Beyond efficiency: the role of lean practices and cultures in developing dynamic capabilities microfoundations

Purpose – The purpose of this study is to theorize and test the relationships among lean operations and lean supply chain practices, learning-and innovation-oriented lean cultures and dynamic capabilities (DCs) microfoundations. Further, this study aims to assess the association of DCs microfoundations with process innovation. Design/methodology/approach – The researchers combine primary data collected from 153 manufacturing firms located in five continents using a survey designed for the purpose of this study with archival data downloaded from the Bureau Van Dijk Orbis database and test the hypothesized relationships using structural equation modelling. Findings – Results support the contribution of lean operations and lean supply chain practices to the development of DCs microfoundations, which further lead to greater process innovation. Additionally, while a learning-oriented lean culture positively moderates the relationships between both lean operations and lean supply chain practices and DCs microfoundations, an innovation-oriented lean culture only moderates the relationship between lean operations practices and DCs microfoundations. Practical implications – This study identifies DCs microfoundations as the key mechanisms for firms implementing lean practicestoachieve greater levels ofprocess innovationand the important role playedby lean cultures. This study provides direction for managers to put in place DCs through lean implementations, enabling their firms to be ready to respond to challenges and opportunities generated by environmental changes. Originality/value – Whilepreviousresearchhasconfirmedthepositiveeffectsofleanpracticesonefficiency, the role of lean practices and cultures in developing capabilities for reacting to environmental dynamism has


Introduction
Lean has traditionally been conceptualized as a continuous improvement methodology to enhance organizational efficiency (Hopp and Spearman, 2021).More contemporary research on the topic has suggested that efficiency is not necessarily the only outcome and that implementing lean also provides firms with greater adaptability to respond to changes in their environments (Cusumano et al., 2021;Netland and Powell, 2016).The Lean Enterprise Institute asserts that lean organizations are more adaptive to changes in their environments than their non-lean peers (LEI, 2022).However, existing empirical studies have not tackled the questions of whether and how lean facilitates adaptability.
The adaptability of firms can be characterized as dynamic capabilities (DCs), which are defined as abilities "to integrate, build, and reconfigure internal and external competencies to address rapidly changing environments" (Teece et al., 1997, p. 516).For example, it is through developing its DCs that Samsung overtook Apple as the leading smartphone maker in the world (Park, 2022;Song et al., 2016).Previous researchers have analysed the relationships of continuous improvement initiatives such as lean with DCs (Dobrzykowski et al., 2016;Gutierrez and Antony, 2020) and prescribed elements of continuous improvement organizational infrastructure for DCs (Anand et al., 2009).Mohaghegh et al. (2021) found lean practices to be positively associated with systematic problem-solving, agile manufacturing and continuous improvement.What remains unaddressed, however, is whether and how implementing lean can help firms to continuously adapt their processes to environmental changes, such as varying customer requirements (e.g.improved response time and customization), harnessing emerging technologies (e.g.Industry 4.0), meeting sustainability objectives (Tortorella et al., 2019;Yu et al., 2020) and reacting to high impact and low-probability events such as pandemics, climate disasters and political conflicts (Alexander et al., 2022).This research aims to understand the relationship between lean implementation and DCs development, manifested by the adaptation of organizational processes in response to changing external conditions.We hypothesize that lean enables the development of DCs microfoundations, understood as the capacities to (1) sense opportunities and threats, (2) seize opportunities and ways to combat threats and (3) transform intangible and tangible assets to maintain competitiveness (Teece, 2007), leading to higher levels of process innovation.
Process innovation indicates the ability to adapt organizational processes to different contexts, which can be especially valuable in environments characterized by changing technologies, varying customer needs and market uncertainties (Piening and Salge, 2015;Teece et al., 1997).Extant literature has highlighted the importance of capacities related to DCs microfoundations, such as sensing, learning, R&D and training for generating process innovations (Pavlou and El Sawy, 2011;Piening and Salge, 2015).
To analyse the relationships between lean, DCs microfoundations and process innovation, we differentiate between lean operations and lean supply chain practices (Azadegan et al., 2013;Furlan et al., 2011;Hofer et al., 2012;Shah and Ward, 2007).In addition, we include the role of learning-and innovation-oriented lean cultures following previous research (Akmal et al., 2022;Bortolotti et al., 2015;Hardcopf et al., 2021;Onofrei et al., 2019;Yu et al., 2020).We hypothesize that process innovation is supported by lean practices implementation through DCs microfoundationssensing, seizing and transformingand strengthened by the presence of organizational learning-and innovation-oriented lean cultures.
Our analysis is based on survey and archival data from 153 lean manufacturing firms located in 34 countries.We address three research questions: (1) how are lean operations and lean supply chain practices associated with DCs microfoundations development?(2) How do learning-and innovation-oriented lean cultures impact the effect of lean operations and lean supply chain practices on DCs microfoundations development?(3) How do DCs microfoundations impact process innovation?Our research contributes to the literature streams on lean and firm environmental adaptation by demonstrating how lean practices and cultural orientations affect DCs microfoundations and process innovation.Our findings reveal how lean firms are more adaptable to their environment, through their enhanced sensing, seizing and transforming capacities.Thus, we provide an extended understanding of the benefits of lean implementation beyond mere operational efficiency.
2. Literature review 2.1 Lean practices Womack et al. (1991) popularized the term "lean" as a management concept (Holweg, 2007) after Krafcik first used the term to describe the Toyota Production System (Krafcik, 1988).Lean thinking and lean practices focus on creating customer value by reducing waste and enhancing product/service features without additional cost.Subsequently, Shah and Ward (2007) categorized lean manufacturing practices into 10 subcategories of practices, which they grouped into 3 constructs (supplier-, customer-and internal practices).

Dynamic capabilities
The DCs view extends the resource-based view (Rumelt, 1984;Wernerfelt, 1984) of sustainable competitive advantage by including an organization's ability to sense and seize new opportunities and reallocate or reconfigure resources to adapt to the environmental changes toward establishing a competitive advantage (Teece, 2007).DCs are positioned as higher-order capabilities for modifying operating routines (Zollo and Winter, 2002) and sustaining competitive advantage across changing environments (Eisenhardt and Martin, 2000;Schilke, 2014;Teece et al., 1997).Extant research has proposed a set of three organizational processes that comprise DCs microfoundations (Teece, 2007).First, sensing refers to the capability to search, explore, identify and shape new or emerging opportunities and threats from the internal and external environments.Second, seizing refers to the ability of a firm to capture the sensed opportunities by means of an infrastructure to exploit them.Third, transforming refers to the capability to reconfigure, change and modify existing resources to succeed with the seized opportunity.
Researchers have described different organizational capabilities as DCs.For example, continuous improvement is a DC that includes developing a vision for managing and improving operations, empowering employees, tapping into tacit knowledge and building a customer-oriented culture (Anand et al., 2009).Similarly, alliance management and new product development are recognized as DCs (Gutierrez-Gutierrez et al., 2018;Schilke, 2014).On the topic of continuous improvement, scholars have recognized the potential of DCs for explaining and understanding how firms achieve and sustain competitive advantages (Anand et al., 2009;Su et al., 2014).In this research, we seek to further explain the role of DCs microfoundations in the context of lean implementations.

Lean practices and DCs microfoundations
For lean practices to result in sustained competitiveness, firms need to develop the capability to select appropriate lean practices and adjust them based on varying needs dictated by firmenvironment interactions (Peng et al., 2011).Past research on relationships between lean/ continuous improvement and DCs can generally be defined by two perspectives (Gutierrez and Antony, 2020) -(1) continuous improvement as an enabler of DCs (e.g.Dobrzykowski et al., 2016;Mohaghegh et al., 2021) and (2) continuous improvement as a DC in itself.The first perspective proposes that the search for process improvements includes the need for alignment with opportunities in the environment, as well as the development of capabilities to adapt resources and exploit these opportunities.Hence, Camis on and Puig-Denia (2016) clearly stipulate that continuous improvement practices can function as vehicles for the development of organizational capabilities, such as organizational learning, though it cannot be considered as a bundle of DCs by itself due to its functional orientation.Secondly, research that does propose continuous improvement to be a DC (Anand et al., 2009) takes the view that emergent organizational processes in continuous improvement initiatives can show high similarity with DCs characteristics when these comprise comprehensive organizational contexts and sustainable organizational learning efforts.These associations between continuous improvement capabilities and DCs, however, have not been studied empirically beyond case studies (Gutierrez and Antony, 2020).
3.1.1Lean operations practices.Lean operations practices facilitate pull production (i.e.demand-triggered delivery), flow (i.e.demand-dictated organization), setup reduction (i.e.demand-based flexibility) and employee involvement (i.e.employee-led improvement) (Shah and Ward, 2007).These practices have in common, the elements of learning from customer interaction, a natural tendency to organize and if needed reorganize operations and decentralized employee-led improvement initiatives.We posit that these lean operations Lean and dynamic capabilities practices positively affect the organizational processes that are identified as DC microfoundations.
The DCs microfoundations of sensing, seizing and transforming are known to be facilitated by specific organizational processes (Teece, 2007).First, a firm's natural openness to new ideas and change, facilitate experimentation and reduce the reliance on pathdependent routines are geared toward sensing and seizing (Teece, 2007;Tushman and Anderson, 1986).Second, a firm's willingness to depart from existing routines when confronted with new business opportunities facilitates transforming.Firms with a tradition of change and improvement are argued to be less susceptible to a "heightened state of anxiety" due to letting go of past routines and more easily engaging in improvement and innovation (Teece, 2007).
Overall, successful implementation of lean operations practices such as pull-based operations that facilitate the production of units needed, at the time needed, and in the quantities needed; improved flow that enhances production movement without frequent stop-and-go operations; and setup reduction that allows firms to predict process output more exactly (Shah and Ward, 2007), warrant ongoing detection and responding to the needs of the firms' environment.Also, implementing pull, flow and setup practices is conditional on employee involvement in improvement efforts (Shah and Ward, 2007) thereby harnessing employee knowledge and experience to materialize on improvement opportunities in reaction to externally driven changes (Teece, 2007).As a result, the implementation of these practices creates an ongoing change orientation (Anand et al., 2009), thereby facilitating the development of the three DCs microfoundations (Teece, 2007;Tushman and Anderson, 1986).
An illustrative example is Volvo Motor Company, which implemented flow and pull practices as part of their lean production system to enable an adequate response to changed product quality and security expectations.As a result, their operations became more flexible and resilient, showcasing the potential of lean practice implementation for DCs microfoundations development (Netland and Aspelund, 2013).In addition, Distelhorst et al. (2017) showed how lean mechanisms increased employee involvement in Nike's factories, leading to improved labour relations, thereby positively responding to current external social demands (e.g.inspection, communication, workplace arrangements, industrial hygiene, etc.).Based on our argumentation above, we propose (Figure 1): 3.1.2Lean supply chain practices.Lean supply chain practices comprise supplier feedback (i.e.good relations with frequent feedback cycles), supplier just-in-time delivery (i.e.supplierinvolved process design), customer involvement (i.e.frequent interaction about needs and expectations) and supplier development (i.e.supplier-involved process improvement) (Shah and Ward, 2007).The common features of these practices are intensive collaborations with suppliers and engagement with customers, and we posit that these lean supply chain practices positively affect the development of DC microfoundations.
Sensing processes depend on a firm's ability to detect and identify changes in customer needs (Nonaka and Toyama, 2007).Engagements with customers and the learning that takes place through interaction processes must then be synthesized by middle management, so that transformative action can be taken in response.In such processes, it is of utmost importance that "local search" is supplemented by searching for information about what is happening in the business ecosystem.In dynamic environments, where a great deal of innovation stems from sources outside of the firm, firms must be able to monitor not only the core of their business ecosystem but also the peripheries.The example of how innovations in microprocessors created novel possibilities for downstream products such as personal computers and other electronics is a manifestation of why it is important to maintain close ties with suppliers for effective sensing capabilities (Chesbrough, 2003).
For seizing and transforming capacities, a firm must be open to new ideas and change, which is not limited to the boundaries of the firm.Path dependencies and reliance on existing routines and assets beyond the boundaries of the firm inhibit seizing processes (Teece, 2007).Vice versa, when new opportunities emerge and investments are made in product or serviceoffering developments that stretch beyond the boundaries of the firm, close supplier collaboration enhances the chances for success (Teece, 2007).Joint innovation processes comprise interfirm interpersonal relationships (Tortorella et al., 2017) that reportedly lead to higher degrees of supply chain flexibility and responsiveness (Teece, 2007).Moreover, interpersonal relations within supply chains enable the exchange of tacit knowledge, thereby enhancing sensing, seizing and transformational capabilities (Eisenhardt and Martin, 2000;Teece, 2007).Additionally, adequately seizing and transforming opportunities may require updating and learning new skills which can be facilitated by alliance arrangements with suppliers or even customers (Branzei and Vertinsky, 2006).
We posit that lean supply chain practices instil intensive collaboration with customers and suppliers strengthening joint capability development, which improves sensing, seizing and transforming capabilities.In support of our argument, earlier research has reported on how lean supply chain practices such as supplier feedback and customer involvement strengthened coordination, integration of information and knowledge transfer within supply chains (Cagliano et al., 2006;Powell and Coughlan, 2020;Tortorella et al., 2017).
An exemplary manifestation of the benefit of close supplier relationships is the case of Philips electronics (currently known as NXP).In March 2000, a fire in a New Mexico-based production site caused severe disruption in the integrated circuit supply chain of several cellphone companies, including Nokia and Ericsson (Sheffi and Rice, 2005).Nokia responded immediately and sent 30 employees to work with NXP to restore operations while solid relationships with other suppliers allowed the negotiation of contracts to ensure that its worldwide production capacity would remain intact (i.e. by means of Nokia's "supplierfeedback and development" lean supply chain practices).Ericsson, on the other hand, was slower in sensing the severity of the problem.By the time it searched for alternative suppliers, their capacities were already committed to Nokia (Sheffi and Rice, 2005).Based upon the reasoning, above we hypothesize: Lean and dynamic capabilities H2.Implementation of lean supply chain practices is positively related to DCs microfoundations development.

Lean organizational culture and DCs microfoundations
The contextual effects of cultural traits on the adoption of lean practices in firms have been researched at multiple levels of analysis.At the higher level of societal norms that influence culture, aggregated traits such as higher degrees of individualism, uncertainty avoidance and power distances are found to positively affect the adoption of lean practices adoption (Pakdil and Leonard, 2017).While the impacts of the broadly categorized traits of national culture do matter, our focus, in the present research, is on cultural traits at the organizational level that serve as the context for the implementation of lean practices (Akmal et al., 2022).As such, we emphasize cultural traits that can be moulded by organizations instead of those that are considered relatively fixed based on national origins.We adopt the definition of organizational culture as "a combination of artifacts (also called practices or forms), values and beliefs, and underlying assumptions that organizational members share about appropriate behaviour" (Detert et al., 2000, p. 851).Existing literature has supported the idea that learning-and innovation-oriented organizational cultures are key ingredients for DCs development, as these orientations support the generation of knowledge that allows a firm to adapt to its competitive environment (Teece et al., 1997;Teece, 2007;Zollo and Winter, 2002).Likewise, existing research has revealed the importance of changeenabling organizational cultural traits, without which firms cannot realize the benefits of lean implementations (Hardcopf et al., 2021;Hines et al., 2004;Onofrei et al., 2019;Yu et al., 2020).
3.2.1 Learning-oriented lean culture.Lean operations and lean supply chain practices are the observable artefacts of lean implementations (Galeazzo et al., 2021;Powell and Coughlan, 2020).The success of these lean practices (and arguably the strength of their effects on DCs microfoundations development), depends on the involvement and commitment of employees and managers in implementing these lean practices; selecting the appropriate practices based on firm-environment dynamics and making ongoing changes to practices so that these fit in the firms' contexts (Anand et al., 2009;Spear and Bowen, 1999).
Lean practices such as pull, setup or suppliers' development are not off-the-shelf solutions.Rather, they are complex to implement and require knowledgeable employees, making learning an imperative to augment lean practice efficacy (De Mast et al., 2021;Onofrei et al., 2019;Powell and Coughlan, 2020;Tortorella et al., 2019).In successful lean implementations, employees improve organizational processes and systems through collaboration, consensus building, suggestion systems and group decision-making, which together constitute a supporting infrastructure for individual and group learning (Spear and Bowen, 1999).Vice versa, an organizational culture characterized by resistance to learning hinders the realization of benefits from implementing lean practices (Wiengarten et al., 2015), both within and outside firms' boundaries (Hines et al., 2004).Examples that substantiate the importance of learning in successful lean practices adoption include Toyota, where learning is a basic part of their day-to-day operations (Spear and Bowen, 1999).Additionally, a learning-oriented culture at software company Wipro is exemplified by the creation of a productivity office based on disciplined learning that increases the effectiveness of lean practices to respond to environmental changes through the identification of best practices and the investigation of new ideas (Staats and Upton, 2011).Hence, we hypothesize that a learning culture enhances the effects of lean operations and lean supply chain practices on DCs microfoundations development: H3a.A learning-oriented lean culture positively moderates the relationship between lean operations practices implementation and DCs microfoundations development.

IJOPM 42,13
H3b.A learning-oriented lean culture positively moderates the relationship between lean supply chain practices implementation and DCs microfoundations development.
3.2.2Innovation-oriented lean culture.Firms' innovation orientations are known to enhance the adoption of lean operational practices by means of an empowering climate for employees to propose and implement improvements (Anand et al., 2009).This is typically because innovation-oriented firm cultures facilitate the search for solutions to (external) changes and this orientation is an important enabler for employees to share ideas and cooperate in the creation of knowledge (Onofrei et al., 2019).Similarly, and resembling the objectives of lean supply chain practices, innovation-oriented firms typically invest more time in supply chain partners to help them achieve organizational capabilities, through activities such as educational programmes for supply chain partners, inter-firm teams, consulting and problem-solving for-and with suppliers and collaborative R&D activities (Solaimani et al., 2019).
Lean implementations are also innovation-enhancing, manifested by continual (process) innovations (Netland et al., 2015).For example, lean practices like setup, flow and pull, provide progressive insight into required product specifications and technological competencies needed, and reportedly improve time-to-market capabilities, thereby positively affecting a firm's innovative capability (Schniederjans, 2018).Moreover, successful lean practice implementations are known to be based on the creation of an infrastructure for generating and harnessing ideas and a culture of ongoing change and innovation (Anand et al., 2009;Rathore et al., 2020).
An example that reveals the role of innovation orientations in successful lean adoption is Volvo Motor Company, which reportedly has enhanced the success of its lean initiative by means of the company's human-centred production, predominantly based on an innovationoriented culture that is attuned to responding to (internal and external) changes.Their innovation orientation enabled the replacement of a regular assembly line with a dock assembly setup to increase teams' responsibilities and joint decision powers (thereby achieving setup reduction and increased flow) (Netland and Aspelund, 2013).Hence, we hypothesize the presence of an innovation-oriented lean culture as accentuating the effects of lean operations and lean supply chain practices on DCs microfoundations development: H4a.An innovation-oriented lean culture positively moderates the relationship between lean operations practices implementation and DCs microfoundations development.
H4b.An innovation-oriented lean culture positively moderates the relationship between lean supply chain practices implementation and DCs microfoundations development.

Dynamic capability microfoundations and process innovation
Process innovations represent changes in the way sequences of activities (processes) are executed in firms.DCs microfoundations enable activities directed at generating, acquiring, integrating and disseminating knowledge for reconfiguring such organizational processes, enabling sustained competitive advantage (Pavlou and El Sawy, 2011), typically by means of enhanced technologies or systems (Teece, 2007).For example, firms conducting activities for sensing, seizing or transforming such as (1) market research (i.e.sensing customer needs, market developments), (2) scanning of patent databases (i.e.technological opportunities), (3) conducting in-house R&D and prototyping, and (4) R&D outsourcing, typically demonstrate increased process innovation activity (Teece, 2007).Developing DCs microfoundations enhances the ability of firms to recognize the value of new information for knowledge creation (Chatterjee et al., 2022).This knowledge makes the firms aware of potentially useful process innovations (Piening and Salge, 2015).Consequently, we expect firms with strong DCs microfoundations to be more flexible, better able to detect and understand the implications of opportunities and threats and better able to make innovation decisions in response to changed competitive conditions (Chatterjee et al., 2022;Gutierrez-Gutierrez et al., 2018).Therefore, we propose the following hypothesis: H5. DCs microfoundations development is positively related to process innovation.

Research methodology 4.1 Sample data
We collected survey data from mid-and senior level lean-managers and practitioners in manufacturing firms between June 2019 and June 2020 by sending a digital questionnaire to the targeted respondents.As our objective is to make generalizations about theory rather than populations, we applied non-probability heterogeneous purposive sampling to select subjects for the study (Saunders et al., 2009).That is, we aimed for substantial variation in the respondents' organizational contexts (globally scoped, differing-firm sizes, functional domains and occupational titles) to enable us to compare and generalize findings.Table 2 lists the descriptive information for our sample.
The information in Table 2 hints at differences between sample and population distributions (i.e.all lean implementing manufacturing firms): a bias towards western countries, large enterprise overrepresentation (average of 82% observed against 35.6% in the European Union alone (Eurostat, 2019), the largest cluster of respondents) and biases in the functions of employment.This suggests that the representativeness of our sample and hence generalizability of our findings are somewhat limited.However, this limitation must be seen in the light of the richness of the data that we achieve through the use of a survey conducted for the express objective of this research study based on carefully constructed scales for pointed measurements of constructs as well as the inclusion of archival data for measuring control variables and the dependent variable.
To enhance research validity and following previous research in operations management (Kortmann et al., 2014), we applied a "key informant approach" (i.e.preference for betterinformed respondents having specialized knowledge about the phenomena under research over more but less knowledgeable respondents) (Kumar et al., 1993).Respondents were selected based on two criteria.First, being educated in lean methodology, defined as at least six days of lean training (medium-level competency that is enabling autonomous lean project leadership), and second, a dedication to lean implementation, defined as at least 25% of daily working time spent on lean implementation (more than one full-time equivalent working day per week).We targeted respondents via co-authors' networks, and each respondent was asked for relevant referrals, which were vetted for function, tenure and experience in lean working environments before being invited to participate.This purposive sampling strategy minimized the chance for selection error and assured experienced and knowledgeable respondents (Eisenhardt, 1989).

Questionnaire design
We used a two-stage scale development procedure following prescriptions by Hensley (1999).First, we put together the questionnaire based on a review of the existing literature and consultations with lean experts for adopting existing scales.Second, we tested the initial questionnaire for clarity and user-friendliness by working with eight lean practitioners based on the principle of saturation (Saunders et al., 2009).This helped in reducing the chances of errors and biases from the perspectives of participants and observers.The research team amended and clarified questions based on these discussions.

IJOPM 42,13
The resulting questionnaire comprised two parts.The first section was designed to collect the demographics of the respondents such as country, position, lean experience, economic sector of the firm and number of employees.The second section sought data on constructs of interest based on existing validated 7-point Likert scales (Section 4.3) using reflective items.On one hand, this design assured reliable respondent assessments (Finstad, 2010), and on the other hand, it generated continuous and likely normally behaving response data as a prerequisite for the chosen information-dense methodological technique (Kline, 2015).

Lean and dynamic capabilities
We prepared the questionnaire in English and Spanish versions using the standard backward translation method (Brislin, 1970).With this technique, a questionnaire is translated and blindly translated back to the original language by another translator to detect anomalies.This procedure is repeated several times until no more anomalies exist.Finally, we paid special attention to writing concise and clearly understandable items (Noar, 2003).
We took four steps in the design and execution of our survey aimed at increasing the response rate and reducing the concern for non-response bias (Goyder, 2019).First, in the early stages of the development of the survey questionnaire, we conducted pre-tests and collected feedback on its clarity, interpretability and multi-medium deployment.Second, we persisted with the data collection for an extended period of one year during which we sent reminders to the targeted respondents.Third, we assured the respondents that the identities of the firms and respondents would remain confidential.Fourth, as an incentive for participation, we offered to share the results of the research with those who responded.We assessed non-response bias by testing for differences between (1) early and late and (2) complete and incomplete responses.Both sets of twosample t-tests on the construct mean scores revealed no significant differences, indicating that the risk of non-response bias was small (Armstrong and Overton, 1977;Whitehead et al., 1993).
We sent the finalized questionnaire to 510 targeted respondents out of which 195 respondents showed interest by starting to respond.In the process of data preparation, we deleted incomplete responses and imputed missing values for less than 5% of the remaining sample by means of single regression imputation (Kline, 2015).A total sample of 153 useable cases for data analysis remained.

Measures
The unit of analysis is lean implementation in the respondents' firms.The constructs of "lean operations practices" and "lean supply chain practices" are operationalized using scales based on Shah and Ward (2007).For measuring "learning-oriented lean culture" and "innovation-oriented lean culture", we adapted scales from Hult et al. (2007).Our measures for the three microfoundations of DCssensing, seizing and transformingare adapted from Jantunen et al. (2018).Finally, we measured the construct for our dependent variable "process innovation" using scale items from Wang and Ahmed (2004).The constructs and items adopted are presented in Appendix 1.We included firm size (Table 1), lean experience (in years) of the firm, industry subtype within the manufacturing sector, and geographical location as control variables in our analysis (Sousa and Voss, 2008).
To alleviate common method bias (CMB) and as a robustness check for our results, we repeated our analyses using an alternative measure of firm performance based on secondary financial company data as our dependent variable.To do this, we merged our survey data which contained company identifying information, with data from the Bureau Van Dijk Orbis database.The Orbis database is a large Moody's Analytics owned global database that includes data on listed and unlisted companies.The data compiled in Orbis is sourced from over 170 different company information providers and standardized to enable comparisons (Bureau van Dijk, 2022).This database provided us with financial information for 100 of the companies represented in our sample.Specifically, we obtained the Return on Capital Employed (ROCE) indicator, which represents how well a firm is generating profit from its capital.ROCE has been used in past studies to measure operations performance (Hernandez-Vivanco et al., 2019;Martins and Lucato, 2018).In addition, we checked the robustness of our results by using scales adopted from Schilke (2014) that were included in our survey questionnaire, measuring strategic and financial performance, instead of process innovation.IJOPM 42,13

Data analysis method
We estimated partial least squares (PLS) based structural equation models (SEM) using the Smart PLS 3 software (Ringle et al., 2015).PLS carries out non-parametric structural equation modelling with interdependent ordinary least squares (OLS), which minimizes residual variances (Henseler et al., 2016).PLS is deemed a suitable method for this research because (1) our research questions call for a predictive study of endogenous variables, (2) the nonparametric character of PLS allows us to analyse variables whose distribution is not normal and (3) PLS allows for the simultaneous estimation of multiple independent equations (Peng and Lai, 2012).Additionally, PLS is particularly recommended over other SEM covariancebased methods for models that use both formative (lean practices and DCs microfoundations) and reflective (cultural orientations and process innovation) constructs (Peng and Lai, 2012).PLS allows for combining the measures for reflective constructs, with their theoretical concepts (i.e.cultural orientations) to serve a certain goal such as revealing its relationship with formative constructs (i.e.lean practices).Due to the constructivist nature of this approach, recent literature suggests operationalizing these constructs by a composite model focusing on explained variance, as employed by PLS (Henseler et al., 2016;Benitez et al., 2020).The use of PLS is increasingly accepted in academic research, demonstrated by numerous recent PLS-based operations management studies (e.g.Blome et al., 2013;Hadid et al., 2016).Table 3 shows the descriptive statistics for the construct variables included in the study.

Model development
The measurement model was subjected to a validation process consisting of multiple steps.First, we ensured the content validity of our scales by using scales that had been validated and used in previous empirical studies.In addition, we pretested our scales with experienced researchers and firm managers to confirm their face validity and reconfirm content validity.Second, we analysed the reliability and internal validity of the scales using confirmatory factor analysis, for which the first-and second-order details are presented in Appendix 1. Internal consistency was assessed using Cronbach's alpha (Cronbach, 1951) and composite reliability analyses (Netemeyer et al., 2003).All our scales have Cronbach's alpha and composite reliability scores exceeding the minimum recommended threshold of 0.7, except Setup (Cronbach's alpha 5 0.671).Third, having defined our measurement model in totality, we conducted convergent validity analysis to assess item-factor loadings (Peng and Lai, 2012).All our factor loadings exceed the minimum suggested threshold of 0.7 except two (0.673 for Flow5 and 0.648 for SupplierDevelopment4).With the aim of preserving content validity, we decided not to exclude the Setup scale and the scale items Flow5 and SupplierDevelopment4 because the factor loadings were only just below the respective thresholds and the other scale reliability and validation requirements were fulfilled.We further evaluated the convergent validity of the scales, checking for values of their average variance extracted (AVE) to be above 0.5 (Hair et al., 2016) (Appendix 1).Finally, the discriminant validity of the constructs was confirmed through compliance with the Fornell and Larcker criterion (Fornell and Larcker, 1981), requiring the value of the square root of the AVE for each construct to be always greater than the corresponding correlations between paired constructs.In addition, the Heterotrait-Monotrait Ratio (HTMT) test showed values lower than the recommended upper limit of 0.9, indicating that the different scale items are not measuring the same construct (Benitez et al., 2020).
For the second-order formative constructs "lean operations practices", "lean supply chain practices" and "DCs microfoundations", we ensured adequate content validity via a rigorous qualitative approach (i.e.literature review) and evaluations of the validity of the constructs by our expert panel (Hair et al., 2016).We checked for multi-collinearity of the first-order constructs by assessing the variance inflation factors (VIFs) in our structural models.The largest VIF value is 2.23, which is lower than the commonly referenced upper threshold of 3 (Diamantopoulos and Siguaw, 2006).In addition, to test the absolute contribution of each first-order construct to the related second-order formative construct, we verified that their loadings are significant (p-values <0.10) (Cenfetelli and Bassellier, 2009).
Finally, we evaluated CMB, following tests involving both procedural and statistical methods.First, a survey pre-test was performed to avoid ambiguity (Podsakoff et al., 2003).Second, as a robustness test, we used data from primary and secondary sources (Spector, 2006).Third, we evaluated the possibility of bias using Harman's single-factor test.The results of the principal component analysis, in which all items were loaded on one construct, resulted in a low explained variance of 31.92%,reducing the concern of CMB in our study (Podsakoff et al., 2003).Fourth, we tested for CMB using the correlation marker technique (Lindell and Whitney, 2001).Our questionnaire included the variables of interest for the study and a variable theoretically unrelated to them (marker variable) but equally susceptible to measurement biases.The method assumes that, when the variance of the common method is present, the marker variable will be just as affected as the variables of interest by the effects of the method.The correlations between the latent variables of interest in the study (i.e.lean practices or DCs microfoundations) were greater than the correlation between the marker variable and the variables of interest.The results of these multiple checks alleviate concerns about the potential effects of CMB on our analyses (Podsakoff et al., 2003;Spector, 2006).

Structural model evaluation and hypotheses testing
To estimate the structural model, we employed Smart PLS-SEM 3.0 which uses the PLS algorithm to generate the standardized path coefficients.In addition, we used a bootstrapping re-sampling procedure to estimate the significance of the path coefficients.To perform the moderation effects analyses, we used interaction terms created as the products of the maineffects and moderating variables.Figure 2 shows the output of this estimation.
Results show positive and significant relationships between lean operations practices and DCs microfoundations (λ 5 0.185, p < 0.01) and between lean supply chain practices and DCs microfoundations (λ 5 0.202, p < 0.01), supporting H1 and H2.In addition, learning-oriented lean culture positively moderates both relationships between lean operations practices and
Relationships between the control variables and process innovation were not significant.We conducted additional analyses of inter-group differences.Feasible group comparisons based on (1) alike contextual economic conditions (European Union membership) combined with (2) the degree of economic, social and political dimensions of globalization (Gygli et al., 2018) (Western European (n 5 92) vs non-Western European (n 5 61)) resulted in consistently and significantly slightly higher mean values measured in the non-Western European group for the constructs sensing and transforming (DCs microfoundations), innovation-oriented lean culture (moderator) and process innovation (dependent variable).As measures for only a few of the constructs are partially different between the groups and are all consistent (comparatively higher in the non-western European group), no distorting effect of the geographical location-bound influences are indicated.
To evaluate the significance of the results, we interpret the model's explanatory power through the path coefficients, significance level and Cohen's f 2 and adjusted R 2 values (Benitez et al., 2020).The path coefficients of the proposed relationships are all significant (except H4b) and range between 0.137 and 0.701.Adjusted R 2 values for the model predicting DCs microfoundations (0.762) and for the model predicting process innovation (0.551) show high predictive power above the 0.36 threshold (Wetzels et al., 2009).Cohen's f 2 statistic measures the relative size of each incremental relationship in the model.Values higher than 0.020 (weak), 0.150 (medium) and 0.350 (large) indicate effect sizes (Benitez et al., 2020).The f 2 values of the confirmed relationships in the model ranged from 0.043 to 1.043.Consequently, good explanatory power is obtained through the model.
Furthermore, based on the proposed relationships in our model, our sample size exceeds the minimum required observations (n 5 144) to make estimations with a minimum statistical power of 80%, a significance level of 5% and with the possibility of estimating R 2 values from 0.10 and higher (Hair et al., 2016).Additionally, we conducted a predictive power analysis for the statistical model, using PLSpredict, to observe how adequate the results are for generating out-of-sample predictions.The results satisfy the established requirements and show high predictive power for the dependent variables (DCs microfoundations Q 2 predict 5 0.524; Process innovation Q 2 predict 5 0.436) (Shmueli et al., 2019).Finally, the standardized root mean squared residual (SRMR) shows the discrepancy between the sample covariance matrix and the model covariance matrix, which allows us to evaluate the goodness of fit for the structural model.The estimate shows an SRMR 5 0.07 lower than the upper threshold (0.08) (Benitez et al., 2020) thus ensuring a good model fit.

Robustness analyses
To assess the robustness of our results, we carried out additional estimations of the structural model.First, we used an independently collected secondary measure of firm performance obtained from the Orbis company database to replace process innovation as the dependent variable.Specifically, we used ROCE, a financial ratio that measures firm's profitability and capital efficiency, as the alternative dependent variable, as this is an accepted measure of a firm's growth opportunities (Hernandez-Vivanco et al., 2019;Martins and Lucato, 2018).The availability of data in the Orbis database reduced the sample size for this test to 100 firms.The results obtained showed similar significance supporting the hypotheses of the proposed IJOPM 42,13 model, including the positive and significant relationship between DCs microfoundations and ROCE (λ 5 0.129, p < 0.05).This result confirms the positive relationship of DCs not only with process innovation, but also with long-term financial measures.
Second, to analyse potential endogeneity issues, we used an instrumental variable twostage least squares (2SLS) approach (Lu et al., 2018).For the relationships between lean operations and lean supply chain practices and DCs microfoundations, we selected a single plausibly exogeneous variable that measures the firm's lean experience as an instrumental variable.Lean experience suffices the "relevancy condition" as it directly and strongly relates to the implementation of lean practices but does not necessarily relate to DCs microfoundations.For the relationship between DCs microfoundations and process innovation, we selected a four-item variable that measures product innovation as an instrumental variable that is directly related to DCs microfoundations but not necessarily to process innovation.We used the Wu-Hausman F test and the Durbin-Wu-Hausman χ 2 test for assessing the 2SLS estimator and endogeneity (Lu et al., 2018) and conducted Stock-Yogo to assess instrument variable relevancy (Lu et al., 2018).All the tests show that endogeneity is not a significant concern (Appendix 2).
Third, we re-estimated the model by replacing the dependent variable, process innovation, with two alternative variables constructed from questionnaire scales measuring strategic and financial performance and related to DCs in previous literature (Jantunen et al., 2018;Schilke, 2014).The results confirmed our main analysis results along with significant and positive relationships of DCs microfoundations with both strategic and financial performance (λ 5 0.532, p < 0.01; λ 5 0.420, p < 0.01).These results provide assurances of the robustness of our results.
Fourth, we estimated an alternative model that considers DCs microfoundations as an exogenous rather than a mediating variable.This alternative model shows a fit index with less explanatory power than our original model (SRMR 5 0.089).
Fifth, we tested the direct effect of lean operations and lean supply chain practices on process innovation.With bootstrapping, our results show significant positive relationships for both cases (lean operations practices-process innovation λ 5 0.130 t-value 5 2.524***; lean supply chain practices-process innovation λ 5 0.142 t-value 5 3.843***), confirming the existence of partial mediating effects through DCs microfoundations.
Finally, as noted in the model development section, we retained two scale-items that had slightly lower factor loadings than the recommended minimum of 0.7.To check for spurious results based on that inclusion, we carried out the estimation of the model without those indicators.The significances of the relationships obtained in these alternative models were not different from those in the original model.

Discussion
The main goal of this study is to analyse the relationships between lean operations and lean supply chain practices, learning-and innovation-oriented lean cultures, DCs microfoundations and process innovation.In this section, we discuss the main implications of this study for research and practice.
6.1 Lean practices and DCs microfoundations development H1 and H2 propose positive impacts of lean operations and lean supply chain practices on DCs microfoundations.Our results support these hypotheses and add to existing empirical evidence of the value of lean practices for operational performance (Dal Pont et al., 2008;Onofrei et al., 2019;Shah and Ward, 2003).Our findings show that lean practices improve the capabilities for sensing opportunities and threats, seizing them and transforming resources to respond to them.These capabilities enhance process innovation that facilitates firm adaptation to environmental changes.Our research offers empirical evidence for the DCs framework explaining the adaptive capacity resulting from lean implementations (Anand et al., 2009;Dobrzykowski et al., 2016).By providing evidence of the positive relationships between lean practices and DCs microfoundations, our research sheds light on the mechanisms through which lean firms respond to environmental changes.
Related to lean operations practices, our conclusions extend previous literature on the benefits of lean, which include lower response time, higher quality, increased flexibility (Chavez et al., 2013;Dal Pont et al., 2008), pollution prevention and employee health improvement (Yu et al., 2020).Further, DCs development through lean operations practices corroborates DCs literature that suggests that better firm responses to outside changes are based on improved internal (quality) systems, how these have been utilized, deployed and enhanced (Dobrzykowski et al., 2016;Teece et al., 1997)."Competitive advantage is not just a function of how one plays the game; it is also a function of the "assets" one has to play with, and how these assets can be deployed and redeployed in a changing market" (Teece et al., 1997, p. 529).Lean operations practices comprise tools and work programmes that, through information sharing and knowledge accumulation, facilitate knowledge codification, articulation and experience, the three predominant requirements for DCs development (Barrales-Molina et al., 2013;Zollo and Winter, 2002).
Similarly, our results extend previously recognized benefits of lean supply chain practices such as collaboration, integration and quality improvement (Cagliano et al., 2006;Powell and Coughlan, 2020;Tortorella et al., 2017) to DCs development within firms, which enables better adaptation to change environments.The outcome of our study is in line with the literature on DCs, which states that alliance management represents one of the clearest examples of DCs that facilitates environmental adaptation through inter-organizational collaboration (Schilke, 2014).Similar to the impact of alliance management, lean supply chain practices enhance the inter-organizational knowledge derived from customers and suppliers' collaboration, leading to higher levels of firm responsiveness to environmental changes (Gligor et al., 2015).

The moderating effects of learning-and innovation-oriented lean cultures
Our results show that both cultural orientationsstrengthen the relationships between most lean practices and DCs microfoundations development.This shows that understanding lean exclusively as an implementation of practices is short-sighted; it is necessary to incorporate lean-supportive cultural elements.These results confirm contemporary views on lean implementation ( Ahlstr€ om et al., 2021;Cusumano et al., 2021).Our results support the notion that learning-and innovation-oriented lean cultures strengthen the relationship between lean practices and the firm's capacity to be on the lookout for and be ready to react to changes in the environment.This resembles the concept of a "lean competitor" that operates in highly dynamic and complex contexts (Ward et al., 2007), thereby departing from an exclusive focus on operational efficiency.
Our results describe how learning amplifies the effectiveness of lean practices (Onofrei et al., 2019;Tortorella et al., 2019).Nevertheless, our empirical analysis did not support the existence of a positive moderating effect of an innovation-oriented lean culture on the relationship between lean supply chain practices and DCs development (H4b).This can be interpreted as suggesting that lean supply chain practices are, by nature, externally oriented and help create an understanding of the competitive environment.Hence, these practices already enable customer-and supplier feedback and the exchange of new ideas, thereby diminishing the moderating impact of a lean cultural innovation-orientation.On the other hand, lean operations practices such as pull systems and setup up time reductions, with their internal focus, are predisposed more than lean supply chain practices, to target efficiency while overlooking customer and supplier requirements.

IJOPM 42,13
This result agrees with previous literature such as Azadegan et al. (2013) who, contrary to hypothesized, found that lean supply chain practicesunlike lean operations practicesimprove performance in dynamic environments, thanks to "intangible benefits" (Kaynak and Pagan, 2003), which include real-time information sharing or collective problem-solving.We propose that these intangible benefits of lean supply chain practices, such as close contact with customers and suppliers, feedback, information sharing or joint development of new ideas, facilitate an innovative orientation, explaining that an innovation-oriented culture does not significantly strengthen the relationships between these practices and the environmental adaptation suggested by DCs microfoundations.On the contrary, according to our results, this innovation-orientation is beneficial for realizing DCs from lean operations practices (Spear and Bowen, 1999).

DCs microfoundations and process innovation
Finally, our results provide support for H5, which asserts a positive relationship between DCs microfoundations and process innovation and thereby corroborate the previously explored relationship between process innovation and its DCs microfoundations antecedents (Piening and Salge, 2015).
Direct correlations between enhanced process innovation capabilities and lean implementations have been explored (M€ oldner et al., 2020).Our research provides evidence for the organizational processes in lean implementations that ultimately lead to process innovation.Specifically, lean operations and lean supply chain practices, strengthened by learning-and innovation-oriented lean cultures, facilitate sensing, seizing and transforming capacities, which benefits process innovation.These results provide insights into the "black box" of DCs (Pavlou and El Sawy, 2011) and add to the results of the few previous studies (e.g.Jantunen et al., 2018;Nikookar and Yanadori, 2021) that examine the outcomes of sensing, seizing and transforming capabilities.

Theoretical contributions
We make three contributions to the theoretical perspective of continuous improvement as DC.First, by assessing the impact of lean practices combined with lean cultures on DCs microfoundations, we extend the benefits of lean implementation identified in previous studies beyond operational performance.Second, this study provides empirical evidence on how lean cultures oriented towards learning and innovation strengthen the relationship between lean practices and DCs development.Third, while DCs are hard to measure without having incidents of changes that activate them, our results do show that DCs microfoundationssensing, seizing and transformingare associated with process innovation, which can be useful for firms in reacting to a variety of environmental changes.
Most importantly, our research brings out the importance of microfoundations of DCs as mechanisms for sustainable results from lean practices.Our findings highlight the importance of lean practices and supportive cultural orientations in firms pursuing successful DCs development.We show that lean practices are important for firms' strategic orientation in response to environmental changes.This conception of lean, as an initiative that facilitates firm adaptation, supports the phenomenon-based perspective on lean (Cusumano et al., 2021), thereby expanding lean's myopic goal of merely improving efficiency.This view holds that the rigid application of lean practices reduces value creation from it and that it is necessary to allow for flexibility in lean practice adoption.Flexibility (i.e.learning and innovating) in the process of lean practice diffusion allows lean practices to take different forms in different contexts (Netland and Powell, 2016).Our results provide evidence that lean practices adoption in learning-and innovation-oriented settings hold significant potential for opportunities and threats detection and consequent assessment and action capabilities (process innovation).This aspect has relevance for environmental changes that lead to widespread disruptions such as those caused by the Covid-19 pandemic or the Russian invasion of Ukraine.These contexts demand from firms an extraordinary capacity to adapt.Previous views that remark on the potential of lean to respond to environmental changes (Van Hoek, 2020) align with our findings, which indicate that it is not only about reducing inventory levels or increasing efficiency, but rather it is about adopting lean practices and harnessing a supportive cultural orientation that fosters capabilities for adaptation to (extreme) environmental changes.

Managerial implications
Our results show managers that when implemented in full, lean positively contributes to the development of DCs microfoundations, which in turn fosters higher levels of process innovation.DCs microfoundations and process innovation enable the firm to respond appropriately to environmental dynamism.Further, in describing the issue of firms not being able to realize the complete benefits of lean by implementing its practices, Spear and Bowen (1999) emphasized the importance of the supportive cultural aspects needed to embrace the lean philosophy.Our research makes that advice practical for managers by pointing to actionable items in practices and cultural orientations that are critical for a holistic implementation of lean.We offer three specific insights for managers: First, we highlight the importance of holistic approaches to adopting lean practices.Our results suggest that long-term sustainable benefits require managers to implement lean as collections of both internally and externally oriented lean practices.Second, we emphasize the importance of learning-and innovation-orientations in lean implementations.Lean implementation leaders should actively influence the creation of such cultural orientations and remove obstructions to nurturing supportive contextual conditions.Third, managers should be cognizant of the value of building DCs when implementing lean.DCs microfoundations development, through lean practices and supporting cultures, help to maintain readiness to respond to dynamic environments.By maintaining a complete repertoire of lean practices and nurturing learning-and innovation-oriented cultures, managers and frontline employees can select the right combination of tools and adopt appropriate directions for responses to gradual changes as well as sudden perturbations in volatile and uncertain environments.

Conclusion, limitations and future research implications
Overall, our adaptive-enabling perspective on lean implementation, as a long-term, DCs developing, strategic approach, shows the potential of lean to fit firms' strategies and competitive environments, thereby providing a perspective on lean beyond the traditional focus on consistent standardized operations and waste and variability reduction.Our findings reveal how lean implementation enables the development of firm-specific processes for the functioning of continuous improvement systems, which subsequently facilitate responding to environmental dynamism, thereby increasing firms' chances of sustaining their competitive advantages.
Our research has limitations, some of which are related to the sampling and analysis.First, the response rate possibly limits the generalizability of the findings and the robustness of the conclusions.Second, our data were obtained from single respondents in each organization and are self-reported.Third, although informed and experienced respondents across the globe responded to our survey, a majority of our sample (65%) consists of European companies.Finally, the cross-sectional nature of this research does not allow observation of the relationships between variables over time, thereby limiting the understanding of the results.

IJOPM 42,13
Toward future opportunities for related research, DCs are built for responding to environmental changes.Thus, future research can include the dynamism of the environment as a moderating factor in the relationships analysed and perhaps differentiate pre-and post-Covid-19 contexts.Including dynamism would lead to more complete analyses of lean practices implementation for DCs development under changing environments and would allow drawing stronger and perhaps different conclusions about the potential of lean implementation to contribute to organizational success.Further, a study of the organizational routines that develop DCs would provide a more in-depth understanding of the relationships suggested in this study.
Figure 1.Proposed model