Financial performance and supply chain dynamic capabilities: the Moderating Role of Industry 4.0 technologies

Abstract

Industry 4.0 digital technologies are becoming indispensable for firms striving to enhance their supply chain capabilities and financial performance, but how these relationships play out in practice remains unclear. To address this issue, this study assesses the relationship between supply chain integration, supply chain agility, and financial performance from a dynamic capability perspective. Further analyses are conducted to establish whether Industry 4.0 digital technologies moderate the association between (a) supply chain integration and supply chain agility and (b) supply chain agility and financial performance. Findings based on the data pertaining to a sample of 274 Swedish manufacturing firms indicate that supply chain agility fully mediates the link between supply chain integration and financial performance. However, while Industry 4.0 digital technologies strengthen the effect of supply chain agility on financial performance, they do not moderate the relationship between supply chain integration and supply chain agility. These findings contribute to the ongoing debate regarding how digital technologies play a role in achieving competitive advantage in interplay with dynamic capabilities related to the supply chain. These findings are relevant for decision-makers, as they address the need for organisational adjustments beyond the mere introduction of Industry 4.0 technologies to fully reap their benefits.

KEYWORDS:

• Agility
• integration
• Industry 4.0
• financial performance
• dynamic capabilitiy

1. Introduction

Rapid globalisation has placed unprecedented pressure on the manufacturing sector, necessitating that firms keep abreast of technological advancements, as well as efficiently respond to changing government policies and customer demands (Kusiak 2021). Evidence has shown that this can only be achieved by drawing on capabilities related to e.g. agility, customisation, visibility, and transparency (Zhong et al. 2017; Aslam et al. 2018; Blome, Schoenherr, and Rexhausen 2013), which becomes even more challenging when producing small quantities of diverse products (Park, Son, and Noh 2020). To maintain competitive advantage and improve financial performance, firms need to continuously align their entire supply chain processes with the dynamics of their environment, as was aptly demonstrated by the current COVID-19 pandemic (Ivanov 2020). Digital technologies, especially those within the Industry 4.0 framework, are particularly important in this endeavour (e.g. Giffi and Gangula 2016; Li et al. 2018; Loonam et al. 2018) as their adoption helps mitigate risks and improve visibility, while also improving Supply Chain Integration (SCI) and Supply Chain Agility (SCA) capabilities (Prajogo and Sohal 2013; Flynn, Huo, and Zhao 2010; Braunscheidel and Suresh 2009; Schoenherr and Swink 2012). However, if firms are to fully benefit from Industry 4.0, they must make use of digital resources and capabilities to achieve seamless and transparent information flow and exchange (Büyüközkan and Göçer 2018).

Although several studies have confirmed the positive performance impact of supply chain capabilities (Fatorachian and Kazemi 2020; Gligor, Esmark, and Holcomb 2015; Kim and Schoenherr 2018; Rai, Patnayakuni, and Seth 2006; Swafford, Ghosh, and Murthy 2008), and SCI is widely considered as a key concept in the Operations and Supply Chain Management (OSCM) literature, its performance benefits, especially those related to the financial bottom line, lack consensus (Ellram and Cooper 2014). Thus, given the potential downsides of over-integration and the significant costs and resources required for fully implementing SCI (Gimenez, van der Vaart, and van Donk 2012), we aim to determine whether the potential financial benefits of SCI might be indirectly realised via SCA, as suggested in the extant OSCM literature based on the dynamic capability view (Braunscheidel and Suresh 2009; Matawale, Datta, and Mahapatra 2016).

Second, the link among various supply chain capabilities and their influence on financial performance has been relatively rarely studied. For example, although Zacharia, Nix, and Lusch (2011) and Gligor, Esmark, and Holcomb (2015) discussed the performance outcomes of SCA and SCI, neither study extended to the relationship between these capabilities. This was the topic of research conducted by Jajja, Chatha, and Farooq (2018), but these authors did not investigate the impact on financial performance. In sum, despite the managerial relevance of the link between supply chain resources, strategies, capabilities, and financial performance, this phenomenon has received limited attention in the OSCM literature (Dehning, Richardson, and Zmud 2007; Mishra et al. 2018; Shi and Yu 2013).

Third, prior research on SCA and SCI tends to focus on the role of organisational factors such as organisational culture (Altay et al. 2018), market-related factors such as environmental uncertainty (Wong, Boon-itt, and Wong 2011; Gligor, Esmark, and Holcomb 2015), and supply chain complexity (Gimenez, van der Vaart, and van Donk 2012). As a result, little is known about the influence of digital technologies, especially within the framework of Industry 4.0, on these capabilities (Tortorella, Giglio, and van Dun 2019). Still, available evidence indicates that Industry 4.0 digital technologies can facilitate timely access to pertinent information, such as that related to market demand, whilst enabling firms to detect and address issues in their external environment in order to make the sourcing, manufacturing, and distribution more efficient (Turkulainen et al. 2017). Moreover, prior studies have pointed to the important implications that Industry 4.0 may hold for SCI and SCA (see Dubey et al. 2018), warranting further exploration of their influence on financial performance.

Therefore, the aim of the present study is to shed light on the impact that digital technologies, SCI and SCA have on firms’ financial performance by answering the following research questions: (1) Does SCA mediate the link between SCI and financial performance? and (2) Do Industry 4.0 digital technologies condition the effect of (a) SCI on SCA and (b) SCA on financial performance? To answer these questions, a cross-sectional survey was conducted with a sample of 274 Swedish manufacturing firms in 2020. By taking a dynamic capability approach, this study intends to make three important contributions. Firstly, we underline that SCA is an important result of SCI without which manufacturing firms will struggle to make the best use of SCI. Secondly, we empirically investigate the contingent effects of Industry 4.0 digital technologies on the relationship between SCI and SCA as well as that of SCA on financial performance. Finally, we provide insights for executives in the manufacturing industry aiming to improve the outcomes of their dynamic capabilities by investing in Industry 4.0 digital technologies.

The remainder of this paper is organised as follows. In Section 2, our theoretical framework based on the dynamic capability perspective is introduced, followed by the development of hypotheses. Section 3 provides information on the data collection and analysis methods, while the main results are presented in Section 4. The theoretical and managerial contributions of our work are presented in Section 5 before concluding the paper in Section 6 by outlining the limitations of this study and proposing some directions for future research.

2. Literature

2.1. Theoretical background

2.1.1. Dynamic capabilities

The term ‘capabilities’ can be defined as ‘the capacity of deploying and combining resources towards a desired end’ (Amit and Schoemaker 1993, 35). Different types of capabilities exist, ranging from those pertaining to basic functions to dynamic, higher-order capabilities of strategic importance to the firm (Collis 1994). In rapidly changing environments, such as those currently encountered by manufacturing firms in result of changes in market structures and technologies (McAdam, Bititci, and Galbraith 2017), dynamic capabilities are essential for business survival and require the capacity for continually integrating, creating, and modifying external and internal competencies (Teece, Pisano, and Shuen 1997). In other words, to sustain their competitive advantage and superior financial performance, companies need to adjust their capabilities to address the changing market dynamics (Teece 2007). Dynamic capabilities are, then, a firm’s processes that use resources to match and even create market changes, and patterns of their effective use tend to vary regarding their reliance on existing knowledge (Eisenhardt and Martin 2000). Dynamic markets require the rapid creation of new knowledge, but managers have been shown to find it difficult to cope with uncertainty (Eisenhardt 1989). Here, Industry 4.0 digital technologies can be assumed to make a difference, as they support efficient processes for the analysis of information and implementation of decisions and can thereby speed up the creation of new knowledge.

The dynamic capability approach has been widely used in OSCM research (Halldórsson, Hsuan, and Kotzab 2015; Chahal et al. 2020). The findings yielded indicate that dynamic capabilities could be generated within a focal firm in collaboration with other partners, whereby operating routines are reconfigured to improve effectiveness. In supply chain management, this process necessitates acquiring the capabilities needed to adequately respond to the changing environment and market requirements. In the increasingly interconnected world, this implies the adoption of Industry 4.0 digital technologies to maximise the utilisation of firm’s resources, while exploring opportunities to share information with partners in the supply chain to better react to market changes and thus outperform competition (Akter et al. 2016; Wamba et al. 2020). Warner and Wäger (2019) underline the importance of dynamic capabilities, especially agility, in the digital transformation process. Agility, broadly defined as the ability to efficiently and effectively respond to market changes and turbulence, has become instrumental in firm’ survival (Altay et al. 2018).

2.1.2. Supply chain integration (SCI)

SCI describes the level of collaboration with key partners in the supply chain (Flynn, Huo, and Zhao 2010). According to Wiengarten et al. (2014), such collaboration improves the efficiency and effectiveness of resource utilisation by all supply chain members.

SCI can take many forms, ranging from internal (inter-departmental or inter-functional) to external (with upstream or downstream actors) integration (Frohlich and Westbrook 2001). However, determining an optimal level of SCI as well as establishing a universal measure of it can be challenging in practice (see Fisher et al. 1994). Integration across the supply chain can be measured by the degree to which flows of information, financials, and materials are exchanged between the focal firm and its partners (Rai, Patnayakuni, and Seth 2006). In this context, information exchange can be achieved by adopting digital information systems to connect supply chains in innovative ways (Kim and Schoenherr 2018). Rapid technological advances also allow firms to seamlessly integrate some internal functions with external partners in order to enhance their performance through efficient coordination of activities and material flows (Ellram and Cooper 2014). In the manufacturing industry, where rapid globalisation has resulted in highly complex supply chains extending beyond national borders, SCI is of paramount importance (Moyano-Fuentes, Sacristán-Díaz, and Garrido-Vega 2016).

2.1.3. Supply chain agility (SCA)

SCA has received attention in the OSCM literature due to its external focus and outcome orientation (Swafford, Ghosh, and Murthy 2006). SCA allows the supply chain members to respond to turbulent and volatile markets in a cost-effective and efficient manner (Blome, Schoenherr, and Rexhausen 2013) by flexibly adapting to changing market demands (Qrunfleh and Tarafdar 2014). Hence, it is regarded as an important ‘seizing’ dynamic capability (Aslam et al. 2018). SCA is fundamental for firms operating in unpredictable environments characterised by rapid fluctuations in demand and supply (Christopher and Towill 2001), as it enables them to adapt and respond to changes in customer expectations (Christopher 2000; Vonderembse et al. 2006). Most importantly, SCA can mitigate the risks imposed by unforeseen changes in the external environment, including demand for new product designs, improvements, functionalities, and increases in production costs (Al-Shboul 2017; Sangari and Razmi 2015). In this context, the speed of such response is an important component of SCA (Gligor and Holcomb 2014; Swafford, Ghosh, and Murthy 2006; Swafford, Ghosh, and Murthy 2008; Eckstein et al. 2015).

2.1.4. Boundary condition: Industry 4.0 digital technologies

Digital technologies, which in the contemporary manufacturing industry are closely related to Industry 4.0, allow for real-time communication and connectedness of objects (e.g. devices, machines, containers, and packages) that are embedded with software and sensors (Hahn 2020; Parente et al. 2020). Industry 4.0 technologies range from ‘base technologies’, including Internet of Things (IoT), cloud computing, big data and analytics, to ‘front-end technologies’, including smart working, manufacturing, product, and supply chains (Frank, Dalenogare, and Ayala 2019). Among other benefits, they enable firms to make use of digitised data to improve current processes but also provide new opportunities to create customer value (de Vass, Himanshu, and Shah 2018). The increasing use of Industry 4.0 digital technologies can be expected to have a great impact on supply chains and their performances in the future (Ben-Daya, Hassini, and Bahroun 2019).

Digital technologies have generally been operationalised as being things-oriented, internet-oriented, or semantic-oriented (Atzori, Iera, and Morabito 2010; Reaidy, Gunasekaran, and Spalanzani 2015). Things-oriented technologies involve the physical attributes of digital technologies, architecturally allowing the real-time capturing of data, which can assist in monitoring, tracking, and controlling supply chain processes. Internet-oriented technologies refer to the usage of a globalised network and its platforms, such as cloud technologies, for facilitating data transmission. Lastly, the semantic orientation of digital technologies involves the capability of hosting, processing, and synthesising data (Atzori, Iera, and Morabito 2010).

However, as the use of Industry 4.0 digital technologies is still relatively limited in the manufacturing sector, their relationships with SCI and SCA remain insufficiently studied (Reaidy, Gunasekaran, and Spalanzani 2015; Ben-Daya, Hassini, and Bahroun 2019). Nonetheless, drawing upon empirical evidence from other fields, it can be assumed that digital technologies might play an important role in sustaining firms’ competitive advantage through enhanced supply chain practices, as they can enable intelligent automatisation and more efficient usage of resources across the supply chain (Verdouw et al. 2016). They can provide the real-time information needed for efficient processes of analysing data and making decisions that are crucial for creating value through dynamic capabilities (cf. Eisenhardt and Martin 2000). Still, to fully benefit from Industry 4.0 digital technologies, supply chain members must address several important issues related to these processes, such as data management, data mining, data security, and data errors (Lee and Lee 2015). For example, as Industry 4.0 digital technologies generate large amounts of data, they require significant architectural storage capacity to handle and manage efficient processing (cf. Núñez-Merino et al. 2020). Similarly, manufacturing firms may lack the tools and analytical competencies required for data mining that would allow firms to move from the simple, experiential and iterative routines typically characterising firms’ capabilities in highly dynamic markets, towards the handling of more complex processes (cf. Eisenhardt and Martin 2000).

2.2. Hypothesis development

2.2.1. SCI and SCA

Customer requirements changes need to be communicated as quickly as possible along the supply chain in order to enable timely action. With frequent market changes taking place in a highly competitive environment, strategic collaborative relationships with supply chain partners can help manufacturing firms in adapting to these changes more efficiently (Ying et al. 2016). Empirical evidence indicates that when supply chain partners communicate, interact, coordinate, and share relevant knowledge in a timely and transparent manner, better decisions can be made in a shorter time (Verdouw et al. 2016). Such close collaboration, however, requires sharing of both formal and informal information (Tuan 2016), which can take the form of regular joint problem solving, planning, and goal-setting activities, thereby proactively preparing for eventual market shifts. Thus, building long-term relationships with supply chain partners will enhance SCA, which will, in turn, enable quicker reactions to changing market requirements (Matawale, Datta, and Mahapatra 2016). Similar arguments have been used in various supply chain contexts. For instance, Dubey et al. (2021) empirically investigated how different components of SCI impact SCA in humanitarian supply chains.

While both SCI and SCA assist in risk mitigation (Nishat, Banwet, and Shankar 2006), SCI is usually seen as a precondition to SCA. In the manufacturing industry, agility is achieved by integrating all inter-organisational processes, ranging from product design to customer services (Gunasekaran 1999). Consequently, SCA can only be realised by smooth internal coordination, as well as through close collaboration with key upstream and downstream actors. Therefore, SCI is a natural antecedent to SCA (Braunscheidel and Suresh 2009). Moreover, close working relationships among supply chain partners help firms prepare for and react to market shifts by, for example, sharing relevant data such as forecasts, delivery times, and stock levels. In line with prior research on the antecedents of dynamic capabilities and their causal relationships (Collis 1994; Schilke 2014), SCI can thus be conceptualised as a lower-order capability used to develop the higher-order dynamic capability of SCA, leading to our first hypothesis:

H₁: SCI is positively related to SCA.

2.2.2. SCA and financial performance

Prior research indicates that SCA is particularly relevant when demand predictability and supply stability are low, as is the case for innovative products with evolving supply processes (Sebastiao and Golicic 2008). SCA represents a higher-order dynamic capability that can generate value also by facilitating more effective ad hoc problem solving (cf. Bingham and Eisenhardt 2011). While developing SCA requires commitment and profound organisational changes (Jain, Benyoucef, and Deshmukh 2008), it enables companies to quickly and effectively adapt to market-driven changes in demand and thus can lead to better competitiveness and superior financial performance (Gligor, Esmark, and Holcomb 2015). According to Al-Shboul (2017), the ability to increase the rate of product customisation, improve delivery performance, and shorten the development and delivery times, improves firms’ market and financial performance. As a result of higher market share, firms can increase their sales which might lead to greater profitability (Sangari and Razmi 2015). Thus, in line with prior research suggesting a positive performance impact of dynamic capabilities (cf. Pezeshkan et al. 2016), SCA as a dynamic capability can be expected to improve the financial performance of firms (Halley and Beaulieu 2009). Recent studies have documented the impact of SCA on different financial performance indicators, including return on assets, return on investment, and return on sales (Inman et al. 2011; Gligor, Esmark, and Holcomb 2015). Accordingly, we hypothesise that:

H₂: SCA is positively related to financial performance.

2.2.3. SCA as a link between SCI and financial performance

Despite abundant research confirming the benefits of SCI, many firms struggle to capitalise on these benefits in practice (Vanpoucke, Vereecke, and Muylle 2017). SCI does not necessarily occur due to firms seeking economies of scale or increased efficiency but may evolve as a natural step in the pursuit of the company’s strategic objectives (Burgelman and Doz 2001). As attaining growth often leads to better financial performance (Rajaguru and Matanda 2019), firms can make use of their internal resources and capabilities to achieve that goal. Moreover, through strategic integration with their supply chain partners, firms can capitalise on their combined resources for the benefit of all involved parties (Wiengarten et al. 2014). Despite ample evidence supporting the benefits of SCI (Ellram and Cooper 2014), there is still limited consensus regarding how such results can be realised. However, several scholars have discussed the possible pitfalls and downsides of over-integration in supply chains (Flynn, Huo, and Zhao 2010; Gimenez, van der Vaart, and van Donk 2012). The recent poor performance of multiple supply chain actors due to the global pandemic and prior research pointing toward a ‘domino effect’ triggered by failing supply chain actors (Hertz 2001) further highlight the potential downsides of integration. These concerns are further augmented by the need for top-level commitment, trust, and investment into SCI without a guarantee of success. Chang et al. (2016) suggest that SCI could increase financial performance via other means, such as operational, relational, or strategic performance. Based on a meta-analysis of prior studies on the relationship between dynamic capabilities and performance, Fainshmidt et al. (2016) point at the importance of considering the hierarchical ordering among capabilities when studying their performance effect. Following this reasoning, we argue that the financial performance outcomes of SCI as a lower-order capability can be realised if SCI translates into other higher-level dynamic capabilities, here SCA. Therefore, we posit that:

H₃: SCA mediates the relationship between SCI and financial performance.

2.2.4. The moderating role of Industry 4.0 digital technologies

There is still a lack of consensus among scholars regarding the requirements, implementation, and potential direct benefits of Industry 4.0 digital technologies (Frank, Dalenogare, and Ayala 2019). However, they provide the supporting conditions for achieving ‘smart’ supply chains (Meindl et al. 2021). Considering that SCI and SCA are key supply chain initiatives to reach sustained competitive advantage, examining the underlying factors and contingencies influencing their fruitfulness is of utmost relevance. Both SCI and SCA require the support of contemporary digital technologies to fully leverage their potential. Meanwhile, we posit that the performance benefits from SCA will be intensified in the presence of such facilitating technologies.

While the relationship between SCI and SCA appears to be fairly straightforward, the circumstances under which SCI could lead to different levels of SCA remain to be established. According to Christopher and Towill (2001), achieving SCA is contingent on contextual factors in which the actors are embedded. In this regard, Gunasekaran (1999) argued that the use of Industry 4.0 digital technologies, tools and equipment, as well as integrated supply chain strategies, can be expected to be critical in reaching higher levels of agility. Dubey et al. (2018) called for investigating how resources and capabilities such as big data analytics could influence the achievement of SCA. Since Industry 4.0 digital technologies enable the provision and access to real-time data regarding supply chain entities and flows, they support the integrative nature of supply chain initiatives to be more agile and responsive to uncertainties. For instance, Chang et al. (2016) and Vanpoucke, Vereecke, and Muylle (2017) pointed out that leveraging on such technologies can enable simplifications and rationalisation of processes that support the sharing of financial resources, strategic planning, and just-in-time delivery between supply chain partners. Similarly, the use of such technologies can help improve the visibility and transparency along the supply chain by providing item identification, monitoring, and tracking, which are key enablers in achieving SCA via SCI (Dubey et al. 2018). Therefore, Industry 4.0 digital technologies have the potential to facilitate not only information creation, but also information sharing and integration (Ben-Daya, Hassini, and Bahroun 2019) which is crucial for the level to which SCI enhances SCA (Al-Shboul 2017). Furthermore, the digital technologies within the Industry 4.0 framework enable further autonomous decision-making. This, in turn, eliminates excessive complexities and shortens lead times. As a result, they can play a supporting role in improving ‘speed’ in reacting to uncertainties, which is central to SCA, especially, for firms that already have higher integrative potential or those which have incorporated system-wide SCI. Therefore, it can be argued that in the firms that have adopted Industry 4.0 digital technologies, the effect of SCI on SCA will be intensified. Hence, we hypothesise:

H_4a: Industry 4.0 digital technologies moderate the relationship between SCI and SCA.

The manner in which digital technologies are adopted is posited to be highly influential on the level of agility that can be attained, and thus on the firm’s performance (Lowry and Wilson 2016). Furthermore, Eckstein et al. (2015) maintained that the financial performance outcomes resulting from SCA are contingent upon other contextual factors. In firms undergoing digital transformation, Industry 4.0 digital technologies can assist in better capitalising on SCA while also enhancing firm-specific and supply chain-specific dynamic capabilities (Warner and Wäger 2019; Mikalef and Pateli 2017). In line with Industry 4.0 digital technologies, Rajaguru and Matanda (2019) investigated how firms relying on both big-data analytics and SCA could enhance their competitive advantage. Similarly, Dubey et al. (2019) identified a positive impact of big-data analytics on financial performance, while Tarafdar and Qrunfleh (2017) found support for the conditional impact of IT capability on the SCA relationship with performance. Since Industry 4.0 digital technologies reduce set-up and processing times, as well as costs associated with labour and materials, they support supply chain capabilities, such as agility, in achieving superior productivity in production processes, and hence, improved financial performance (Dalenogare et al. 2018). Meanwhile, according to Sangari and Razmi (2015), the level of IT-related capabilities for developing and sensing how to efficiently respond to changing requirements often differs between partners in the supply chain. Thus, Jia et al. (2020) call for further elucidating the role of digital technologies in the relationship between inter-organisational practices and financial performance in supply chains. Since Industry 4.0 digital technologies enable the real-time optimisation of supply chain activities, thanks to information provision and analytics, they have the potential to improve the outcome of SCA. However, the heavy investments and cost incurred due to implementing Industry 4.0 digital technologies might hamper improved financial performance, especially in the short-term (Dalenogare et al. 2018). Nonetheless, we argue that such technologies could improve financial performance, especially in those firms that have incorporated SCA. We posit that digital technologies facilitate benefitting from SCA by allowing firms to quickly sense changes in market requirements and adapt the supply chain accordingly in terms of information processing, product supply, and provision. In this context, Industry 4.0 digital technologies are instrumental, as they can enhance the outcomes of SCA. Therefore, we hypothesise:

H_4b: Industry 4.0 digital technologies moderate the relationship between SCA and financial performance.

Figure 1 illustrates the conceptual model used in this study.

Figure 1. Conceptual model.

3. Data and method

To test the above hypotheses, data was collected through a cross-sectional survey in which representatives of Swedish manufacturing firms participated. Swedish manufacturing firms are frequently studied in the supply chain literature as Sweden is considered a pioneer in focusing on supply chain activities and their influence on innovation (Beheshti et al. 2014). Moreover, Swedish firms are high-tech manufacturing forerunners with high degree of competence in digital technologies, such as Industry 4.0 (Thelen 2019). In order to ensure the sampling of firms possessing enough shared knowledge and capital to be able to invest in digital technologies, focus was given to manufacturing firms (SNI codes 10–32; NACE Rev. 2) with more than 50 employees. This resulted in an initial sample of 1727 firms, retrieved from the database ‘Retriever Business’ in January 2020. This database receives its data from Statistics Sweden and includes information on Swedish companies, including number of employees, contact information, board information and annual reports. The database also provides access to group structures as well as complete financial statements for public companies and allows to search for companies in a particular industry, region or size. Search results can be exported to Excel for further processing (Retriever 2020). Using the random stratified sampling method, 1000 firms were selected, as the dataset was normally distributed with respect to firm sizes. The data collection process, which took place in the Spring of 2020, commenced with developing the survey in the ESMaker cloud-based application, which was chosen due to its flexibility and simplicity. To evoke interest in the study and increase the participation rate, prior to distributing the questionnaire via email, all 1000 firms were contacted by phone. After several attempts at making this initial contact, 696 candidates answered the calls. Out of these, 191 firms were not interested in participating in the study due to either tight schedules, no current or planned investment in Industry 4.0 digital technologies, or other unspecified reasons. As a result, 502 firms received an email with the questionnaire, along with a brief explanation of the aim of the study and a confidentiality disclaimer. After two reminder emails, 286 complete responses were obtained, of which 274 were valid, yielding a 27.4% overall response rate, which is deemed adequate for performing structural equation modelling (Rajaguru and Matanda 2019). The majority of the respondents held managerial roles and were responsible for the supply chain, procurement, production, plant, and/or logistics operations. Table 1 below shows the sample characteristics, including firm size, age and standard technology intensity based on the SNI code 10-32 (Statistics Sweden 2017).

Table 1. Characteristics of the respondent firms.

		Frequency	Percentage
Technology intensity	Low technology	82	30%
	Low-medium technology	85	31%
	Medium-High technology	94	34.3%
	High technology	13	4.75%
Age	Less than 25 years	51	18.6%
	Between 25 and 50 years	121	44.1%
	More than 50 years	102	37.2%
Size in employees	<99	17	6.2%
	100–249	145	54%
	250–499	59	21.5%
	500–999	32	11.7%
	>1000	21	7.6%
Managerial grade of respondent	Supply chain	72	26.2%
	Procurement and logistics	65	23.7%
	Production and plant	57	20.8%
	Strategy and innovation	42	15.3%
	General management	16	58.3%
	Operations	12	4.3%
	Other	10	3.6%

As the survey was conducted in Sweden, the questionnaire was developed in English and all relevant materials were subsequently translated to Swedish for reporting purposes with the help of a bilingual supply chain and manufacturing specialist working in two different manufacturing firms. In addition, the questionnaire was pilot tested on 10 academics and business professionals, and their feedback was used to refine the questionnaire layout and wording as needed (Collis and Hussey 2013). Moreover, the nonresponse bias was assessed by segregating the sample into early (144) and late (130) respondents (Lambert and Harrington 1990). These groups were compared with respect to 12 variables, and the analyses of variance confirmed that there were no significant differences in their characteristics.

3.1. Measurement

The questionnaire items pertaining to both dependent and independent variables were adopted from extant research and participants were required to rate them on a five-point Likert-type scale, ranging from 1 (strongly disagree) to 5 (strongly agree).

3.1.1. Dependent and independent variables

Supply chain agility is a dependent variable capturing a higher-order, dynamic capability by indicating the degree to which the supply chain can respond to changes in design requirements or costs, accommodate a large number of product improvements, provide new products to market, and effectively adjust production capacity (Al-Shboul 2017). Supply chain integration is an independent variable representing a lower-order capability used to develop the higher-order dynamic capability of SCA (Schilke 2014) and measures the supply chain partners’ involvement in joint problem solving, production planning and demand forecasting, as well as the degree of information sharing with respect to changing needs and goal setting activities (Wiengarten et al. 2014). Industry 4.0 digital technologies is also an independent variable, reflecting the degree of the participating firms’ involvement in digital technologies such as cloud computing, big data analysis, Internet of Things (IoT), and additive manufacturing and robotic systems. In particular, it relates to the use of these technologies to (1) monitor, track, and control the supply chain processes remotely; (2) make autonomous supply chain decisions; (3) use information to optimise supply chain activities in real-time; (4) process large volumes of data and apply data analytics for decision making; and (5) strengthen inter- and intra-organisational information sharing within the supply chain (de Vass, Himanshu, and Shah 2018). IoT is the key enabling technology in the Industry 4.0 project that was originally conceived by the German government to make use of information and communication technologies to improve manufacturing (BWMI 2021). Not only are these two notions used largely interchangeably by managers in the manufacturing industry, also prior research has acknowledged as one of the building blocks of Industry 4.0 framework (Ben-Daya, Hassini, and Bahroun 2019). In fact, Industry 4.0 framework was developed and initiated with the idea to incorporate IoT in a way that ideally smart manufacturing factories could be realised by machine automatically operating and exchanging information (cf. Hofmann and Rüsch 2017; Ben-Daya, Hassini, and Bahroun 2019).

3.1.2. Control variables

To ensure that findings were not influenced by certain firm characteristics, firm size, sales and age were adopted as control variables, as these factors are indicative of a firm’s capacity to implement digital technologies and other supply chain practices (Gligor, Esmark, and Holcomb 2015). For the purpose of the present study, the number of employees was used as a measure of firm size, annual revenue was the sales indicator, and age corresponded to the number of years that the firm has been in operation. Given the focus on technological development in this study, technology intensity in the industry in which the firm operates was also included as a control variable. For this purpose, the participating firms were divided into four categories – low, medium-low, medium-high and high technology – based on NACE Rev. 2. These categories were used to form sets of industry dummy variables in line with the approach taken by Eurostat and Statistics Sweden (Statistics Sweden 2017). These four control variables were retrieved from the database Retriever Business, as outlined above. Finally, market turbulence was also controlled in the model, as the responsiveness of supply chains is highly dependent on market demand (Ying et al. 2016). To determine market turbulence for their particular sector, the respondents were asked to rate the degree of changes in customer preferences, customers’ desire to look for new products, and the differences between the needs of new and existing customers (Akgün and Keskin 2014).

3.2. Data screening

To ensure data quality and eliminate potential outliers, data screening was performed, whereby questionnaires submitted by individual respondents were deemed outliers if more than 10% of the questions were unanswered, or if dishonest or ‘random’ answers were detected. If only a few values were missing, questionnaires were considered acceptable and missing values were replaced by averages prior to the analysis. The sample data did not consist of extreme values as outliers since the evaluation of kurtosis shows that all items are within the threshold of (> / < ±1), thus the variance in the items of sample data is retained.

3.3. Exploratory factor analysis

Exploratory factor analysis was conducted using the maximum likelihood extraction method. To determine if the items pertaining to each construct were adequately correlated and met the validity and reliability standards, the Bartlett’s test and Kaiser-Meyer-Olkin measure were assessed and satisfactory (KMO = .874, χ² = 3633.175, df = 231). Thus, as the correlation matrix was adequate and significantly different from an identity matrix, the dataset was deemed suitable for factoring. In addition, the exploratory factor analysis considering the Varimax rotation test grouped the items into four respective constructs, as anticipated by the proposed model. To determine the optimal number of factors to retain, the Kaiser-Guttman criterion eigenvalue approach was adopted. As the cumulative variance explained (61.8%) was above the minimum threshold of 50% it was unlikely that random errors in the data would adversely affect the correlations. Moreover, the eigenvalues indicated that all four constructs in the proposed model should be retained. Table 2 shows the rotated component matrix with component loading.

Table 2. Constructs and loading factors.

Constructs and items	Standardized loading
Supply Chain Agility
Our supply chain responds quickly to our changing requirements of design	0.65
Our supply chain responds quickly to our changing requirements of costs	0.66
Our supply chain responds quickly to introduce large numbers of product improvements	0.84
Our supply chain responds quickly to offer/introduce new products to the market	0.71
Our supply chain responds effectively to adjust capacity in production	0.65
Supply Chain Integration
We share production planning and demand forecast information with key supply chain partners	0.64
We and our key supply chain partners inform each other in advance of our changing needs	0.75
We include our key supply chain partners in our planning and goal setting activities	0.68
We regularly solve problems jointly with our key supply chain partners	0.68
Industry 4.0
Industry 4.0 digital technologies provide individual item identification of products, packages, containers and pallets	0.80
Industry 4.0 digital technologies provide monitoring, tracking and controlling of supply chain entity processes remotely	0.88
Industry 4.0 digital technologies provide autonomous supply chain decisions	0.89
Industry 4.0 digital technologies provide information to optimise supply chain activities in real-time	0.90
Industry 4.0 digital technologies provide large volumes of data to apply data analytics for decision making	0.85
Financial Performance (compared to three main competitors)
… we have higher growth in profit	0.77
… we have higher growth in market share	0.63
… we have higher growth in return on investments	0.90
… we have higher growth in return on assets	0.89

3.4. Confirmatory factor analysis

AMOS 27.0 was used to conduct confirmatory factor analysis (maximum likelihood estimation) while retaining all constructs in the conceptual model. The results presented in Table 2 below indicate that good internal consistency was achieved, as the factor loadings of all items on their respective constructs predominantly exceeded 0.7 (Hair et al. 2010). Only five loadings were below 0.7, but all exceeded 0.6, thus meeting the minimum threshold of 0.5. Hence, composite reliability was achieved. Convergent validity was further examined by assessing the model fit parameters. The confirmatory factor analysis yielded the following results: χ² = 206.0; df = 128; χ²/df = 1.609; CFI = 0.971; GFI = 0.921; AGFI = 0.894; TLI = 0.966; RMSEA = 0.47, which confirmed that all fit parameters were within the acceptable range for SEM measurement model fit indices (Byrne 2013).

3.5. Validity and reliability

As evident from the convergent validity results reported in Table 3 below, the acceptable threshold of 0.7 was met for all constructs. Discriminant validity of the model was also examined by analysing the average variance extracted (AVE) and maximum shared variance (MSV) following the validity determination criteria provided by Hair et al. (2010). As the model’s AVEs for each construct were greater than the respective MSVs, the average variance extracted for each latent variable indicated independence of dimensions, thus confirming that discriminant validity has also been achieved. Next, to illustrate the reliability of the items (i.e. the consistency of the item-level errors within a single factor), the Cronbach’s alpha was assessed. As demonstrated in Table 3, all constructs had a Cronbach’s alpha value above 0.7, confirming that all items within the construct had consistent loads and were thus reliable. The descriptive statistics and correlation matrix are provided in Table 3.

Table 3. Mean value, standard deviations (SD), Composite reliability (CR), Average variance extracted (AVE), Cronbach’s (α) and correlation variables.

	Mean	SD	CR	AVE	MSV	Α	1	2	3	4
1. Industry 4.0	3.36	1.11	0.937	0.748	0.084	0.936	0.865
2. Supply chain agility	2.75	0.70	0.830	0.501	0.287	0.808	0.255***	0.704
3. Supply chain integration	3.15	0.84	0.812	0.521	0.287	0.806	0.290***	0.536***	0.722
4. Financial performance	3.07	0.69	0.879	0.650	0.114	0.871	0.157*	0.388***	0.251***	0.806

Notes: *p < 0.05; **p < 0.01; ***p < 0.001

3.6. Common method bias

We have conducted the procedural approach in our survey study to assess the possible common method bias (Podsakoff et al. 2003). To do so, we followed three steps. Within the first step, following Azadegan et al. (2020), we selected the respondents with the managerial roles including level of seniority to ensure we have the most reliable organisational information. In the second step, we reduced possible common-rater results by providing a high level of confidentiality to our respondents. Third, we decreased the possibility of sociably desired responses by clarifying that there are not any true or false answers on survey questionnaires so that they could answer the survey items with a high level of honesty. In addition, to assess the risk of the presence of common method bias, two tests were conducted. First, Harman’s single factor test was performed to understand the presence of common method bias. Following, Kassinis and Soteriou (2003), if a single factor could account for more than half of the total variance, the common method variance appeared. Based on the factor analysis of all items, four factors with eigenvalues more than 1.0 accounted for 61.36% of the total variance. The results showed that when all items were constrained to one factor, the variance was equal to 27.36%, which was under the recommended threshold of 50%. The second test was related to an unmeasured latent factor approach. The unmeasured latent factor, together with a comparison of standardised regression weights, was performed. The critical ratios of differences in regression weights were below 0.20 and z scores of all sub-groups were non-significant at p < 0.50 (cf. MacKenzie and Podsakoff 2012).

3.7. Endogenous assessment

To assess whether our analysis might be misleading due to endogeneity bias, we have used the control function method to estimate whether our results were subjected to self-bias (Ullah, Zaefarian, and Ullah 2020). In this method, the residuals for our predictor were attained by regressing one of the independent variables (i.e. SCA) on the moderating variable (i.e. Industry 4.0), the instrumental variable (i.e. market turbulence), and other control variables. In this method, normally, the instrumental variable should affect SCA, which is an endogenous variable, but should not effect the dependent variable (financial performance). Thus, we used market turbulence as an instrumental variable as it is expected that SCA will be greater where the market is volatile (Ying et al. 2016). However, we did not find any indication that market turbulence directly affects financial performance (Zhou, Mavondo, and Saunders 2019). Equation 1 explains the residual as: (1) $\text{SCA}=\alpha 0+\alpha 1\mathrm{I}4+\alpha 2\mathrm{M}\mathrm{T}+\text{Controls}+\zeta$(1) wherein SCA indicates SCA, I4 refers to Industry 4.0, and MT refers to market turbulence. In the first step, the results show that market turbulence is positively related to SCA (α₂ = 0.35, p < 0.01), suggesting the strength of the market turbulence as an instrumental variable. In the second step, we regress the dependent variable by including the residuals as a control variable to assess the self-selection bias using: (2) $\begin{array}{rl}\text{FP}& =\text{β}0+\text{β}1\mathrm{S}\mathrm{C}\mathrm{A}+\text{β}2\mathrm{I}4+\text{β}3\mathrm{S}\mathrm{C}\mathrm{A}\times \mathrm{I}4\\ & +\text{β}4\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{d}\mathrm{u}\mathrm{a}\mathrm{l}+\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}+\zeta .\end{array}$(2)

FP indicates the financial performance of the firms. To assess if our results are leading to self-selection bias, we have compared the results in equation 2 with equation 3, by which we have excluded the residuals that were attained from equation 1. (3) $\begin{array}{rl}\text{FP}& =\text{β}0+\text{β}1\mathrm{S}\mathrm{C}\mathrm{A}+\text{β}2\mathrm{I}4+\text{β}3\mathrm{S}\mathrm{C}\mathrm{A}\times \text{I}4\\ & +\text{Controls}+\zeta .\end{array}$(3)

The comparison of results between equations 2 and 3 in Table 4 below shows that our analyses are not influenced by self-selection bias, as the patterns of significance are almost the same.

Table 4. Regression test for self-selection bias.

	Model 1 (include correction term)	Model 2 (include correction term)
SCA	0.298**	0.283**
Industry 4.0	0.107	0.102
SCA×Industry 4.0	0.120*	0.122*
R2	0.16	0.15

Note: SCA: supply chain agility.

n=274, ** p < 0.01, * p < 0.05.

4. Results

4.1. Hypothesis tests

The structural model for testing the hypotheses in the proposed model was developed in AMOS 27.0. For this purpose, a bootstrap analysis with 5000 samples at a 95% confidence interval was conducted. The findings confirmed that the overall fit of the structural equation models was adequate (see Table 5 below). The hypotheses of this study were tested after controlling for firm age, size, sales, level of technological intensity, and market uncertainty.

Table 5. Mediation results.

Relationship	Indirect effect			Direct effect			Total effect
	Effect	s.e	95%CIs	Effect	s.e	95%CIs	Effect	s.e	95%CIs
SCI → SCA → FP	0.271**	0.070	[0.116,0.421]	0.130	0.094	[-.277,.127]	0.156**	0.094	[0.69,0.318]

Notes: S.E in parentheses. Bootstrapping 95% confidence interval bias corrected in square brackets (based on n = 5000 subsamples).

*p < 0.05; **p < 0.01; ***p < 0.001.

The relevant p-values and standardised path coefficients reveal that SCI enables SCA (β = 0.554; p < 0.001) and that SCA significantly enhances financial performance (β = 0.296; p < 0.01), confirming H1 and H2. In addition, mediation analysis was also conducted to determine whether SCI exerts a significant indirect effect on financial performance. We have assessed the direct, indirect and total effects in mediation analysis (cf. Hayes 2009). The results show the direct effect in the baseline model between SCI and financial performance was not significant (β = .130; p > .10), whereas the indirect effect was statistically significant (β = 0.271; p < 0.01). This indicates full mediation of SCA in the impact of SCI on financial performance. Full mediation reveals the effect of SCI on financial performance is realised through SCA and thus hypothesis 3 is supported. Table 5 shows the mediation results including total effects, indirect effect and direct effect.

Next, model 2 was developed to test the moderating effect of Industry 4.0 digital technologies on the relationship between SCI and SCA. The results reported in Table 6 below reveal that the moderating effect of Industry 4.0 digital technologies was not significant (β = .021; p > .10), thus rejecting Hypothesis 4a. In addition, model 3 in Table 6 tested the moderating effect of Industry 4.0 and SCA on financial performance. The result indicates that Industry 4.0 positively moderates the path from SCA to financial performance (β = .121; p < .05), thus supporting hypothesis 4b.

Table 6. Structural model results.

Dependent Variable → Independent Variable	Model 1	Model 2	Model 3
SCI → SCA	0.554***	0.429***	0.561***
	(0.072)	(0.070)	(0.073)
	[0.382, 0.682]	[0.222, 0.613]	[0.387, 0.686]
SCA → FP	0.296**		0.287**
	(0.094)		(0.093)
	[0.127, 0.479]		[0.111, 0.465]
SCI → FP	0.130		0.114
	(0.84)		(0.086)
	[−0.082, 0.318]		[−0.094, 0.309]
I4 → SA		0.093
		(0.035)
		[−0.022, 0.204]
I4 × SCI → SCA		0.021
		(0.035)
		[−0.022, 0.204]
I4 → FP			0.093
			(0.040)
			[−0.044, 0.218]
I4 × SCA → FP			0.121*
			(0.044)
			[−0.019, 0.288]
Size (CV) → FP	−0.171		−0.172
	(0.091)		(0.090)
	[−0.477, 0.126]		[−0.477, 0.123]
Sales (CV) → FP	0.197		0.184
	(0.074)		(0.073)
	[−0.082, 0.489]		[−0.092, 0.472]
Age (CV) → FP	−0.022		−0.034
	(0.059)		(0.059)
	[−0.137, 0.104]		[−0.152, 0.094]
Low-med tech (CV) → FP	0.101		0.109
	(0.109)		(0.108)
	[−0.048, 0.253]		[−0.042, 0.257]
Med-high tech (CV) → FP	0.056		0.072
	(0.105)		(0.105)
	[−0.081, 0.194]		[−0.069, 0.209]
High tech (CV) → FP	−0.002		−0.008
	(0.203)		(0.201)
	[−0.161, 0.137]		[−0.163, 0.137]
MT (CV) → FP	−0.091		−0.088
	(0.066)		(0.066)
	[−0.231, 0.046]		[−0.231, 0.052]
R2 (SA)	0.31	0.36	0.31
R2 (FP)	0.15	–	0.16
χ^2	277.068	211.268	306.663
Df	1.565	123	203
CFI	0.961	0.955	0.960
GFI	0.917	0.930	0.916
TLI	0.949	0.931	0.946
RMSEA	0.046	0.051	0.043

Notes: S.E in parentheses. Bootstrapping 95% confidence interval bias corrected in square brackets (based on n = 5000 subsamples).

*p < 0.05; **p < 0.01; ***p < 0.001.

4.2. Robustness test

To verify the validity of our findings, we have assessed the robustness of our results by including two measures (ROI and EBITDA) for financial performance in our study. Earnings before interest, taxes, depreciation, and amortisation (EBITDA) and return on investment (ROI) have been accepted by OSCM scholars as a reliable accounting-based financial measures to capture operational profitability which is independent of capital investments or non-operating expenses (Shi and Yu 2013; Shockley and Turner 2015). Following existing studies that investigated supply chains and financial outcomes (Vickery et al. 2010; Quintana-García, Benavides-Chicón, and Marchante-Lara 2021), we have retested the model by including ROI and EBITDA as two measures of financial performance and found consistent results. Hypotheses 1 and 2 were supported (β = 0.524; p < 0.001 and β = 0.128; p < 0.10). The mediation result was positive and significant (β = 0.08; p > 0.10 for direct effect and β = 0.119; p < 0.1 for indirect effects). Also, hypothesis 4a was rejected (β = 0.008; p > 0.10) and 4b was supported (β = 0.167; p < 0.05) respectively. The overall findings of the retested model, including new measurements for financial performance, replicated the hypothesis results and thus validate our findings.

5. Discussion

In this study, we tested the mediating role of SCA on the relationship between SCI and financial performance and found support for its full mediation effect on this relationship. Interestingly, in the presence of SCA, the direct relationship between SCI and financial performance is not statistically significant but becomes significant when SCA is assumed to exert an indirect effect. These results provide evidence for the earlier assertion that SCI is an antecedent of SCA (Braunscheidel and Suresh 2009), which we conceptualise as a higher-order dynamic capability (cf. Schilke 2014). Our findings further indicate that SCA exerts a positive impact on financial performance, which is in line with existing literature (Gligor, Esmark, and Holcomb 2015; Al-Shboul 2017). Therefore, superior performance outcomes from SCI as a lower-order capability can be realised by developing the higher-order dynamic capability of SCA, which is in line with performance effects predicted by Collis (1994).

In addition, we investigated the moderating effect of Industry 4.0 digital technologies on the relationship between SCI and SCA, as well as on the relationship between SCA and financial performance. Although we did not find support for its moderating role in the link between SCI and SCA, Industry 4.0 digital technologies appear to play a significant conditional role when it comes to how SCA impacts financial performance. As a result, companies that have implemented Industry 4.0 digital technologies and have developed the required capabilities can be expected to derive a greater positive impact from SCA in terms of their financial performance. This prompts us to conclude that the dynamic capability of leveraging Industry 4.0 digital technologies supports agile manufacturing firms in achieving superior financial performance.

5.1. Theoretical contributions

We make various contributions to theory and the existing body of literature. We advance the current theoretical knowledge of the interrelationships among supply chain dynamic capabilities (Aslam et al. 2018; Blome, Schoenherr, and Rexhausen 2013) and their role on gaining a competitive financial edge. Specifically, we address the call made by leading scholars to extend the dynamic capability perspective beyond the firm boundaries, to explain how resources and capabilities can lead to completive advantage over external market actors (Schilke, Hu, and Helfat 2018). We engage in the ongoing call for empirical evidence elucidating how low and high-order dynamic capabilities are intertwined in the hierarchy of dynamic capabilities (Schilke 2014). From this perspective, we argue that SCI as a lower-order dynamic capability can be used to develop SCA as a higher-order dynamic capability, which in turn facilitates superior financial performance (Ambrosini, Bowman, and Collier 2009; Collis 1994). Moreover, drawing on the dynamic capability perspective, we discuss that to effectively prepare for and react to the dynamics in the market, manufacturing firms need to ensure that pertinent information is transferred across the supply chain partners in a timely manner. In a frequently changing industry such as the manufacturing sector, this requires continuous and regular information exchange. In addition, supply chain partners need to be included in planning and goal-setting processes, as well as problem-solving activities because, according to Braunscheidel and Suresh (2009), external integration between suppliers and customers is the strongest predictor of SCA. Gathering and sharing information, especially related to the market, which is central to SCI, improves the capability to seize and quickly react to opportunities, which is a main attribute of SCA. Therefore, we contribute to advancing theory regarding how dynamic supply chain capabilities can be created (Aslam et al. 2018). In this realm, our study also advances the current understanding of the role of dynamic capabilities in firms’ efforts to achieve a competitive market position. Specifically, we demonstrate that developing SCA, which enables manufacturing firms to quickly respond to changes in a market environment, allows them to effectively leverage internal and external resources and capabilities to make quicker decisions and react promptly to new market conditions. Hence, consequently, such firms can be expected to financially outperform competition in the marketplace.

Importantly, our study indicates that SCI alone is insufficient to ensure financial benefits and that, instead, superior performance requires SCI as an antecedent to feed into SCA. This finding is in line with the argument put forward by several authors that a single-sourcing integrational approach could increase the risk of supply chain setbacks, pointing to potential drawbacks of focusing solely on SCI (Flynn, Huo, and Zhao 2010; Gimenez, van der Vaart, and van Donk 2012). In sum, our study suggests that manufacturing firms cannot expect direct financial returns from investing in SCI, as they also need to develop SCA. Disentangling these relationships can help to shed light on the performance implications of higher- and lower-order dynamic capabilities, as propagated by Fainshmidt et al. (2016).

Moreover, while various possible factors conditioning the effect of SCA have been identified in prior research, surprisingly Industry 4.0 digital technologies have been relatively neglected (Vanpoucke, Vereecke, and Muylle 2017). Since Industry 4.0 digital technologies facilitate real-time information provision, their effective implementation across the supply chain can improve its effectiveness and efficiency when responding to market changes. Moreover, such technologies allow automated item (e.g. products, packages, containers, and pallets) identification, which is crucial for enhancing SCA. Thus, by investigating the conditioning effect of Industry 4.0 digital technologies on the link between SCA and performance, we address this paucity of research outlined above. The rationale behind this finding is that, by being able to monitor, track, and control all activities and processes along the supply chain, Industry 4.0 digital technologies allow SCA initiatives to be leveraged more efficiently. For example, SCA can be enhanced by applying data analytics for real-time decision making. It also allows entities along the supply chain to make autonomous decisions, which expedites response to any changes while minimising risks. As a result, SCA can translate into superior financial performance if supported by Industry 4.0 digital technologies, as previously suggested by Warner and Wäger (2019). While Tarafdar and Qrunfleh (2017) investigated the conditional impact of such technologies in the broader framework of IT systems, we find support for such a role of contemporary Industry 4.0 digital technologies in the context of the manufacturing industry. From the dynamic capability perspective, our findings demonstrate that Industry 4.0 digital technologies play a supporting role in reinforcing the financial outcomes resulting from the higher-order dynamic capability of SCA (Amit and Schoemaker 1993; Teece, Pisano, and Shuen 1997; Schilke 2014). This benefit can be attained because Industry 4.0 digital technologies can help firms and their supply chain partners to quickly sense, process, and translate changes in market requirements (cf. Eisenhardt and Martin 2000). Thus, we argue that by incorporating such digital technologies, firms can attain much greater benefits than merely outperforming competition operationally, as previously suggested by Akter et al. (2016) and Wamba et al. (2020).

Contrary to our expectation, our findings did not support the moderating role of Industry 4.0 digital technologies in the relationship between SCI and SCA. This result could be attributed to challenges related to data management, data mining, data security and data errors (Lee and Lee 2015), and confirms that SCA development requires organisational processes that go beyond the use of technologies.

5.2. Managerial implications

Our findings have several managerial implications. Firstly, as the market demands regarding supply chain efficiency are higher than ever (Yao and Askin 2019), firms should focus on developing SCA to facilitate achieving their financial and operational goals. Secondly, our findings suggest that SCA is most effectively enhanced through close collaboration with key supply chain partners, including joint planning, goal setting, and problem-solving, as well as timely and transparent information sharing. However, as many manufacturing firms are still not fully aware of the many benefits of SCI, managers need to develop such lower-order capabilities feeding into developing the higher-order dynamic capability of SCA. Thirdly, according to our findings, to increase SCA, managers should also invest in Industry 4.0 digital technologies and in developing and maintaining the competences of all employees to ensure that their firms can quickly respond to any changes in market requirements. As these technologies facilitate the tracking and monitoring of supply chain processes, they enable real-time information sharing and provide larger volumes of data for analysis and decision-making. This would in turn increase the financial performance and the degree of the competitiveness of manufacturing firms.

Finally, our findings indicate that, when Industry 4.0 digital technologies are implemented to a significant degree, integration among supply chain partners does not necessarily improve SCA. This result is attributed to the challenges related to data management, data mining, and data security related to Industry 4.0, which might reduce the supply chain partners’ willingness to work closely with each other.

6. Limitations and further research

Our study has some limitations which provide possibilities for further research. Firstly, our study focused mainly on the responsiveness aspect of agility, while overlooking alertness (Gligor and Holcomb 2014; Sangari and Razmi 2015). This limitation could be addressed by investigating how Industry 4.0 affects the supply chain partners’ alertness to changes in the supply chain itself and within the external environment. In this context, the role of Industry 4.0 digital technologies can also be investigated, as they can help manufacturing firms monitor, track, and make autonomous decisions. Secondly, since supply chains are highly influenced by disruption risks (Ivanov, Dolgui, and Sokolov 2019), digital technologies are likely to mitigate such risks; as, by increasing SCA, firms are better prepared to respond and adapt to changes in turbulent markets. Even though market turbulence was not the focus of the current investigation, disruption risk is likely to influence the strategies manufacturing firms adopt in responding to market changes and should be examined further. A further limitation stems from focusing on Swedish medium-sized and large manufacturing firms. While Sweden is a frontrunner in digitally-enabled manufacturing, and because medium-sized and large firms often have enough capital to invest in Industry 4.0 digital technologies, the results reported here might not apply to smaller manufacturing firms or other sectors and countries. Thus, it would be beneficial to conduct a similar study focusing on smaller firms, or those operating in different countries or industries, as this would help determine if SCA and digital technologies play the same roles as those identified here. Given that small and medium-sized firms account for around 90% of all businesses worldwide, such investigations would be of importance. It should also be mentioned that while we have taken measures to minimise the negative role of endogeneity and common method bias, along the lead of other scholars (Wamba et al. 2020), we call for having multiple informants from each firm or carrying out longitudinal studies against these backdrops. Finally, we have focused on supply chain wide capabilities in our conceptualisation. Specifically, although we acknowledge that integrative activities at the firm level are crucial, in operationalising SCI, we have primarily focused on integration with external supply chain actors. Future studies could incorporate more comprehensive operationalizations to include inter- as well as intra-organisational measures.

Acknowledgement

The authors acknowledge the support for this research from Jan Wallanders och Tom Hedelius Stiftelse samt Tore Browaldhs Stiftelse foundations, as well as the e-merge project financed by the European Regional Development Fund (via the Swedish Agency for Economic and Regional Growth - Tillväxtverket). The authors thank the participating firms and anonymous reviewers for valuable feedback and comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Notes on contributors

Mohammad H. Eslami

Mohammad H. Eslami is Assistant Professor at Jönköping International Business School (JIBS), Sweden. His research interests are in the field of Innovation Management and supply chain management. He is specially interested in supply chain collaboration with focus on the role of digital technologies. His research has been published in Journal of Business research, Industrial Marketing Management, Journal of Engineering and Technology Management and Journal of Business & Industrial Marketing, among others.

Hamid Jafari

Hamid Jafari is an Assistant Professor at Jönköping University, Sweden. His research primarily revolves around dynamic capabilities and digitalisation in supply chains. He has participated in various research projects focusing on omni-channel retailing and e-commerce, and AI. His research has appeared in several peer-reviewed academic journals including Journal of Retailing and Consumer Services, Industrial Management & Data Systems and Computers & Industrial Engineering.

Leona Achtenhagen

Leona Achtenhagen is a Professor of Entrepreneurship and Business Development at Jönköping International Business School (JIBS) in Sweden and Visiting Professor at LUT University's School of Business and Management in Finland. She is the Director of JIBS’ Media, Management and Transformation Centre (MMTC). Her research interests focus around the impact of digitalization, globalization and the SDGs on entrepreneurship and business development in different contexts.

John Carlbäck

John Carlbäck holds a Master Degree in International Logistics and Supply Chain Management from Jönköping International Business School (JIBS) in Sweden. He works with strategic procurement and supply chain management in the healthcare industry. His research interest lies in the fields of Supply Chain Management, Operations Management, and Digital technologies. Primarily, John is interested in the modern supply chain; lowering costs and grow opportunities through digital technologies.

Alex Wong

Alex Wong received his Master of Science Degree in International Logistics and Supply Chain Management from Jönköping International Business School (JIBS) in Sweden. He is an Operational Purchaser and Production planner at Swedish manufacturing industry. His research interest is in the field of Operations Management and Data Analytics. He is particularly interested in the focus of data-driven decisions in the production industry.