Lean leadership across different national cultures: a comparative study

Abstract

Most studies on organisational lean transformation have been conducted in the North-Western hemisphere. This questions the cross-cultural generalisability and understanding of managers’ leadership style that is required to effectively guide lean transformations. Our mixed-methods study compares effective Brazilian to Dutch lean managers’ behaviours and values and builds on a total of 43 in-depth interviews and 100 surveys with focal managers, their bosses, and subordinates. While self-transcendence and openness-to-change values were similar across cultures, Brazilian lean managers were perceived to show more relations-, change-, and task-oriented behaviours than the Dutch ones. The Brazilian managers also strongly identified with lean leaders, noting that lean values were quite different from Brazilian ones. Thus, Identity Theory is introduced to explain these striking differences and three propositions are formulated to guide future longitudinal mixed-methods research across the globe. Multinational organisations are advised to build a strong ‘lean identity’ across their plants to stimulate lean-leadership development.

Keywords:

• Lean leadership
• national culture
• lean maturity
• identity theory

1. Introduction

Over the past decade, scholars have started accounting for the so-called ‘soft side’ or ‘human dimension’ of lean adoption (Bortolotti, Boscari, and Danese 2015; Magnani, Carbone, and Moatti 2019). Also more popular literature has highlighted the need to address the whole lean ‘iceberg’ for organisational lean transformations to work out well: hence, including both the tools and the human factors (Hines et al. 2008). One human factor that has been increasingly associated with lean management is leadership style (Netland, Powell, and Hines 2019; Da Costa Noguiera, Sousa, and Moreira 2018; Bianco et al. 2021), which can be characterised by specific attributes, such as observable behaviours and underlying values (Van Dun, Hicks, and Wilderom 2017). Although so far the majority of studies have assumed a generalisable leadership style that is most effective for lean adoption, which also informed the content of (in-company) lean leadership development programmes, recently lean scholars noted that these attributes tend to be influenced by national cultures (Seidel et al. 2019). For instance, following Hofstede’s and Schwartz’s frameworks, attitudes towards lean adoption and managers’ behaviours can be negatively affected by high levels of ‘masculinity’ (Erthal and Marques 2018), as well as ‘power distance’ and ‘uncertainty avoidance’ (Taras, Steel, and Kirkman 2012).

However, when accounting for cultural differences, lean studies have been mostly situated in Northern America or Western Europe, with India as a rare exception (Danese, Manfè, and Romano 2018). Consequently, comparisons of managers’ effective lean behaviours and values across the globe are lacking, limiting our understanding of whether and how the lean approach and national cultures affect each other (Danese, Manfè, and Romano 2018). This leads to less fruitful investments in lean leadership development initiatives, and, consequently, less sustainable lean adoptions in certain parts of the world, whereas increasingly globalising supply chains demand world class manufacturing. In addition, so far theory is lacking to explain why lean might potentially fare better under certain conditions (Åhlström et al. 2021). Thus, we explored lean leadership behaviours and values of managers situated in an under-researched but economically developing continent, namely Latin America (in Brazil to be precise), and related the findings to data from a previous Dutch study (see, Van Dun, Hicks, and Wilderom 2017) to answer the following research question: How may effective lean managers’ behaviours and underlying values overlap or differ across national cultures?

This study answers the call for considering potential cross-cultural differences in lean management studies (Erthal and Marques 2018; Danese, Manfè, and Romano 2018), as well as the identified need for more empirical studies in Latin America (Aguinis et al. 2020). As such, this study provides in-depth insights into the impact theory-based cross-cultural differences might have on how we define effective lean leadership behaviours and values, for instance when it pertains the lean adoption process of multinational firms. Interestingly, the findings brought to light that cross-cultural differences seem to have less of an impact on lean leadership behaviours and values compared to (1) one’s personal identification with being (or becoming) an effective lean leader; and (2) one’s lean maturity. Thus, Identity Theory (Ashforth and Schinoff 2016; Stryker and Burke 2000), which has been associated with lean before (Skovgaard-Smith, Soekijad, and Down 2020), is introduced to explain these striking differences. This study also has clear practical importance, as it contributes to managers’ understanding of what constitutes effective lean leadership across countries (Meyer 2017). Given the popularity of lean management across the globe, it is essential to learn that building an organisational ‘lean identity’ is crucial for developing effective lean leadership within organisations, even in countries that could be considered polar in terms of their national-cultural characteristics. Such knowledge has important implications for developing leadership recruitment and development programmes, and will, in turn, lead to more effective and sustainable lean adoption across countries.

In what follows, first the theoretical roots and the mixed-methods research design are explained, in Section 2 and 3, respectively. The empirical findings presented in Section 4 are then interpreted and further explored in Section 5 by utilising Identity Theory, which may contribute to advancing cross-cultural lean leadership theorising. This section also presents three propositions for future research that stem from our study’s findings. Section 6 then elaborates on the managerial implications, whereas Section 7 acknowledges the limitations of our study as well as suggestions for future research.

2. Theoretical background

2.1. Lean leadership

Lean management aims to continuously improve work processes for an increased customer value, by reducing waste through solving persisting operational problems within the organisation and/or supply chain (Womack, Jones, and Roos 1990). Since it gained popularity in the 90s (Samuel, Found, and Williams 2015), organisations have realised that adopting lean often requires a complete transformation towards becoming a learning organisation (Hines, Holweg, and Rich 2004; Tortorella et al. 2020). In the past decades, scholars and practitioners alike have grown their awareness of the importance of managers’ role in such a lean transformation. Operations Management scholars tend to refer to these managers as ‘lean leaders’ (Netland, Powell, and Hines 2019; Seidel et al. 2019; Van Dun, Hicks, and Wilderom 2017; Van Elp, Roemeling, and Aij 2022; Mann 2009), i.e. managers at different hierarchical levels that effectively lead lean initiatives in their unit. In their literature review, Seidel et al. (2019) distinguished publications on the role of lean leader’s attributes, their influence process, and the impact of context on lean leadership. They also stressed that lean leadership should not be considered as a new style but rather an application of generic leadership styles in a lean work context.

In particular, given its focus on embracing continuous improvement and innovation, lean leadership may resemble the transformational leadership style (Seidel et al. 2019). Through the emphasis on creating a shared vision to guide inspirational changes in tandem with committed employees, transformational leadership is one of the most prevalent leadership styles in the domain of Organisational Behaviour, alongside more transactional leadership (Siangchokyoo, Klinger, and Campion 2020). When it comes to their behaviours, transformational leaders tend to display mainly relations-oriented behaviours (e.g. showing individual consideration and listening) and change-oriented behaviours (e.g. encouraging innovation and facilitating learning) (Behrendt, Matz, and Göritz 2017), while task-oriented behaviours (e.g. sharing information and task monitoring) tend to be associated with transactional leadership (Hoogeboom et al. 2021). Transformational leader behaviours are associated with positive outcomes including organisational change (Peng et al. 2021), idea generation and implementation (Watts, Steele, and Den Hartog 2020), and follower job satisfaction/performance (Siangchokyoo, Klinger, and Campion 2020). Transformational leaders are assumed to also influence their followers through their espoused values (Siangchokyoo, Klinger, and Campion 2020).

Previous studies have indeed suggested that values drive people’s behaviours. For instance, the refined basic values theory, developed by Schwartz et al. (2012), distinguishes universal individual-level values grouped into four clusters on two continua, namely self-transcendence vs. self-enhancement, and openness-to-change vs. conservation. The different combinations and relative importance of these sometimes conflicting values then inform one’s behaviour and explain differences among individuals’ behaviours (Schwartz et al. 2017; Torres, Schwartz, and Nascimento 2016).

More specifically, the few studies on lean leadership grounded in Schwartz’s basic values theory found mainly self-transcendence and openness-to-change type work values to promote improvement behaviours that benefit others in the organisation (Van Dun, Hicks, and Wilderom 2017; Van Dun and Wilderom 2016, 2021). Given the universality of the values depicted in Schwartz’s theory (Rudnev, Magun, and Schwartz 2018), individuals embracing the lean philosophy and its values are expected to behave consistently and in a similar way across different cultures. Still, the relative emphasis placed on certain values might differ across cultures which means that lean managers’ behaviours might not be so identical across the globe, and therefore national cultural differences are considered to be key factors in explaining the variations in behaviours.

2.2. National cultural differences

In the cross-cultural comparison literature, national culture is defined as ‘a unique set of attitudes, beliefs and behaviours shared by a group of people’ (Matsumoto 2007, 1291; Hofstede, Hofstede, and Minkov 2010). This set identifies preferred ways of thinking and acting that can ultimately influence, for instance, communication patterns and, thus, leadership styles (Meyer 2017).

To better understand such influence, one of the most famous models developed to map out national-cultural differences is Hofstede’s cultural taxonomy (Meyer 2014). Hofstede, Hofstede, and Minkov (2010) identified six national-cultural dimensions, namely:

1. power distance: The extent to which the less powerful members of institutions and organisations expect and accept that power is distributed unequally;

2. individualism-collectivism: The extent to which the ties between individuals are loose vs strong. For instance, whether one is expected to look after themselves and their immediate family vs after a larger group;

3. masculinity-femininity: The extent to which social gender roles are clearly distinct vs whether those tend to overlap;

4. uncertainty avoidance: The extent to which individuals feel threatened by uncertain or unknown situations;

5. Confucian dynamism (or long-term vs short-term orientation): The extent to which individuals foster virtues oriented towards future rewards vs towards the past and present; and

6. indulgence-restraint: The extent to which individuals allow relatively free gratification of basic and natural human drives related to enjoying life and having fun vs suppressing them by means of strict social norms.

Although not immune to criticisms (McSweeney 2002; Taras, Steel, and Kirkman 2012), many Operations Management studies integrating national culture and effective lean (leader) behaviours have predominantly relied on Hofstede’s dimensions mostly given its easy accessibility, popularity and clarity (Erthal and Marques 2018; Cagliano et al. 2011).

Yet, several other reliable cultural frameworks exist (Kaasa 2021), such as Schwartz’s (2011) cultural values orientation theory. Albeit more complex and elaborated, this theory is particularly interesting for our research since it specifies universal, national-level values that underpin (national) cultures (Fog 2021). Schwartz’s theory comprises the following three cultural value orientations forming a circumflex and based on six polar a priori dimensions: ‘autonomy (intellectual and affective) vs. embeddedness’, ‘egalitarianism vs. hierarchy’, and ‘mastery vs. harmony’. The validity and reliability of this theory has been corroborated in numerous studies across the globe (e.g. Morales et al. 2019; Lee et al. 2011). Over the past decade, research has also started comparing Schwartz’s and Hofstede’s theories in the attempt of offering comparisons and parallels between these pivotal frameworks (see, Kaasa 2021; Fog 2021; Maleki and De Jong 2014; Dobewall and Rudnev 2014). See, also, Table 1 for a summary. As presented in Table 1, previous studies have shown that Schwartz’s ‘hierarchy vs. egalitarianism’ can be associated with Hofstede’s ‘high-low power distance’, Schwartz’s ‘mastery vs. harmony’ with Hofstede’s ‘masculinity-femininity’, and Schwartz’s ‘autonomy (intellectual and affective) vs. embeddedness’ with Hofstede’s ‘individualisms-collectivism’. The remaining Hofstede’s dimensions are blended and part of Schwartz’s circumflex.

Table 1. Literature-based comparison of national-level cultural values to individual-level work values and behaviours.

National level		Individual level
Hofstede’s dimensions^a	Schwartz’s cultural values	Schwartz’s basic human values		Lean values^b	Lean behaviours^c
High power distance	Hierarchy	Power	Self-enhancement	1. Achievement-orientation 2. High-quality Performance  3. Self-reflection	Task-oriented Behaviours: initiating structure sharing factual information with employees offering timely problem-solving support correcting and task monitoring nurturing team members’ skills
Masculinity	Mastery	Achievements
Individualism	Autonomy (intellectual and affective)	Stimulation Self-direction	Openness to Change	4. Continuous Improvement  5. Courage  6. Creativity  7. Innovation  8. Freedom of choice  9. Open minded 10. Constructive feedback	Change-oriented Behaviours: learning from mistakes fostering intellectual stimulation soliciting employees’ ideas
Low power distance	Egalitarianism	Universalism	Self-Transcendence	11. Honesty 12. Participation/Teamwork 13. Responsibility 14. Helpfulness 15. Equality 16. Trust in people 17. Integrity 18. Customer focus 19. Justice 20. Information sharing 21. Employee potential	Relations-oriented Behaviours: being considerate showing respect and appreciation rewarding self-management skills demonstrating humility active listening and admitting mistakes engaging employees
Femininity	Harmony	Benevolence
Collectivism	Embeddedness	Tradition Conformity Security	Conservation	22. Humility 23. Persistence 24. Respect for People

^aPapers supporting the parallels between Hofstede and Schwartz’s models: Kaasa (2021), Fog (2021), Dobewall and Rudnev (2014), Maleki and de Jong (2014), Nardon and Steers (2009), Basabe and Ross (2005), and Schwartz (2004). Hofstede’s uncertainty avoidance, Confucian dynamism, and indulgence/restraint dimensions are blended in the other Schwartz’s dimensions. ^bRetrieved from: Van Dun and Wilderom (2016, 2021) and Van Dun, Hicks and Wilderom, 2017. Underlined values are from Brown and Treviño (2009). ^cRetrieved from Baer and Frese (2003), Behrendt et al. (2017), Camuffo and Gerli (2018), Camuffo and Micelli (1997), De Treville and Antonakis (2006), Lam et al. (2015), Poksinska et al. (2013), Robinson and Schroeder (2009), Spear and Bowen (1999), Staats et al. (2011), and Van Dun et al. (2017).

Table A1. Values Q-sort forced distribution.

Acknowledging these parallels allows us to rely on a solid theory specifically built around national values, i.e. this paper’s focus, whilst keeping an eye on Hofstede’s framework. Accounting for Schwartz’s cultural values orientation theory also enables us to draw strong associations between national values and Schwartz’s basic values theory, which has been used before when studying lean manager’s values, as noted in the previous section.

Since national culture is a key contextual variable that impacts ‘the nature of the social interactions and values, and therefore on the relationship between lean leaders and their followers’ (Seidel et al. 2019, 8), it is now crucial to consider its interplay with lean leadership.

2.3. Impact of national cultural differences on lean leadership

Various scholars have explored lean leadership in different national cultures, with sometimes contradictory and counterintuitive findings. For instance, Gelei et al.’s (2015) study of Hungarian lean leaders highlighted not only the importance of leaders’ communicative behaviours (e.g. improvement-oriented, consultative, and inspirational), but also the weak positive impact of their micromanaging and non-delegation types of behaviours. The positive outcome of leader’s micro-management and not delegating tasks to their followers is not surprising given Hungary’s relatively high scores on power distance, masculinity, and uncertainty avoidance (Taras, Steel, and Kirkman 2012); for instance, micro-management and not delegating (improvement) tasks might show the leader’s superiority in problem solving. Italian lean managers, despite the relatively high score of Italy in the power distance dimension implying authority and hierarchy, expressed more supportive behaviours rather than micromanaging, as it was instead expected (Camuffo and Gerli 2018). In Sweden, lean leaders were found to contribute to individual empowerment and teamwork together with creating group consensus and harmony, e.g. through stimulating joint goal achievement and learning (Poksinska, Swartling, and Drotz 2013). In line with this, Van Dun et al. (2017) found that Dutch effective lean managers showed significantly more relations-oriented behaviours (e.g. active listening and agreeing) as well as ‘positive’ task-oriented behaviours (e.g. informing by structuring conversations), which perfectly matches Hofstede’s dimensions of low power distance and masculinity. They also noted significantly less ‘negative’ behaviours such as task monitoring, providing negative feedback and defending one’s own position (Van Dun, Hicks, and Wilderom 2017). Similar behavioural patterns were revealed by Van Dun and Wilderom (2021). Indeed, people in North-European countries tend to treasure such empowerment and harmony aspects far more than in Latin American countries, in which a more confrontational and hierarchical relationship between team members is to be expected (Cagliano et al. 2011; Poksinska, Swartling, and Drotz 2013).

These results were also reflected in Erthal and Marques’ (2018) work. They found that power distance and masculinity in Japanese organisations tended to hinder lean adoption. This is most likely due to the reticence of younger employees to share their opinions with senior members, since this behaviour is perceived as rude and questioning seniority knowledge. Strikingly, it was implied that the country where the lean philosophy originated did not have an ideal score on its cultural dimensions for lean to be implemented: instead, a low-middle score on power distance and masculinity seems to be ideal for lean to flourish (Erthal and Marques 2018). Similarly, they argued that Northern Europeans might struggle due to their high individualism and low uncertainty avoidance and organisations in Latin America might also face barriers when implementing lean, given their high power distance, low uncertainty avoidance, and relative short-term orientation (Erthal and Marques 2018). Nevertheless, from our literature review we found that the majority of lean leadership studies have been performed in Europe, whereas examinations of lean leadership in Latin America are scarce (with as notable exceptions: Tortorella et al. 2018; Tortorella, Van Dun, and De Almeida 2019). These few available studies in Brazil also showed contrary results. Whereas Tortorella et al. (2018) found that task-oriented leader behaviours were associated with higher levels of lean manufacturing implementation than leaders displaying relations-oriented behaviours, Tortorella, Van Dun, and De Almeida (2019) found that lean leaders tend to develop both their task- and relations-oriented behavioural repertoire over time, as they grow more mature in terms of lean adoption.

Hence, combining these facts, we decided to select as country context for our study the Netherlands as a North-European country and Brazil as a Latin American country. Furthermore, on the basis of both Schwartz’s and Hofstede’s frameworks, Brazil and the Netherlands provide evident differences in some important dimensions particularly relevant to lean: Table 2 clearly depicts that Brazil and the Netherlands score quite the opposite on each of the dimensions, except for indulgence vs. restraint.

Table 2. Hofstede et al.’s (2010) dimensions: Brazil versus the Netherlands.

Dimensions	Brazil	The Netherlands
Power distance	69 (High)	38 (Low)
Individualism vs collectivism	38 (Collectivism)	80 (Individualism)
Masculinity vs femininity	49 (Masculinity)	14 (Femininity)
Uncertainty avoidance	76 (High)	53 (Medium)
Confucian dynamism (or long-term vs short-term Orientation	44 (Long-term orientation)	67 (Short-term orientation)
Indulgence vs restraint	59 (Indulgence)	68 (Indulgence)

Note: values are average and score out of 100.

From a more deductive approach, by combining Schwartz’s and Hofstede’s works (Tables 1 and 2), relations-oriented behaviours, such as recognising and supporting, are expected to be shown more frequently by Brazilian lean leaders because Brazil scores higher on Hofstede’s collectivistic dimension and Schwartz’s embeddedness pole. On the contrary, change-oriented behaviours, which encourage innovation, are proposed to be less shown by Brazilian lean managers less compared to Dutch ones, most likely due to Brazil scoring lower on individualism and higher with regards to Hofstede’s ‘uncertainty avoidance’ and Schwartz’s autonomy dimension. Similarly, task-oriented behaviours, among which problem solving and monitoring operations, are also not expected to be frequent in Brazil, perhaps due to the high level of Hofstede’s masculinity and power distance, or Schwartz’s mastery and hierarchy, compared to the Netherlands. In sum, rigorous and in-depth explorations and comparisons of how effective lean managers’ values and behaviours differ across two different cultures, like Brazil and the Netherlands, is still missing. In this exploratory mix-method study we thus aim to address this negligence.

3. Methodology

3.1. Research design

To explore whether lean leader values and behaviours might differ in an entirely different culture compared to the North-Western, we conducted a comparable follow-up study of Van Dun, Hicks, and Wilderom (2017) with a Brazilian sample. This research design was called for by Danese, Manfè, and Romano (2018, 596): ‘exploratory studies should replicate the successful cases achieved in some countries to highlight how to adapt the lean approach to the NC of less explored countries, thus providing suggestions for gaining the maximum from lean implementation.’

As such, we gained access to the methods and data of Van Dun, Hicks, and Wilderom (2017) and gathered new data about Brazilian lean middle managers. Following the requirements of Bonett (2021): (a) the population sample was similar to the one used in the original study (i.e. effective lean middle managers, their bosses, and subordinates); (b) the same mixed-methods research design and treatment conditions were applied: surveys and interviews; and (c) the same variables were used as well as (d) the same measurement scales.

3.2. Sampling and sample

A call for participants was distributed during the first author’s presentation at a large, Brazilian lean management conference, with four lean-exemplar company visits initiated by a Brazilian university, and the network of a lean consulting firm. Managers of 14 organisations in the Southern part of Brazil responded. During telephonic or in-person meetings with a senior manager of each of those organisation’s they were asked to nominate a highly effective lean middle manager within their organisation. In this screening process nine organisations dropped out, either because they felt that their middle managers were not yet effective enough, or their organisations were too small, and they did not have middle managers. The five other senior managers nominated in total seven effective middle managers leading lean-adopting departments; all seven agreed to participate. The selection procedure of the six Dutch middle managers can be found in Van Dun, Hicks, and Wilderom (2017).

Similar to the Dutch sample, the seven Brazilian managers worked in different sectors (production or service), were predominantly male, had more than one year of lean experience (three had more than five years of lean experience), and had varying department sizes (see Table 3). The focal managers had an average age of 40; one of them had a bachelor’s degree, the other six had finished a master’s program. Given the diversity of their backgrounds, the sample was deemed representative for the Brazilian population of effective lean leaders. We conducted 25 interviews with them, and their seven senior managers and 11 subordinate team leaders. In addition, we surveyed them as well as the 32 remaining subordinate team leaders (n = 57). Thus, we used purposive sampling whereby we invited all subordinate team leaders (also beyond those selected team leaders we interviewed), which enabled a representative sample of the population of interest. The majority of the team leaders who directly reported to the focal middle managers did participate in this survey; only a couple of them were not able to respond due to sickness absence or annual leave. Hence, there were no signs of non-response bias. The Brazilian sample was also comparable to the original Dutch sample utilised in this study, which consisted of in total 18 interviews (including their four senior managers, six subordinate team leaders, and two internal consultants) plus 43 completed surveys (Van Dun, Hicks, and Wilderom 2017).

Table 3. Context of the focal effective lean middle managers.

Industry					n per method		Lean tool adoption	Operational performance improvement
Brazilian managers	Department	Gender	Direct reports	Lean maturity (experience with lean in months)	Interview	Survey	Mean (std. dev.)	Mean (std. dev.)
1. IT services	Product development	Male	45	12	4	15	2.68 (0.69)	3.66 (0.72)
2. IT services	Product management	Male	35	24	4	15	2.58 (1.04)	3.41 (0.80)
3. Manufacturing	Production	Male	90	120	3	10	2.18 (0.78)	3.54 (0.54)
4. Manufacturing	Production engineering	Male	6	60	2	4	2.96 (0.44)	3.70 (0.96)
5. Manufacturing	Product management	Male	154	60	4	5	3.03 (0.50)	4.00 (0.85)
6. University	Human resources	Female	3	18	4	4	2.07 (0.57)	3.44 (0.41)
7. University	Human resources	Female	3	24	4	4	1.89 (0.75)	3.08 (0.08)
Sub total					25	57
Dutch managers^a
8. Trucks	Production	Male	22.50	120	3	3
9. Coffee	Production	Male	125	>24	1	1
10. Energy	Call centre	Male	165	18	3	12
11. Energy	Call centre	Male	200	15	4	11
12. Energy	Call centre	Male	110	9	3	8
13. Energy	Human resources	Female	22	8	4	8
Sub total					18	43
Total					43	100

^aAdapted from Van Dun, Hicks, and Wilderom (2017, 178). Note. Lean tool adoption and operational performance improvement had not been measured in the Dutch sample.

3.3. Data collection

3.3.1. Survey

A survey was distributed among all interviewees as well as all the remaining subordinate team leaders of the focal managers. The survey comprised the following validated scales (see, also, Appendix B; note that the Dutch sample only completed the ‘lean middle managerial behaviours’ scales):

Lean tool adoption. Each respondent indicated the extent to which lean tools were employed in their department, based on the list of nine lean tools identified by Fullerton, Kennedy, and Widener (2014), such as: standardisation, 5S, and kaizen. Their responses were captured on a 5-point scale with as scale labels ‘not at all’, ‘a little’, ‘to some extent’, ‘considerably’, and ‘a great deal’. The Cronbach’s alpha was 0.85.

Department-level operational performance improvement. To double-check the effectiveness of each focal middle manager, the operational performance improvement of the department led by them was measured using the six-item operations performance construct developed by Fullerton, Kennedy, and Widener (2014). We asked them to rate how each performance indicator (e.g. cycle time) had changed in the past three years. The respondents rated each item on a 5-point scale with as scale labels: ‘significantly worsened’, ‘moderately worsened’, ‘little or no change’, ‘moderately improved’, and ‘significantly improved’. Taking the example of cycle time: the higher the score, the more the cycle time had reduced in the past three years which is considered an operational performance improvement. The Cronbach’s alpha of this construct was 0.85.

Lean middle managerial behaviours. The focal managers’ behaviours were measured using the Hierarchical Taxonomy of Leadership Behaviours questionnaire, which includes three previously validated sub dimensions (Yukl 2006, 2012; Yukl, Gordon, and Taber 2002), namely: task-oriented (7 items¹; α = 0.78), relations-oriented (8 items; α = 0.85), and change-oriented behaviours (14 items; α = 0.88). The 7-point scale ranged from ‘never’ to ‘always’. An example item is: ‘My direct supervisor/I clarifies/clarify role expectations for task performance’.

3.3.2. Interviews

All 25 Brazilian and 18 Dutch one-hour interviews (with the focal managers, their senior managers, subordinate team leaders, and internal consultants) followed the same semi-structured approach (the interview guide is in Appendix A). First, open questions explored the respondent’s background and views regarding lean and leadership in general. Then, the Critical Incidents Technique (Flanagan 1954; Bott and Tourish 2016) was used whereby interviewees were asked to recall a moment or situation the focal lean manager displayed an act of truly effective lean leadership. Finally, the Q-sort methodology (Brown 1996; Fu et al. 2010) was employed whereby interviewees were asked to sort 24 leader work values (see, Appendix A). The Q-sort followed a forced-distribution procedure in terms of how important they perceived the values to be for the focal manager, on a 9-point scale ranging from 1 = ‘not important’ to 9 = ‘highly important’. The values fit Schwartz et al.’s (2012) four values clusters: self-transcendence (e.g. ‘participation and teamwork’), openness-to-change (e.g. ‘continuous improvement’), self-enhancement (e.g. ‘achievement-orientation’), and conservation (e.g. ‘respect for people’). All audiotapes were transcribed.

3.4. Data analysis

3.4.1. Quantitative data analysis

In terms of the quantitative data, we combined the views from both the focal managers and subordinate team leaders. First, based on the survey, their department’s average lean tool adoption was calculated for each focal manager’s department. We also determined each department’s operational performance improvement to double-check the manager’s effectiveness via more objective measures than only through senior managers’ nomination.

To compare the Brazilian and Dutch datasets in terms of lean managers’ values, means and standard deviations were calculated for both Q-sort datasets collected during the interviews, after which we ranked all 24 values from highest to lowest scores. This enabled us to compare both rankings, in particular the ten highest ranked values for each countries’ sample. In addition, after screening the normality of the data, we performed either independent samples t-tests or Mann–Whitney U tests to determine significant differences between both countries’ samples in terms of each of their score on Schwartz’s (2012) clusters: self-transcendence, self-enhancement, openness-to-change, or conservation. In addition, for exploratory purposes we combined both countries’ datasets and ran exploratory independent samples t-tests comparing the managers based on their lean adoption maturity level, which has been shown an important factor before (Tortorella, Van Dun, and De Almeida 2019; Van Dun and Wilderom 2021; Lameijer et al. 2021). In particular, we compared more mature lean middle managers (i.e. more than two years of lean adoption experience) with the managers who had less experience with lean adoption (i.e. two years or less).

In terms of the behavioural survey data, we used Harman’s single-factor test with an exploratory factor analysis to verify common method bias (Malhotra, Kim, and Patil 2006). The test encompassing all variables yielded a first factor that accounted for 33.80% of the total variance. Because no single factor explained most of the variance (i.e. > 50%; Hair et al. 2014), common method bias issues were disregarded. Then, we checked whether they were normally distributed (Hair et al., 2014). Although the change-oriented behaviour scale matched the normality criterion, the task- and relations-oriented behaviour scales were not normally distributed. Hence, we ran an independent samples t-test for the change-oriented behaviour scale and a Mann–Whitney U test for the other two scales to compare the perceived lean manager task-, relations-, and change-oriented behaviours across both samples. This way we meant to examine to what extent the data matched our assumption that there might be cross-cultural behavioural differences between Brazil and the Netherlands. We also explored possible differences between the more mature lean middle managers and the managers who had two years or less experience with lean adoption.

3.4.2. Qualitative data analysis

The Brazilian and Dutch qualitative data were then explored to allow us to understand the dynamics more in depth and explain potentially unexpected findings from the quantitative data analysis. The Critical Incident Technique interview transcriptions were analysed via thematic analysis (Braun and Clarke 2006) by two coders who independently read and coded four Brazilian interviews (i.e. about 10% of the entire interview dataset). This analysis was conducted deductively, by integrating the authors’ interpretation of the data with the literature-based lean values and Yukl’s (2012) task-, relations-, and change-oriented behavioural categories (see Table 1). Firstly, we read through the whole set of interviews and identified all the behaviours pointed out by our participants. Secondly, guided by Yukl’s taxonomy that we also used in the quantitative part of our analysis (i.e. task-, relations-, or change-oriented behaviours), we then precisely coded all the behaviours into their respective taxonomy (also inspired by Yukl’s items used in the quantitative part of this study; Appendix B). An example of a code we applied is: ‘task-oriented – Sharing feedback and information’. The inter-coder reliability was 87.3%, which points to reasonably strong agreement among both coders, thereby curbing researcher bias. The remaining 39 transcriptions were then coded by the two separate coders.

After coding all the data, we checked for consistency between the respondents in each case (being the focal middle manager) to explore any potential differences in terms of how they reflected on focal leader’s behaviours. Based on this screening no major differences were identified between the sources. In terms of the cross-case analysis, our scope was comparing across national cultures, meaning that we compared the quotes identified in each behavioural category to examine any potential differences between the two sub samples: the Brazilian versus the Dutch sample. In addition, the Brazilian Q-sort outcomes were compared to the Dutch sample. Thus, we assured idiosyncrasy and transparency of reasoning to guarantee a careful analysis of the qualitative data (Ketokivi and Choi 2014).

3.4.3. Triangulation of data

Triangulation is conceptualised as the utilisation of several methods for examining the same phenomenon, so that the study’s credibility increases (Johnson, Onwuegbuzie, and Turner 2007; Hussein 2009). In this step, we achieved ‘method triangulation’ by comparing the data obtained from the quantitative and qualitative analyses (i.e. surveys and interviews) to develop a chain of evidence (Carter et al. 2014). In addition, we applied ‘investigator triangulation’ given that all three authors were involved in analysing the data, among which a Dutch and a Brazilian investigator to allow for a better and more in-depth, context-sensitive interpretation of the rich data (Carter et al. 2014; Barratt, Choi, and Li 2011). Thirdly, we collected and compared data retrieved from multiple sources such as the focal middle managers, their bosses and subordinate team leaders, i.e. ‘data source triangulation’ (Carter et al. 2014), as described in Section 3.4.2.

We revisited the quantitative and qualitative analyses (see Sections 3.4.1 and 3.4.2) from both the Brazilian and Dutch datasets to document the data, including all data sources and reflections, ideas, and insights. Our findings were synthesised and categorised into three dimensions determined upfront, namely: (i) lean leader’s behaviours and values versus national cultural differences, (ii) effect of lean maturity on lean leaders’ behaviours, and (iii) explanation of effective lean leaders’ values and behaviours. We looked for commonalities among all data sources in each dimension and created a table to summarise the findings across the different methods. Details in the examples provided, arguments used by the interviewees to justify answers, the similarity in responses from maintenance and production leaders, observed characteristics, and facts were all used to support this classification (Narasimhan 2014). Such triangulation allowed us to raise insightful findings discussed in section 5 of the manuscript, which were then used as arguments to formulate three propositions for future research validation and theory testing.

4. Results

4.1. Quantitative findings

4.1.1. Lean managers’ effectiveness

As shown in Table 3, concerning the level of lean adoption at each Brazilian manager’s department, the two university units scored lowest (mean = 1.89 and mean = 2.07, respectively, indicating a little lean tool adoption, whereas the information technology and manufacturing departments were slightly more intense lean adopters (mean = 2.68). Indeed, the university units adopted predominantly standardisation, kanban, and kaizen, while the other firms’ departments adopted almost all of the lean practices we queried in the survey. In terms of operational performance improvement, all of the focal Brazilian lean managers’ departments had improved in the past three years (mean = 3.55, ranging between 3.08 and 4.00): a proxy for the focal managers’ assumed effectiveness.

4.1.2. Lean managers’ work values

Table 4 shows that the ten highest ranked values based on the Q-sort distribution were similar across the effective Brazilian and Dutch lean managers samples, although averages (on a 9-point scale) vary slightly. These values tend to be predominantly clustered as self-transcendence and openness-to-change values according to Schwartz’s basic values theory. Zooming in on the specific order of the ranked values, we did notice subtle differences. For instance, Brazilian managers placed ‘trust in people’ and ‘participation and teamwork’ at the 1st and 3rd place, respectively, whereas Dutch managers only placed them at the 8th and 6th positions. In this specific example, these values are associated with relations-oriented behaviour and their different positions could potentially help to explain the differences in terms of frequency of behaviours as will be discussed in the next section. In addition to this list, during the interviews multiple Brazilian respondents spontaneously added the importance of ‘macro managing (remain calm and in control)’. This is a more conservation-oriented type of value. Dutch respondents, on the other hand, spontaneously added ‘candour (be open and vulnerable’, which is a more openness-to-change type of value.

Table 4. Q-sorted values of the effective lean managers, Brazilian vs. Dutch sample.

	Brazilian (n = 25)		Dutch (n = 18)^a
Values (cluster)	Mean (std. dev.)	Values (cluster)	Mean (std. dev.)
1. Trust in people (ST)	7.12 (1.88)	1. Continuous improvement (OtC)	6.78 (1.52)
2. Continuous improvement (OtC)	6.96 (1.18)	2. Responsibility (ST)	6.78 (1.44)
3. Participation/teamwork (ST)	6.92 (2.00)	3. Honesty (ST)	6.50 (1.76)
4. Respect for people (CON)	6.88 (1.68)	4. Respect for people (CON)	6.39 (2.35)
5. Honesty (ST)	6.52 (1.94)	5. Integrity (ST)	6.28 (1.60)
6. Responsibility (ST)	6.36 (1.38)	6. Participation/teamwork (ST)	6.28 (2.30)
7. Customer focus (ST)	6.28 (2.22)	7. Achievement-orientation (SE)	6.28 (2.93)
8. Integrity (ST)	5.76 (2.44)	8. Trust in people (ST)	6.22 (1.86)
9. High-quality performance (SE)	5.24 (2.45)	9. Customer focus (ST)	5.89 (2.17)
10. Achievement-orientation (SE)	5.16 (1.97)	10. High-quality performance (SE)	5.67 (2.14)

Note. ST: Self-transcendence; SE: Self-enhancement; OtC: Openness to Change; CON: Conservation. ^aAdapted from Van Dun, Hicks, and Wilderom (2017, 179).

Comparing the values clusters across both samples and cultures, the t-tests showed that self-enhancement values were scored significantly higher in the Dutch sample compared to the Brazilian one (see, Table 5). On the contrary, the conservation values were scored significantly higher in the Brazilian sample. Upon comparing the lean maturity levels, we did not find any significant values-endorsement differences between managers with a high versus low lean maturity.

Table 5. Perceived behaviours and values of the effective lean managers.

	Brazilian		Dutch					Low lean maturity		High lean maturity
Values	Mean	Std. dev.	Mean	Std. dev.	t/Z	df	p	Mean/Median	Std. dev.	Mean/Median	Std. dev.	t/Z	df	p
Self-transcendence	5.60	0.42	5.38	0.48	−1.55	41	0.129	5.54	0.47	5.54	0.41	−0.013	41	0.990
Self-enhancement	4.40	1.44	5.70	1.32	3.03	41	0.004	5.09	1.49	4.62	1.62	0.94	41	0.355
Openness to change	4.20	0.63	4.35	0.56	0.80	41	0.431	4.20	0.64	4.40	0.52	−0.95	41	0.347
Conservation	5.24	1.05	4.43	1.20	−2.36	41	0.023	5.00	1.23	4.33	1.05	−0.373	41	0.724
Behaviours	Mean/Median	Std. dev.	Mean/Median	Std. dev.	t/Z	df	p	Mean/Median	Std. dev.	Mean/Median	Std. dev.	t/Z	df	p
Task-oriented	5.75	0.62	5.00	0.86	−4.29	98	0.000	5.20	0.77	5.89	0.79	−3.73	98	0.000
Relations-oriented	6.00	0.81	5.50	0.83	−2.22	98	0.026	5.62	0.85	5.70	0.79	−0.42	98	0.678
Change-oriented	5.77	0.66	4.97	0.61	6.24	98	0.000	5.32	0.72	5.78	0.74	−2.68	98	0.009

Note: Depending on the normality of the data, we report here the results of independent samples t-tests or Mann–Whitney U tests. The behaviours were scored on a 7-point Likert scale; the values were ranked on a 9-point scale.

4.1.3. Lean managers’ behaviours

In terms of behaviours, the Brazilian managers adopted significantly more change-oriented behaviours (e.g. encouraging innovation) than their Dutch counterparts, but also significantly more task- and relations-oriented behaviours (see the independent samples t-test results in Table 5). This finding was against our initial, theory-based assumption and contrasts the fact that they shared similar national-cultural and work values with the Dutch as explained above.

We also compared the perceived behaviours of the lean managers with a low lean maturity to those with a higher lean maturity (Table 5). The statistical tests showed that the more mature Brazilian and Dutch lean managers scored higher on both task-oriented and change-oriented behaviours compared to the managers with less lean experience; hence, regardless of their national-cultural background. Given the atypical findings upon comparing the managers in terms of their national-cultural background, and the fact that the Brazilian sample overall consisted of more experienced lean managers, it might thus well be that the managers’ lean maturity level also plays a key role in explaining the significant behavioural differences found here.

4.2. Qualitative findings

The qualitative findings provided more in-depth insights into the unexpected differences between the behaviours of effective lean managers in the Brazilian and Dutch samples. Table 6 provides example quotes taken from the interview transcripts from both samples and will be used to illustrate the diverse range of specific behaviours lean leaders need to display, such as: sharing feedback and information (task-oriented), providing support, encouragement, and recognition (relations-oriented), and leading (the change) by example (change-oriented).

Table 6. Illustrative quotes comparing Brazilian and Dutch sample.

Coded behaviours	Brazilian quotes and behavioural sub codes (in bold)	Dutch quotes and behavioural sub codes (in bold)
Task-oriented	Sharing feedback and information: Boss, #2: ‘He also shares feedback and information with employees […] so that we all know what is happening.’	Monitoring operations: Focal manager, #8: ‘When I just started my job, we had all these shelves stuffed with materials. I thought to myself: “this cannot be correct”. So, you start talking to them, observe, check the dust on the shelves. “How often do you use this?” “Well, I don’t know” “But what is the purpose?” This is how I start the conversation.’
	Organizing work activities to improve effectiveness: Boss, #2: ‘He took the responsibility over the problem, which he saw as a challenge and kept the control, but at the same time he kept employees also involved and let them contribute to the solution.’	Resolving issues: Focal manager, #10: ‘We received an email that we would loose 10,000 clients to another department. That news started to spread among employees. I then visited the three teams that would be affected by this, joined their weekly meeting and explained the reasons why.’
		Collecting and disseminating information: Focal manager, #12: ‘Collecting information, then taking a step back to observe what is going on and then sharing this with the team. Trying to transfer the awareness development to other people.’
Relations-oriented	Listening to ideas of employees: Boss, #2: ‘He is always open for suggestions and new ideas for better quality and collaboration.’	Publicly recognising contributions and accomplishments: Focal manager, #9: ‘When you see that the lines are running smoothly, you see the team leaders give their team compliments, when the team works well, and it even leads to results! That is when I say, “See, this is the evidence. We can do it!”.’
	Providing support and encouragement: Subordinate, #4: ‘He gives his people the option to develop different skills and let them become open minded about their work […] he wants you to brainstorm and really wants to talk to his people.’	Publicly recognising contributions and accomplishments: Focal manager, #10: ‘Yesterday, I also sent an email to everyone to tell them “guys, I am proud of the way this went”. Because it truly goes well!’
		Publicly recognising contributions and accomplishments: Focal manager, #12: ‘Last Friday I also, rather spontaneously, broke our company policy by calling a gathering, with the seniors, to share the story I just told you and open a bottle of Champagne. (…) Celebrating successes is key.’
Change-oriented	Encouraging viewing things from different perspectives: Subordinate, #4: ‘He encourages his employees to be more open-minded and see their work from different perspectives’	Building coalition of stakeholders to get change approved: Subordinate, #10: ‘He first discussed it in the management meeting and then also involved us asking: “Guys, this is what we are going to do. What do you think? Does anyone have any ideas? And, eh, are we in?”.’
	Leading by example and modelling exemplary behaviour: Boss, #3: ‘He teaches, leads, and develops the people who work with him. […] He is one of the role models […] He is a very calm guy which gives him success in terms of workplace transformation. (…) He is a very flexible guy, and he led some projects in the past very technical about installations […] During very difficult times he is calm […] and he led the guys of the product development, and he had a very good influence.’	Building coalition of stakeholders to get change approved: Focal manager, #11: ‘I see the seniors and team leaders as a kind of tandem: I strongly involved them, in various sessions. I asked them to think along, like: “How can we improve our customer satisfaction?”.’
	Expressing confidence team can attain objectives: Boss, #3: ‘for lean manufacturing you need to look for the implementation […] and try to change the culture all the time […] So the change management of this kind of manager I think it is the most important one can have. (…) They needed to believe again that change is possible for everybody, also for managers. They need to see that it is possible, the real physical change.’	Leading by example and modelling exemplary behaviour: Boss, #13: ‘She is not afraid to, yes, provide feedback! And I also see her role modelling the things we want to see.’

4.2.1. Lean managers’ task-oriented behaviours

Both Brazilian and Dutch lean managers adopted task-oriented behaviours, but their delivery or approach slightly differed. As underlined by the interviewees, the Brazilian leaders mainly shared information, top-down, and kept control over the change process. For example, as explained by the second lean middle manager’s boss: ‘He took the responsibility over the problem, which he saw as a challenge and kept the control’.

The Dutch managers also shared their observations and information with their teams but, in doing so, aimed to advance employees’ awareness about their task execution by asking questions, as illustrated by the following quote: ‘(…) you start talking to them, observe, check the dust on the shelves. “How often do you use this?” “Well, I don’t know” “But what is the purpose?” This is how I start the conversation’. There was thus quite some resemblance between both samples, although the Dutch leaders seemed more prone to use an inquiry-type style (i.e. asking questions that probe towards an outcome) that might have been perceived as less directly task-oriented by followers.

4.2.2. Lean managers’ relations-oriented behaviours

Relations-oriented behaviours seemed to be considered also particularly important and appreciated by subordinates in both samples. For instance, one of the Brazilian subordinates noted that their manager gave ‘his people the option to develop different skills and let them become open minded about their work […] he wants you to brainstorm and really wants to talk to his people’. The Brazilian managers clearly aimed for collaboration, as illustrated by the quotes in Table 6.

The Dutch managers also developed connections within their team. In particular, many examples were shared by participants illustrating how leaders gave compliments to their teams, and even specifically celebrated successes. One focal manager mentioned that the week before he had ‘broken company policy’ to gather his team and ‘open a bottle of Champagne.’ As he noted: ‘Celebrating successes is key’. In other words, in their own ways, both types of lean managers fostered the relations within their teams.

4.2.3. Lean managers’ change-oriented behaviours

Also, the qualitative data showed that the Brazilian managers were found to advocate change in their departments. For example, a subordinate underlined that their manager ‘encourages their employees to be more open-minded and see their work from different perspectives’. One of the focal Brazilian lean managers recalled a situation where he promoted change and ownership as well as stimulated his staff not to paralyse when things go wrong. By showing the way, he built the solution together with his staff. His senior manager confirmed this: ‘He teaches, leads, and develops the people who work with him. […] He is one of the role models […] He is a very calm guy which gives him success in terms of workplace transformation’. This anecdotal evidence thus seems to match the survey-based findings.

Furthermore, the Dutch managers were strongly focussed on change, but in a more dialogic way. For instance, the managers actively involved their staff members to participate in idea generation, sharing, and implementation. They then tend to ask specific questions like ‘What do you think?’ or ‘How can we improve our customer satisfaction?’ Similar to the Brazilian manager mentioned in the previous paragraph, the female Dutch lean manager was also considered by her boss to be a true change agent to her staff: ‘I also see her role modelling the things we want to see’.

4.2.4. Summary of overlap and differences between the values and behaviours of Brazilian and Dutch lean managers

Table 7 summarises the key findings across the different methods employed, focussing on the overlap and differences between the values and behaviours of Brazilian and Dutch effective lean middle managers. It can be concluded that the cross-cultural differences we expected prior to the mixed-methods data collection, based on our review of the literature, were not so prevalent from the quantitative and qualitative data analysis. Instead, albeit the fact that there were of course some minor differences, there were also many overlaps between both samples. Also, upon comparing the low- versus high-maturity samples, we did not find significant differences in terms of the values while there were significant differences in terms of their behaviours (Tables 5 and 7). This seems to imply that it might be harder to change values systems than it is to change behaviours. Hence, as a next step, we decided to explore the possible links between the lean managers’ behaviours and values.

Table 7. Summary of the findings: Overlap and differences between the values and behaviours of Brazilian and Dutch effective lean middle managers.

	Brazilian	Dutch
Values
Q-sort (Table 4)	Mainly self-transcendence and openness-to-change values	Mainly self-transcendence and openness-to-change values
t/Z tests (Table 5)	Similar degree of self-transcendence and openness-to-change values; significantly lower self-enhancement and higher conservation values compared to Dutch sample	Similar degree of self-transcendence and openness-to-change values; significantly higher self-enhancement and lower conservation values compared to Brazilian sample
Interviews	Respondents spontaneously added ‘macro managing (remain calm and in control)’	Respondents spontaneously added ‘candour (be open and vulnerable)’
Behaviours
Survey (Table 5)	High mean scores on task-, relations-, and change-oriented behaviours	High mean scores on task-, relations-, and change-oriented behaviours
t/Z tests (Table 5)	Significantly higher scores on all three behavioural categories compared to Dutch sample	Significantly lower scores on all three behavioural categories compared to Brazilian sample
Interviews (Table 6)	Respondents gave examples of all three behavioural categories	Respondents gave examples of all three behavioural categories
Impact of lean maturity
Values (Table 5)	No differences	No differences
Behaviours (Table 5)	More task- and change-oriented behaviours among mature lean managers	More task- and change-oriented behaviours among mature lean managers

4.2.5. Additional exploration of the link between lean managers’ behaviours and lean values

Despite Brazilian managers’ stronger inclination towards conservation-oriented values, they showed an eagerness to promote change because it is key within lean. For example, the third Brazilian focal manager believed profoundly in lean: ‘We need to share these values and it is important that you as a manager really accept and really believe in them.’ In this sense, he strongly identified his own values with lean values because ‘if you to do something different from your values I think this is bad for you. You cannot do this and go home and sleep well and think that everything is OK.’ This strong identification with the lean values was also stressed by other managers and could further explain why all three behavioural categories were so frequent compared to the Dutch counterparts.

On top of this strong identification with lean values, the Brazilian managers were aware that those lean values contrasted the more traditional Brazilian cultural values and way of doing (also evidenced by their significantly higher ranking of conservation values, Table 5). For instance, a focal manager noted: ‘Another big improvement opportunity is that even though our products can only be stocked for 20 days, they [contract managers] usually start to negotiate with the clients only at the end of the month. That is Brazilian culture. Because of this we sometimes have too much in stock’. Similarly, a senior manager explained: ‘In Brazil we have a cascade or hierarchy model and control; this is really hard to change in a short amount of time. So, we are trying to change the mindset of those people who will need to involve all the people, who is engaged and committed to the transformation and who understands the purpose’.

Although the Dutch managers did not explicitly reflect on the differences between the lean values and their own cultural values, one of them noted that he had read that leaders in Japan, compared to American leaders like Jack Welch (former director of General Electric), ‘use less I-statements and instead say “we” or “our company” more often. (…) In my opinion this is the essence of the difference between traditional and lean leadership’. This quote shows his conviction that lean might require a different kind of leadership and overcoming cultural differences.

Hence, participants clearly perceived that lean values were different from their cultural values and saw the importance of embracing the lean values even when that could imply changing their own and others’ mindset and values system.

5. Discussion and conclusions

Addressing the scarcity of research on lean manager’s values and behaviours across national cultures, the triangulation of findings of this mixed-methods study point to several interesting similarities between Brazilian and Dutch managers. Overall, both Brazilian and Dutch lean managers’ values constellation were quite similar and fit Schwartz et al.’s (2012) self-transcendence and openness-to-change clusters. Brazilians were more uncertainty-avoidant as illustrated by their higher ranking of conservation values though, while Dutch lean managers seemed to be more individualistic given their higher ranking of self-enhancement values. These outcomes are aligned with what we had expected based on Hofstede et al.’s (2010) framework and Schwartz’s cultural values theory. Although values are widely considered to influence behaviours (Bardi and Schwartz 2003; Torres, Schwartz, and Nascimento 2016), compared to the Dutch sample, the Brazilian lean managers had higher average scores for both task-, relations-, and change-oriented behaviours, thereby showing behaviours related to both transformational and transactional leadership (Behrendt, Matz, and Göritz 2017). The qualitative findings however indicated that both Brazilian and Dutch lean managers showed similar behaviours and that especially the Brazilian leaders strongly identified themselves with lean leadership, noting that lean values were quite different from traditional Brazilian ones. These findings do have important theoretical implications, which leads to propositions for future research, as is delineated next.

In particular, identity theory (Ashforth and Schinoff 2016; Stryker and Burke 2000; Skovgaard-Smith, Soekijad, and Down 2020) can help to explain these results better. This theory posits that how individuals interpret the world is affected by their identities and their constituting values (Ashforth and Schinoff 2016). An ‘identity’ is ‘an actor’s self-definition, how the actor answers the question, “who am I?” or “who are we?”’ (Ashforth and Schinoff 2016, 113). Among the many personal and social identities that an individual may have (e.g. being a female, being a manager, being religious etc.), cultural identity (e.g. being Dutch, being Brazilian etc.) is a very powerful identity since it unconsciously frames how people see the world and what can be deciphered as ‘normal’ and accepted (Amiot et al. 2018). Consequently, when a behaviour is perceived as being outside the usual way of doing things within a community, it stands out as ‘different’.

From our qualitative findings, it was clear that Brazilian managers were perceived by subordinates as constantly displaying lean behaviours since the managers identified greatly with the lean philosophy and values. Furthermore, lean philosophy was perceived as a different way of working compared to how things were usually run in Brazil. For instance, although Brazilians are typically identified as ‘uncertainty avoidant’ (Hofstede et al. 2010), lean may actually require Brazilian managers to embrace and adopt more openness to change values, consequently displaying more change-oriented behaviours. The quest to strive for continuous improvement and embrace a more tolerant approach to ambiguity may thus lead Brazilian managers to emphasise their change orientation far more than what they would normally do. As a result, this emphasis could be perceived by Brazilian subordinates as a culturally a-typical behaviour, leading them to report such behaviours more often compared to Dutch respondents given that change and tolerance towards uncertainty are a more natural part of the Dutch cultural identity. Still, even one of the Dutch lean managers emphasised the merits of lean leadership over ‘traditional’ leadership. These observations can be understood through the Identity Theory principle of sensemaking, which states that ‘individuals are inclined to search for cues about how things work and what is expected’ (Ashforth and Schinoff 2016, 120). In other words, when developing towards being an effective leader of the lean transformation, managers as well as their subordinates tend to look for and emphasise observable cues, such as behaviours, that align with the desired outcome, in this case: displaying lean leadership.

Additionally, in line with the notion of identity salience and centrality of identity theory, individuals often enact the identity that is more important to them in a specific context (Ashforth and Schinoff 2016). Hence, to emphasise even further the contrast with the ‘more traditional’ Brazilian way of working, Brazilian lean managers fully displayed lean values and accompanying behaviours, identifying themselves more with those values than their own cultural ones. This also aligns with the findings by Roccas and Sagiv (2010, 35) who postulated that especially in ‘tight cultures’ in which individuals are more likely to take the social context rather than their own values into account when choosing among behavioural options, individual’s behaviours are ‘more likely to be affected by social norms and expectations’ (here: to act as an effective lean leader). Thus, it could be that identifying themselves as ‘a Brazilian lean leader’ was actually more important for a Brazilian manager’s intra-organisational status than being ‘a Brazilian leader’, which then could have led them to show more desirable lean behaviours, as perceived by their subordinates. Thus, grounded in Identity Theory, we propose:

Proposition 1:

When developing towards a lean leader, managers tend to emphasise more their effective lean leader identity – and thus translate its associated values into behaviours – than other (national-cultural) identities they possess.

Although lean maturity level of the lean managers was not decisive in terms of their values constellation, more experienced lean managers did score higher in terms of task- and change-oriented behaviours. As proposed by the S-curve theory (Netland and Ferdows 2016; Negrão et al. 2020), real lean performance improvements and organisational-cultural change start to appear after two years – the cut-off point to determine lean maturity in this study. In this phase of lean adoption, managers might need to consolidate the many proposed changes in terms of task execution and process improvements. A recent review of the change management literature also confirmed that in order to institutionalise change into the company culture, managers must clearly communicate these changes and at the same time align existing organisational structures and task routines with the newly introduced practices and processes (Stouten, Rousseau, and De Cremer 2018). This might thus explain lean managers’ stronger focus on task- and change-oriented behaviours over time.

At the same time, due to their maturity, lean managers might have started to identify with being a lean leader even more than before. In addition, their bosses and subordinates have had more opportunities to spot a lean manager’s transition towards truly leading the lean transformation. Lean maturity, or time, is thus another factor that is playing a key role in a lean manager’s identification process (Ashforth and Schinoff 2016). Therefore, in line with the S-curve theory, the more experience a manager has with adoption lean, the more they are perceived to embrace lean task- and change-oriented behaviours. This effect may not have been so apparent for relations-oriented behaviours because those are also key in earlier change management stages of lean adoption: employees’ change acceptance is shaped by the quality of the relationships they perceive with their managers (Stouten, Rousseau, and De Cremer 2018). The findings of this study thus only partially confirm the conclusions of Tortorella, Van Dun, and De Almeida (2019) who associated both task- and relations-oriented behaviours with higher lean maturity. Our second proposition is thus:

Proposition 2:

Lean maturity is more influential than cross-cultural differences in terms of explaining effective lean managers’ task- and change-oriented behaviours.

Lastly, this study connected Hofstede et al.’s (2010) cultural dimensions to Schwartz’s (2011) cultural values orientation theory and, ultimately, to Schwartz et al.’s (2012) basic values theory. This way, we aimed to clarify how national-cultural values and individual values are interlinked, while also providing a more fine-grained perspective of how values can impact lean managers’ behaviours. Hofstede’s framework, albeit popular and well-established in Operations Management studies (Erthal and Marques 2018), cannot properly explain individual behavioural divergences (McSweeney 2002). We thus argued that Schwartz’s theories, which encompass and bridge different levels of values, can better serve grasping such differences. As observed in our findings, the combination of Schwartz’s cultural values orientation theory and basic values theory offered an adequate framework to better explain the differences on lean leaders between the Brazilian and Dutch contexts. For example, conservation type values in which Brazilian managers scored higher align with Schwartz’s cultural value of embeddedness (see Table 1). In other words, for Brazilian managers it is important to fit in the new paradigm of being a lean leader, which explains their emphasis on, and awareness of, supportive lean leader behaviours that stand out from what they are normally used to. While for Dutch managers, who scored higher on self-enhancement type values which are aligned with Schwartz’s cultural value of mastery, it is important to feel they master the task at hand, leading them to think they are more obviously a lean leader, even if they are not such effective lean leaders from their boss’ or subordinates’ point of view. Supported by both quantitative and qualitative findings, the proposed theoretical arrangement provides a more solid basis to build insights related to effective lean managers’ values and behaviours, adding a different perspective through Schwartz’s framework to the one derived from the Hofstede’s taxonomy. Hence, we advance the following proposition:

Proposition 3:

The combination of Schwartz’s cultural values orientation theory and basic values theory can be more helpful in explaining effective lean managers’ values and behaviours than Hofstede’s taxonomy.

This study thus advances current knowledge on effective lean adoption across cultures (Danese, Manfè, and Romano 2018). Through a robust mix-method design, we underline how, whilst lean values seem to be more universally shared and recognised, effective lean leader behaviours can be perceived differently depending on the importance that individuals place on their cultural identity versus their lean leader identity, as well as their lean maturity.

6. Practical implications

Most organisations that have started to adopt lean have understood the importance of involving organisational leaders at different hierarchical ranks to display exemplary lean behaviour. This study shows that although the personal work values of leaders are of importance in influencing their display of such lean behaviour, cross-cultural differences in those values and behaviours should not be exaggerated. Instead, multinational organisations with plants or locations across different countries need to consider building a strong ‘lean identity’ within their firms and as part of their lean leadership training programs. Based on our study, it seems that such a lean identity and the status it might potentially bring, especially if higher-level leaders also emphasise the importance of being a lean leader, can overcome any cross-cultural differences in developing lean leadership. Future action-learning type lean leader training intervention studies, which facilitates participants in their sensemaking of lean leadership, could be of practical value as well to accelerate lean leadership development. Secondly, given the lean maturity effects we found herein, (HR) managers are advised to differentiate their training efforts between novices and more advanced lean leaders.

7. Limitations and future research

As all research, this study has also its limitations. While our relatively small sample fits the exploratory goal of the study and is customary for qualitative middle-range studies (Soltani et al. 2014), the results might not be generalisable. Thus, building on the rich insights uncovered by this mixed-methods study, future quantitative research could consider a bigger sample size at all (multiple) levels, to enable regression analyses and structural equation modelling and, therefore, obtain statistically stronger results. Secondly, Brazil’s regional cultural differences must be taken into account; the respondents all came from the South of Brazil which corresponds with a higher score on individualism, long-term orientation, and indulgence and reduced uncertainty avoidance (Hofstede et al. 2010). Hence, future studies might benefit from widening the geographical scope to really capture the diversified, regional culture in Brazil. Thirdly, the cross-sectional character of the study is a constraint given that lean adoption may require different leadership behaviours over time (Tortorella, Van Dun, and De Almeida 2019). Future, longitudinal quantitative research is thus advised which should also consider self-other agreement and replicate Van Dun et al.’s (2017) video-analyses. Lastly, while we had unique access to a Brazilian and Dutch dataset, we also relied on a convenience sampling approach. Hence, future research should select and compare countries more systematically, like Japan, based on national cultural characteristics (Erthal and Marques 2018). Especially given the ongoing expansion of lean adoption across the global supply chain (Powell and Coughlan 2020; Bortolotti et al. 2016), such studies must test the universal applicability of lean leadership as we define it today.

Ethics statement

This research has been approved by the ethics committee of the Faculty of Behavioural, Management and Social Sciences, University of Twente, the Netherlands (approval number: BCE18639).

Acknowledgements

The authors thank all participants and Laura van Leeuwen, MSc., for her help in collecting and analysing the data.

Disclosure statement

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

Additional information

Notes on contributors

Desirée H. van Dun

Dr Desirée H. van Dun is an Assistant Professor in Organisational Behaviour, Change Management & Consultancy at the University of Twente, the Netherlands, and Visiting Researcher at Cardiff Business School, Cardiff University, United Kingdom. Besides having 10 years of lean management consulting experience, she obtained a cum laude Ph.D. degree in Organisational Behaviour and Operations Management; the thesis won various local and international awards. Her research has appeared in journals such as the International Journal of Operations & Production Management, European Management Journal, Journal of Manufacturing Technology Management, International Review of Industrial and Organisational Psychology, Creativity and Innovation Management, and International Journal of Qualitative Methods. She serves as an Associate Editor for Creativity and Innovation Management and reviews for various other international journals. Van Dun is currently studying effective human behavioural dynamics for digital and green organisational transformation, including leading lean (and green), agile, and Industry 4.0 technology adoption. https://people.utwente.nl/d.h.vandun

Guilherme L. Tortorella

Dr Guilherme L. Tortorella is a Professor of Industrial Engineering at the University of Melbourne, Australia. With more than 200 publications in peer-reviewed journals, he is the Editor-in-Chief of the Journal of Lean Systems, and Associate Editor of Operations Management Research, International Journal of Lean Six Sigma, and Production Journal. He has worked for 12 years as a Manufacturing and Continuous Improvement Manager in the automotive industry, with experiences in Brazil, Uruguay, Mexico, the UK, and the USA. In the last 8 years, he has dedicated his career to academia, focussing his research on topics, such as operational excellence and industry digital transformation.

Lara Carminati

Dr Lara Carminati is an Assistant Professor in Organisational Behaviour, Change Management and Consultancy at the University of Twente, the Netherlands. She obtained her PhD in Management, Organisational Behaviour, at the Surrey Business School, United Kingdom. Her research has been published in peer-reviewed international outlets such as Frontiers in Psychology, Current Psychology, Qualitative Health Research, Health Policy, and Healthcare Management Review. Carminati reviews for various international journals and is currently researching the interplay between professional identity and AI in different organisational settings.

Notes

1 The item ‘Clarifies quality standards for task performance’ was removed as it had to be omitted from the Dutch dataset.

Appendix A.

Interview guide

Introduction

Briefly explain the goal of the research. Data will be treated confidentially and will be anonymized. There are no wrong answers.

1.	Can I audio record this conversation?
2.	Could you please briefly tell me about your position within your organisation? (number of subordinates, responsibilities, tenure)

Lean organisation

• 3. How would you describe a lean organisation?

• 4. How do you view the role of a manager in a lean organisation?

• 5. Do you see your own organisation as a lean organisation? Why?

• 6. When did your organisation start adopting lean principles?

• 7. What do you do differently since your organisation started transitioning to lean, compared to previously?

Leadership

This study focuses on effective middle managers in lean organisations.

• 8. In your opinion, why are you/is he or she an effective manager?

1. What do you/does he or she do specifically?

2. Is this distinctive of lean middle managers (as opposed to other, non-lean organisations)?

• 9. Which values do you/does he or she consider important? Values are defined as ‘the frame of reference which someone carries with them in all situations as a guideline for his/her actions.’

  1. Is this distinctive of lean middle managers (as opposed to other, non-lean organisations)?

Critical incident technique

• 10. Please imagine a recent and specific moment when you were/the concerning lean middle manager was in contact with your/his/her immediate employees and showed very successful leadership. Can you describe this situation and tell us exactly what happened?

• 11. When did this situation occur?

• 12. What caused this situation?

• 13. What did you/the concerning Lean middle manager do at the exact moment? What where his/her behaviours? Behaviours can be described as 'what leaders and managers do during their work as leaders’.

  1. Probe to pinpoint specific behaviours.

  2. Is this distinctive of lean middle managers (as opposed to other, non-lean organisations)?

• 14. Which personal values where dominant in this situation according to you?

  1. Which specific values were shown by the Lean middle manager in this story?

  2. Probe to pinpoint specific values.

  3. Is this distinctive of lean middle managers (as opposed to other, non-lean organisations)?

Q-sort of values

• 15. Which values are, in your opinion, most applicable to you/this highly effective middle manager? Please distribute the cards in the normal distribution (Table A1):

Available values cards: See Table 1, column ‘lean values’.

After the Q-sort:

• 16. Would you like to explain something further in terms of your distribution?

• 17. Why did you make these choices? Discuss the most and least important values.

• 18. Did you experience any difficulties during the card distribution?

Conclusion

• 19. Would you like to add anything to the conversation?

Thank you for your participation!

Appendix B.

Survey scales

Lean tool adoption (Fullerton, Kennedy, and Widener 2014)

5-point scale with as scale labels ‘not at all’, ‘a little’, ‘to some extent’, ‘considerably’, and ‘a great deal’.

To what extent has your facility implemented the following:

1. Standardisation

2. Lean Cells

3. Reduced setup times

4. Kanban system

5. One-piece flow

6. Reduced lot sizes

7. Reduced buffer inventories

8. 5S

9. Kaizen (continuous improvement)

Department-level operational performance improvement (Fullerton, Kennedy, and Widener 2014)

5-point scale with as scale labels: ‘significantly worsened’, ‘moderately worsened’, ‘little or no change’, ‘moderately improved’, and ‘significantly improved’.

Indicate how your facility’s operations have changed over the last three years:

1. Scrap and rework

2. Machine setup times

3. Queue and move times

4. Machine downtime

5. Lot sizes

6. Cycle time

Lean middle managerial behaviours (Yukl 2012)

7-point scale ranged from ‘never’, ‘almost never’, ‘sometimes’, ‘regularly’, ‘more than regularly’, ‘often’, to ‘always’.

Task-oriented behaviours:

My direct supervisor/I…

1. Organises work activities to improve effectiveness

2. Develops milestones and action plans for a project

3. Determines what resources are needed to do a project

4. Clarifies role expectations for task performance

5. Facilitates collection and dissemination of information

6. Actively monitors operations and performance

7. Resolves immediate questions or problems from team

8. Clarifies quality standards for task performance (removed)

Relations-oriented behaviours:

My direct supervisor/I…

1. Actively listens attentively to a person’s concerns

2. Actively provides support and encouragement

3. Socialises with team beyond work to build relationships

4. Publicly recognises contributions and accomplishments

5. Provides individual role and/or behaviour coaching

6. Consults with members on decisions affecting them

7. Helps team members (as a group) resolve conflicts

8. Keeps team (group) informed about upcoming actions

Change-oriented behaviours:

My direct supervisor/I…

1. Leads by example and models exemplary behaviour

2. Encourages viewing things from multiple perspectives

3. Expresses confidence team can attain objectives

4. Creates sense of urgency, promotes change

5. Studies other projects to get ideas for improvements

6. Envisions exciting new possibilities for the organisation

7. Develops strategies linked to organisation’s vision

8. Builds coalition of stakeholders to get change approved

9. Creates task force to guide implementation of change

10. Suggests symbolic changes that affect the work

11. Empowers members to implement new strategies

12. Announces, celebrates progress supporting changes

13. Encourages/facilitates learning by team members

14. Experiments with new approaches