 |
µ + 2σ | 1 | 2 |
Collective bargaining coverage | µ - 2σ | 4.5 | 8.9 |
µ - σ | 3.3 | 6.5 |
µ + σ | 1 | 1.8 |
µ + 2σ | -0.2 | -0.6 |
Trade union density | µ - 2σ | 5.1 | 10.4 |
µ - σ | 3.7 | 7.6 |
µ + σ | 1 | 2.1 |
µ + 2σ | -0.2 | -2.1 |
Job vacancy rate | µ - 2σ | 2.8 | 5.8 |
µ - σ | 2.2 | 4.6 |
µ + σ | 1.2 | 2.2 |
µ + 2σ | 0.6 | 1.4 |
Regional labour market tightness | µ - 2σ | 6.1 | 12.2 |
µ - σ | 3.4 | 6.8 |
µ + σ | 0 | 0.2 |
µ + 2σ | -1.5 | -3.3 |
Overall, results displayed in Table
3 confirm that the moderating effect of employee bargaining power on the productivity premium associated with ILM access is not only statistically significant but also economically meaningful.
6 Discussion and conclusions
Employment regulation limits SMEs’ ability to reduce unused labour. Consequently, SMEs often retain unproductive workers whose wage exceeds their productivity. These constraints may be circumvented when an SME belongs to a business group and has access to its ILM. In this study, we investigate whether SMEs with ILM access enjoy a labour productivity premium. Moreover, because intragroup reallocation of redundant workers often involves changing employment conditions, we investigate whether the aforementioned potential productivity premium depends on employee bargaining power.
Using a comprehensive dataset that covers 119,801 European SMEs (2011–2019), we find strong support for our baseline hypothesis that SMEs with greater access to a group’s ILM—which occurs when there is a higher concentration of sister companies in the same industry-region—have higher labour productivity. This result suggests that the flexibility provided by ILM access to make efficiency-based labour adjustments that are not subject to EPL regulations enhances SME productivity. This efficient functioning of ILMs may further boost productivity by promoting knowledge spillovers within the group if redeployed employees bring with them valuable knowledge and best practices gained in other sister companies. Further, we find strong evidence that this productivity premium is attenuated by employee bargaining power. These findings are consistent with a growing research stream that highlights the role of business group affiliation in explaining SME behaviour and performance (e.g., Eduardsen et al.,
2022; Guzzini & Iacobucci,
2014; Lechner & Leyronas,
2009). We add to this literature by focusing on ILM access as an important element for understanding SME labour productivity. Our results, jointly with the prevalence of group-affiliated SMEs with ILM access (around 17% of our sample have at least one close sister company), underscore the need to consider the embeddedness within these interfirm networks when analyzing SME productivity.
In addition, we contribute to the nascent study of the value-added function of groups’ ILMs (together with Belenzon & Tsolmon,
2016; Cestone et al.,
2016, Faccio & O’Brien,
2021; Huneeus et al.,
2021; Jung et al.,
2019). The novelty of our results is that it is not belonging to a business group per se that matters; instead, it is the increasing option to redeploy employees among other sister companies that is important. In fact, in contrast to previous studies, we identify an asymmetry between group-affiliated SMEs with ILM access and group-affiliated SMEs without ILM access.
Moreover, no previous research has considered the bargaining power of labour in the feasibility and value-added of intragroup reallocations. The focus of previous works has been on how ILM access adds value in the presence of external labour market frictions, namely, hiring and firing costs imposed by EPL (Belenzon & Tsolmon,
2016; Cestone et al.,
2016; Faccio & O’Brien,
2021). Because intragroup labour adjustments are exempt from EPL, these studies have observed that stricter EPL put group-affiliated firms in a better position (compared to stand-alone firms) to reduce their pool of unused labour. Thus, it follows from this research that group affiliation increases labour productivity when EPL is stricter. Interestingly, our results show the opposite effect: the labour productivity premium is higher in countries with low EPL strictness. Nevertheless, we believe that this finding is not incompatible with the arguments laid out in previous studies. The flexibility to adjust labour without incurring in EPL penalties is the core mechanism through which ILM access adds value for member firms. However, in addition to high external adjustment costs, a high value in the EPL index is related with a stronger employee bargaining position, which may hinder intragroup transfers of redundant personnel. Therefore, our results suggest that, although avoiding EPL penalties is the core value-adding mechanism of ILM access, too strict EPL can neutralize the feasibility and value-added of this type of organizational market.
Finally, we answer the calls for studying the benefits/costs of group affiliation in developed economies, which have been typically neglected (Aguilera et al.,
2020; Carney et al.,
2011; Holmes et al.,
2018). It is generally thought that the ‘institutional voids’ (IV) perspective (Carney et al.,
2018; Leff,
1978)—one of the most widely used theories in the business group literature—is unable to explain (1) why business groups have not faded away or (2) why affiliated firms continue to have a more than acceptable economic performance in advanced economies (Cainelli et al.,
2022; Carney et al.,
2011; Belenzon & Berkovitz,
2008). This limitation justifies the lack of research on the consequences of group affiliation in such a context (Locorotondo et al.,
2012). Here, our study follows recent extensions of the IV perspective that propose a more general transaction costs story of business groups: groups’ internal markets can overcome the failures of arms-length market contracting in any type of economy (see, for instance, Belenzon et al.,
2013; Cainelli et al.,
2020,
2022). Leveraging on this idea, we challenge the conventional wisdom that business groups are gap-fillers only in less developed settings. In this way, our results reconcile the conflicting predictions of the IV theory with the dominant economic role of business groups in European countries (Cainelli et al.,
2011; Carney et al.,
2011; Colpan & Hikino,
2018).
6.1 Limitations and further research
Our paper has some limitations that open new avenues for future works. A potential limitation is that our ILM access variable (constructed with ownership data) is static because Amadeus database only reports the ultimate owner data for the last available year. However, we do believe that ownership patterns remain stable over time, as our sample is mostly made up of private firms and small business groups. Here, exploring how different labour market frictions affect the dynamics of group affiliation opens interesting paths for future research.
Relying on SIC codes to capture group ILM access can also be seen as a shortcoming. For instance, we consider that a group firm has no ILM access if it operates in the same region as other three sister companies but does not share the same two-digit SIC code with any of them. However, two group firms which are classified as unrelated according to the two-digit SIC code classification might be in fact vertically related and they could actually redeploy workers who have knowledge and skills that are applicable throughout the value chain of a product/service. Nonetheless, this potential shortcoming, if anything, should bias our analyses against finding support for the proposed hypotheses to the extent that we might be including in the group of firms without ILM access companies that indeed benefit from this type of organizational market.
A last weakness of our work is that, as much of the business group literature, firm’s affiliation to a group (which subsequently affects our ILM access measure) is identified on the basis of ownership and control relationships. However, even without ownership ties, some firms can be associated by multiple links, such as strategic alliances, franchising, subcontracting, and/or social relations (like family ties) through which they can coordinate to share resources and achieve mutual objectives (Granovetter,
1995; Lincoln et al.,
2017). Studying if the access to network-internal resources provided by such non-ownership ties substitute/complement external markets is a promising avenue for future studies. It would be similarly interesting to investigate the position of the affiliate within the group, especially in the case of pyramidal structures (Belenzon et al.,
2019b). Considering the affiliate’s position could unpack different degrees to which affiliates participate in, identify with, and are controlled by the group—which in turn reflects the extent to which they have more/less access to group-wide resources.
Finally, another promising avenue for future research would be to analyse which characteristics of ILMs generate a higher productivity premium. For instance, asymmetries in company size within the ILM network may affect intra-group reallocations. Workers may find it more appealing to be redeployed to a larger company in the ILM network but may resist being redeployed to a smaller company. Smaller companies are typically less attractive and secure employers than larger ones. Additionally, moving from a large to a smaller company may result in a loss of bargaining power, as large companies often have unions or employee associations negotiating on behalf of their members. As a result, in ILMs with asymmetries in company size, the redeployment of redundant employees may only be feasible in one direction (from small to large companies), whereas in ILMs between companies of similar size, it can occur between all companies in any direction, potentially increasing the productivity premium documented in this work.
6.2 Practical implications
Our work has some practical implications for managers, policymakers, and trade union representatives. Being able to swiftly adapt the workforce to changing economic conditions is more important than ever in the post-COVID-19 scenario, where record employee attrition rates and labour shortages disrupt SMEs everywhere. In this context, many SMEs may leverage their group membership to generate value in terms of a superior ability to adjust and maximize the efficiency of human capital. Specifically, our findings help managers by emphasising the conditions under which one can expect to see ILMs of business groups as a labour-productivity advantage. We provide evidence that internally reallocating workers is more feasible and valuable in contexts associated with low employee bargaining power. Therefore, managers in business groups should consider bargaining power when committing resources to develop formal policies that support/foster employee intragroup redeployment. For example, promoting a corporate culture that views group internal reallocations positively is more likely to pay off in subnational regions with lower labour market tightness and/or in industries with lower job vacancy rates. Other policy examples can be (i) including special provisions where redundant employees of a group firm “A” have the opportunity to apply for a job in a sister company “B” before group-external candidates are considered; or (ii) providing support/training programs if the candidates to the redeployment do not meet all of the skills and experience requirements for a different position.
Our results are also relevant for trade union representatives and policymakers interested in enhancing job security and employee welfare. Both should be aware of the potential benefits and drawbacks of inter-firm ILMs for the reallocation of redundant employees. On the one hand, access to a business group’s ILM may provide higher job security and opportunities for employees to acquire new skills and gain from their involvement in different dimensions of the group’s operations. On the other hand, conducting efficiency-based inter-firm reallocations may lead to changes in employment conditions, such as reduced bargaining power and less favourable pay and benefits. Trade union representatives and policymakers can use this research to negotiate with employers within business groups regarding the creation of ILMs and the rules that govern internal reallocations. It is important for all parties involved to advocate for policies that provide SMEs with greater access to ILMs, which may lead to higher labour productivity and better employment opportunities, thus leading to a win-win situation for employee and employer. However, it is equally important to ensure that the rights and interests of workers are protected during intra-group reallocations.
Acknowledgements
We would like to thank the editor, Rui Baptista, and two anonymous reviewers for their valuable feedback. Similarly, we are grateful to Julia Bodner, Rafael Corredoira, Manuel González-Díaz, Dongwook Kim, Gustavo Lannelongue, Vicente Salas-Fumás, Wolfgang Sofka, Chandra Thapa, and Luis Vázquez-Suárez for comments and suggestions on previous versions of this paper. We have also benefitted from the comments of participants at the IV Edition of the Brown Bag Seminars in Business & Economics of the University of Zaragoza, 14th International Accounting & Finance Doctoral Symposium (IAFDS), XXIII Seminario Luso-Español de Economía Empresarial (SLEEE), XXX International Conference of the Asociación de Economía y Dirección de la Empresa (ACEDE), 2021 Annual Conference of the British Academy of Management (BAM), 41st Annual Conference of the Strategic Management Society (SMS), and 82nd Annual Meeting of the Academy of Management (AOM).
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.