INTRODUCTION
In a globalized world, the question of whether leadership has universal effects becomes crucial. This study focuses on transformational leadership, which consistently has been demonstrated to be associated with positive employee effects (for a meta-analysis see Judge & Piccolo, Reference Judge and Piccolo2004) and performance (Lowe, Kroeck, & Sivasubramaniam, Reference Lowe, Kroeck and Sivasubramaniam1996; Bass, Reference Bass1999; Avolio, Reichard, Hannah, Walumbwa, & Chan, Reference Avolio, Reichard, Hannah, Walumbwa and Chan2009). Building on ‘the happy-productive employee thesis’ (Wright & Staw, Reference Wright and Staw1999), which suggests that employee well-being and performance are associated, we test the universality claim of transformational leadership. The universality claim assumes that the defining transformational leadership behaviours are universally endorsed as outstanding (Den Hartog, House, Hanges, Ruiz-Quintanilla, & Dorfman, Reference Den Hartog, House, Hanges, Ruiz-Quintanilla and Dorfman1999) and have robust effects (Bass, Reference Bass1997). Inspired by Lonner (Reference Lonner1980), Bass (Reference Bass1997) describes five types of universals: the simple (when a phenomenon is constant across the world), the variform (when a general principle is constant, but may be enacted differently across culture), the functional (when relations are constant and invariant across cultures), the systematic (when a sequence or organization of behaviours are constant across cultures), and the variform functional type (when a relationship is constant but the magnitude may vary across cultures). In contrast to many universality studies, we do not investigate the constancy, enactment, and effects of transformational leadership across countries and cultures. Instead, we hold the national context constant and investigate a participant sample varying on national and cultural backgrounds. Contrasting the universalities to the contingency theory of leadership, we test the universality claim of transformational leadership with regard to both its perception and its effects relative to three outcomes: job satisfaction, self-rated health, and well-being.
Over the past decade, transformational leadership has dominated leadership research (Walumbwa & Wernsing, Reference Walumbwa and Wernsing2012; Dinh, Lord, Gardner, Meuser, Liden, & Hu, Reference Dinh, Lord, Gardner, Meuser, Liden and Hu2014), and it has had a massive impact on both leadership science and practice (Shamir, House, & Arthur, Reference Shamir, House and Arthur1993; Conger, Kanungo, & Menon, Reference Conger, Kanungo and Menon2000; Antonakis, Reference Antonakis, Avolio and Sivasubramaniam2012). The transformational leader works to ensure alignment between the needs of followers and company goals, thereby aiming for performance beyond expectation. In the transformational leadership literature, performance beyond expectation refers to the followers being influenced to transcend their self-interest for the greater good of the organization, as opposed to transactional leadership, which is limited to inducing basic exchanges with followers (Bass, 1985). Indeed, transformational leadership has been associated with employees’ image of the ‘ideal’ leader (Bass & Avolio, Reference Bass and Avolio1989). This notion of transformational leadership as being close to a prototypical perception of the ‘ideal’ leader is coherent with the view that transformational leadership is universal – thus transcending cultural boundaries (Bass, Reference Bass1997, Reference Bass1999) and evoking similar positive effects across countries (Bass & Bass, Reference Bass and Bass2009; Jung, Yammarino, & Lee, Reference Jung and Yammarino2009). The universalistic notion of a unitary construct of transformational leadership has, nonetheless, been debated (Bass & Bass, Reference Bass and Bass2009) and heavily criticized (Van Knippenberg & Sitkin, Reference Van Knippenberg and Sitkin2013), reflecting a more general, scientific call for contextualized perspectives in organizational behaviour research (Klimoski, Reference Klimoski2013; Härtel & O’Connor, Reference Härtel and O’Connor2014).
Current trends of globalization and diversity at work further underline the need for investigation of the proposed universality of transformational leadership (Kumar & Chhokar, Reference Kumar and Chhokar2012). These trends pose several challenges for managers (Dickson, Castaño, Magomaeva, & Den Hartog, Reference Dickson, Castaño, Magomaeva and Den Hartog2012), for instance, that of knowing how individuals with different national and cultural backgrounds perceive managerial actions. Many companies employ individuals from different national and cultural backgrounds and the composition of employee groups is both changing and dynamic. Investigating the universality of transformational leadership becomes highly important for companies with such composite employee groups. Cultural sensitivity, interpreted as being open to a variety of different environments and cultures, is indeed a competence, which global leaders perceive to be important to acquire (Terrell & Rosenbusch, Reference Terrell and Rosenbusch2013) and subsequently apply. With the purpose of facilitating global leadership skills, the challenges of globalization and diversity call for an increased understanding of how leadership is perceived and fosters outcomes across employee nationality and culture – and consequently to which extent transformational leadership is universal or contingent upon the national and cultural backgrounds of employees.
The present study contributes with an empirical investigation of the perceptions and outcomes of transformational leadership across native and immigrant employees, thereby contributing to the exploration of the universality or cultural contingency of transformational leadership. There are different ways to conceptualize the construct of culture. In the current study, we adopt a common approach, which is to use proxies (country of birth and citizenship) in order to characterize and operationalize culture (for a similar approach see Soares, Farhangmehr, & Shoham, Reference Soares, Farhangmehr and Shoham2007). Our research question is the following: Is the perception and effects of transformational leadership universal across native and immigrant employee groups within a shared national context (Denmark)? As opposed to other cross-national studies, our study thus contributes to the extant leadership literature in the following five ways: (1) it pushes for a theoretical development of transformational leadership theory towards embracing the conditions and potential limits of its universality, (2) it offers a different approach to the testing of transformational leadership universality by performing a within-country study, thus addressing a highly relevant condition for companies with composite employee groups, (3) it identifies a need for a more explicit integration of the leadership perspective with aspects and consequences of globalization, (4) it is performed within a natural organizational context, thus contrasting with the experimental student sample settings commonly applied in other studies (Jung & Yammarino, Reference Jung, Bass and Sosik2001; Ergeneli, Gohar, & Temirbekova, Reference Ergeneli, Gohar and Temirbekova2007; Jung, Yammarino, & Lee, Reference Jung and Yammarino2009), and (5) it applies a repeated measures study design consequently enabling the investigation of both causal relations and change over time, thus responding to calls for follow-up studies within both leadership (Kuoppala, Lamminpää, Liira, & Vainio, Reference Kuoppala, Lamminpää, Liira and Vainio2008) and immigration research (Olesen et al., Reference Olesen, Carneiro, Jørgensen, Rugulies, Rasmussen, Søgaard and Flyvholm2012).
This paper is structured in the following way: first, we outline the theoretical background of our hypotheses development, followed by a description of the methods by which we test our hypotheses. We then present and discuss our results and finally propose our conclusions and recommendations for future research and practise.
THEORY AND HYPOTHESIS DEVELOPMENT
Based on a review of the leadership literature more than 40 years ago, Stogdill (Reference Stogdill1974) concluded that there are almost as many definitions of leadership as there are persons, who have attempted to define the concept. Leadership has for instance been defined in terms of both traits, behaviours, and role relationships (Day & Antonakis, Reference Day and Antonakis2013). Most definitions, however, reflect the assumption that leadership involves a process whereby one person exerts intentional influence over others with the purpose of guiding, structuring, and facilitating activities and relationships in organizations (Yukl, Reference Yukl2002). Transformational and transactional leadership are two important examples of leadership styles that focus on intentionally influencing followers.
The roots of transformational leadership are often attributed to the works of political scientist James Burns (Reference Burns1978). Burns (Reference Burns1978) used the term ‘transforming leadership’ to describe leaders who seek to satisfy the higher order needs of their employees in order to engage them. He argued that transforming leadership occurs when leaders and followers raise one another to higher levels of motivation and morality. This should be seen in contrast to transacting leadership, which motivates followers by appealing to their self-interest through the exchange of pay and status for work effort.
Transformational leadership remains interesting to study for several reasons: First, it belongs to one of the most dominant leadership paradigms, often referred to as The New Leadership School (Day & Antonakis, Reference Day and Antonakis2012, Reference Day and Antonakis2013; Dinh et al., Reference Dinh, Lord, Gardner, Meuser, Liden and Hu2014). Second, research has consistently found transformational leadership to be positively associated with employee performance (Lowe et al., 1996; Bass, Reference Bass1999; Judge & Piccolo, Reference Judge and Piccolo2004; Avolio et al., Reference Avolio, Reichard, Hannah, Walumbwa and Chan2009). Third, transformational leadership has been associated with employees’ job satisfaction (Braun, Peus, Weisweiler, & Frey, Reference Braun, Peus, Weisweiler and Frey2013), well-being (Arnold, Turner, Barling, Kelloway, & Mckee, Reference Arnold, Turner, Barling, Kelloway and Mckee2007), and potentially also health – even though research on this relation still remains inconclusive (Nyberg, Bernin, & Theorell, Reference Nyberg, Bernin and Theorell2005). Fourth, consistent claims to universal transformational leadership effects have been made (Bass, Reference Bass1997). Inspired by the massive attention, impact, reputation, and universality claim of transformational leadership, the current study investigates whether transformational leadership is indeed universal, or if it is contingent upon nationality and cultural background. The identification of boundary conditions has been targeted as crucial in leadership research (Antonakis, Avolio, & Sivasubramaniam, Reference Antonakis2003). Certainly, in the evolution of theories, transformational leadership has now reached the point at which critical review and identification of such boundary conditions are relevant.
Contingent and universal leadership paradigms
In the leadership literature, two general paradigms of leadership have emerged: the contingency paradigm and the universalistic paradigm. Both paradigms are represented within transformational leadership research.
Within the contingency paradigm, leadership is understood to be dependent on situations (Hersey & Blanchard, Reference Hersey and Blanchard1969), cognitive states and traits (Fiedler, Reference Fiedler1986), or decision-making processes (Vroom & Jago, Reference Vroom and Jago2007). Further, a cultural contingency model of leadership has been proposed, focussing on cross-cultural leadership and organizational change (Muczyk & Holt, Reference Muczyk and Holt2008). Within the cultural strand of transformational leadership research, studies have investigated whether certain cultures elicit higher degrees of transformational leadership (Ergeneli, Gohar, & Temirbekova, Reference Ergeneli, Gohar and Temirbekova2007) and whether individual elements of transformational leadership are weighed differently across cultures and countries (Ardichvili & Gasparishvili, Reference Ardichvili and Gasparishvili2001). However, the cultural strand has mainly focussed on racial and national differences (Shelton, Reference Shelton2007). Jung and Yammarino (Reference Jung, Bass and Sosik2001) found, for example, that transformational leadership has stronger effects on group potency among Asian Americans, while stronger effects on self-efficacy among Caucasian Americans. With Hofstede’s dimensions of national culture as point of departure, the GLOBE project found differences in preferred leadership across national clusters (see e.g., House, Javidan, Hanges, & Dorfman, Reference House, Javidan, Hanges and Dorfman2002; Wanasika, Howell, Littrell, & Dorfman, Reference Wanasika, Howell, Littrell and Dorfman2011). More recently, Engelen, Schmidt, Strenger, and Brettel (Reference Engelen, Schmidt, Strenger and Brettel2014) found that national culture affects the strength of the relation between top management’s transformational leadership behaviours and innovation orientation. Further, Dickson et al. (Reference Dickson, Castaño, Magomaeva and Den Hartog2012) established that culture predictably matters for the emergence, development, and selection of leaders, while Leong and Fischer (Reference Leong and Fischer2011) found that leadership behaviours covary with cultural values, such that managers in more egalitarian cultures reveal more transformational leadership behaviours.
The universalistic approach, also referred to as the ‘one-size-fits-all’ approach (Shamir, Reference Shamir2012), suggests that there are universal leadership models that hold generic leadership functions and behaviours (see for instance, Hamlin, Reference Hamlin2007). Bass (Reference Bass1997) describes transformational leadership to be ‘a systematic behavioural universal’, underpinned by an underlying theory and model, which is systematically universal. Only slight contingencies and variation may occur because ‘the same concepts may contain specific thought processes, beliefs, implicit understandings, or behaviours in one culture but not another’ (Bass, Reference Bass1997: 132). In support of this, Walumbwa, Orwa, Wang, and Lawler (Reference Walumbwa, Orwa, Wang and Lawler2007) find that, although transformational leadership explained more variance in their American sample, transformational leadership had similar impacts on employees’ commitment and satisfaction across the United States and Kenya. An example of such cultural variance within the universalistic approach is proposed by Jung, Bass, and Sosik (Reference Jung and Avolio1995) suggesting that, despite being universal, transformational leadership is more likely to appear in collectivistic rather than individualistic cultures. Walumbwa and Lawler (Reference Walumbwa and Lawler2003) found that collectivism moderated the impact of transformational leadership on employees’ commitment and satisfaction in China, India, and Kenya. Although both quantitative (Den Hartog et al., Reference Den Hartog, House, Hanges, Ruiz-Quintanilla and Dorfman1999) and qualitative studies (Karakitapoglu-Aygun & Gumusluoglu, Reference Karakitapoglu-Aygun and Gumusluoglu2013) have supported the notion of universal transformational leadership, the universalistic notion of a unitary construct of transformational leadership has recently been severely criticized by Van Knippenberg and Sitkin (Reference Van Knippenberg and Sitkin2013). Among other things, they suggest to investigate the potential variation of transformational leadership by its conceptual dimensions, levels (e.g., individual and group), and outcomes.
In summary, the extant literature still represents contingency and universality as two paradigms with fundamental differences in the role and importance they ascribe to culture and nationality in leadership. Following the recent developments in transformational leadership research, the current study challenges and investigates this divide.
Transformational leadership falls within one of the identified culturally endorsed implicit leadership theories of the GLOBE study (Javidan, Dorfman, de Luque, & House, Reference Javidan, Dorfman, de Luque and House2006), that is the charismatic, value-based leadership. However, the method of comparing a phenomenon (transformational leadership) across countries – as is done in the GLOBE project (House et al., Reference House, Javidan, Hanges and Dorfman2002) – may not fully answer the question of universality. In part because nationality may not be the only relevant contextual factor for employees’ perception of leadership (Jepson, Reference Jepson2009) – in part because employment contexts may vary significantly across nations, thereby potentially confounding the results. Our design allows us to investigate whether transformational leadership differences, related to the country of birth and citizenship of employees, can be established beyond the probable assimilation occurring for employees within a shared national work context. According to the Hofstede cultural dimensions, the Danish culture is characterized by being highly individualistic and feminine, with a very low power distance and uncertainty avoidance (Hofstede, Reference Hofstede1984). This cultural orientation suggests that culturally fitting leadership in Denmark should be coaching and allowing for employee autonomy. Furthermore, as there is a low need for work life structure and predictability, effective management should be supportive and involving (https://geert-hofstede.com/denmark.html). Still, there are different understandings of the relation between cultural dimensions and leadership. Walumbwa, Lawler, and Avolio (Reference Walumbwa, Lawler and Avolio2007) suggest that collectivistic cultures respond more positively to transformational leadership than individualistic cultures, the latter being more responsive to transactional leadership. To clarify relations between national and cultural background and the perception and effect of leadership more research is needed in this area.
Hypotheses
The hypotheses, we develop in this study, refer to both the universality in perception and outcomes of transformational leadership. With regard to employees’ perception of leadership, research does not draw a conclusive picture. In the defence of universalism, Den Hartog et al. (Reference Den Hartog, House, Hanges, Ruiz-Quintanilla and Dorfman1999) find that several transformational leadership attributes are universally attributed to outstanding leadership. Lankau and Chung (Reference Lankau and Chung2009) also find a high level of cross-cultural consistency in the perception of what constitutes a successful manager, a finding that would argue for cultural convergence towards Western leadership theories and universality. Contrarily, Yan and Hunt (Reference Yan and Hunt2005) argue that cultural differences do exert an influence on employees’ perception of leadership and leadership effectiveness, and Popper and Druyan (Reference Popper and Druyan2001) find significant differences in leadership perceptions across nationalities. Gerstner and Day (Reference Gerstner and Day1994) further find that culture and origin is associated with the degree (most, moderately, or least) to which traits are considered to be characteristic of business leaders. Thus, cultural groups may potentially vary in their understanding of what constitutes effective leadership. Differences in perception may be related to different employee interpretations, as was shown with regard to the impact of justice perceptions in the study by Avery, Tonidandel, Volpone, and Raghuram (Reference Avery, Tonidandel, Volpone and Raghuram2010) and/or to leaders’ different treatment of employees, as shown in the applicant accent study by Hosoda and Stone-Romero (Reference Hosoda and Stone-Romero2010). At the cognitive psychological level, the relation between culture and leadership is by Hanges, Lord, and Dickson (Reference Hanges, Lord and Dickson2000) explained by the development and impact of values and self-constructs developed throughout life. Cultural meaning systems are shared between individuals and are composed by beliefs, norms, and values. Hanges, Lord, and Dickson (Reference Hanges, Lord and Dickson2000) suggest that these cultural meaning systems are well established in individuals and activate or inhibit which traits are associated with leadership. Notwithstanding a potential convergence towards Western leadership theories, we propose that the universality theory will not find support and employees with different national origins (native vs. immigrant) will perceive transformational leadership differently. We therefore hypothesize that:
Hypothesis 1: Native and immigrant employees will differ in their perception of transformational leadership.
With reference to the effects on employee outcomes of leadership in general and transformational leadership in particular, most research has been performed within a Western context (Steers, Sanchez-Runde, & Nardon, Reference Steers, Sanchez-Runde and Nardon2012). Nonetheless, some positive associations between transformational leadership and work outcomes (e.g., job performance, affective organizational commitment, turn-over intentions) have also been found in a non-Western context (Miao, Newman, & Lamb, Reference Miao, Newman and Lamb2012). However, these findings do not respond to the question of whether the effects of transformational leadership remain universal or not. In an experiment, Jung and Avolio (Reference Jung, Yammarino and Lee1999) tested the effects of leadership contingent upon cultural orientations and found that students with individualistic orientations (Caucasians) would produce more with a transactional leader than with a transformational and that students with collectivistic orientations (Asian) would produce more with a transformational than a transactional leader. From the perspective of outcomes, we therefore propose that there will be an impact of national and cultural background (native vs. immigrant) such that the universality theory will not find support with regard to the effects of transformational leadership on individual outcomes. We suggest the following:
Hypothesis 2: Transformational leadership affects individual outcomes (job satisfaction, self-rated health, and well-being) differently for employees with native Danish and immigrant backgrounds.
We propose this hypothesis to be confirmed in analytical models both with and without control for the baseline outcome levels. When controlling for baseline levels, the outcome reflects the change over time in job satisfaction, self-rated health, and well-being that is affected by transformational leadership.
METHODS
Procedure and participants
Data stem from a large cohort study investigating work environment and self-rated health of employees in the Danish public sector, specifically within the elder care. Data were collected from the fall of 2006 to the spring of 2007 (T1) and from the fall of 2008 to the spring of 2009 (T2). At both times, questionnaires were administered in the 35 Danish municipalities that agreed to participate in the study (out of a total of 98 invited municipalities). All measures were assessed at both time points. At T1, 15,697 employees received the questionnaire and a total of 10,065 questionnaires were completed and returned (response rate 64%). At T2, 13,945 employees received the questionnaire and a total of 8,437 were completed and returned (response rate 63%). The questionnaire was answered by 5,206 individuals at both T1 and T2. As we focus on employee perceptions and outcomes of transformational leadership, we excluded 552 managers from the study sample. Due to nonavailable information on country of birth and citizenship, we furthermore excluded 1,707 respondents. Thus excluding dropouts and newcomers, the final sample for analysis consists of 2,947 respondents (2,836 native Danes and 111 immigrants), in that manner representing a larger sample than applied in prior studies (Popper & Druyan, Reference Osborn, Hunt and Jauch2001).
For the native group, the mean age was 48 with an average of 11 years of employment. In total, 98% were women and all participating managers were female. Of the 111 immigrant participants, 70.8% were born in Europe, 0.9% in North America, 3.8% in South and Central America, 7.5% in Africa, 11.3% in Asia, and 5.7% in the Middle East. The mean age within the immigrant group was 46 with an average of 7 years of employment at the current workplace. In total, 92% were women. The female predominance of the sample is a consequence of the elder care profession targeted in the study. The educational programmes for elder care professionals vary in duration (7 weeks, 12 months, 14 months, 36 months, and 42 months). Consequently, in the care of the elderly, the employees have different educational backgrounds and tasks. The distribution of educational level was similar for the Danish and immigrant groups.
Measures
Transformational leadership
The independent variable transformational leadership was measured by a 7-item approximated Danish translation of the 7-item Global Transformational Leadership scale (Carless, Wearing, & Mann, Reference Carless, Wearing and Mann2000). The measure targets the leadership behaviours of the immediate manager. An example item is: ‘My manager communicates a clear and positive vision of the future’. Response categories were on a 5-point Likert scale ranging from ‘to a very large extent’ to ‘to a very small extent’. Cronbach’s α was 0.92 at T1 and 0.93 at T2. The Global Transformational Leadership scale 7-component measure of transformational leadership has been found to have high convergent validity with the lengthier Multifactor Leadership Questionnaire (MLQ) and the Leadership Practices Inventory (LPI) (Carless, Wearing, & Mann, Reference Carless, Wearing and Mann2000).
Job satisfaction
Job satisfaction was measured using a 2-item measure focussing on client-related work. An example item is: ‘Do you find that you become happier by working with clients?’ Response categories were on a 5-point scale ranging from ‘to a very large extent’ to ‘to a very small extent’ Cronbach’s α was 0.83 at T1 and 0.82 at T2.
General self-rated health
General self-rated health was measured using a single item from the SF-36: ‘In general, would you say your health is …’ (Bjørner et al., 2005). Response categories were on a 5-point Likert scale ranging from ‘excellent’ to ‘poor’. This single-item measure has been found to perform as well as the SF-36 multi-item self-reported health (DeSalvo, Fan, McDonell, & Fihn, Reference DeSalvo, Fan, McDonell and Fihn2005).
Well-being
Well-being was measured using the WHO-5 well-being scale (Bech, Olsen, Kjoller, & Rasmussen, Reference Bech, Olsen, Kjoller and Rasmussen2003). This scale has been widely used and validated in several languages across the world (Topp, Østergaard, Søndergaard, & Bech, Reference Topp, Østergaard, Søndergaard and Bech2015). An example item is: ‘How much of the time in the last two weeks did you feel active and energetic?’ Response categories were on a 6-point scale ranging from ‘all the time’ to ‘at no time’. Cronbach’s α was 0.86 at T1 and 0.88 at T2. For reasons of clarity and comparison, we transformed response values for all measures into a 0–100 scale (100 indicating the highest possible rating of each item).
Immigrant status
Information on immigrant status (records of participants’ country of birth and citizenship) was obtained and imported to our data set from the Danish Central Office of Civil Registration. A participant was categorized as Dane, if at least one of the parents was a Danish citizen. A participant was categorized as immigrant, if none of the parents were Danish citizens.
Data analysis
To investigate the perception of transformational leadership, we performed the following three procedures: (1) independent sample T-tests to investigate potential differences in mean levels of transformational leadership (scale and item levels), job satisfaction, well-being, and self-rated health between the native Danish and the immigrant group; (2) principal component analyses with Varimax rotation to examine potential differences (on the entire sample, the native Danish group, and the immigrant group) in pattern and magnitude of item loadings on transformational leadership; and (3) intercorrelations, means, and standard deviations for the seven transformational leadership items separately for the native Danish and the immigrant groups in order to study the pattern of association between items for the two groups.
With the purpose of investigating the outcomes of transformational leadership across time, we performed a general linear model controlled for age and gender. We analysed two models: a Model 1 (without control for baseline outcomes) and a Model 2 (with control for baseline outcomes). All analyses were performed in IBM Statistics SPSS 20. The study followed the guidelines of and was approved by The Danish Data Protection Agency.
RESULTS
Table 1 displays the intercorrelations, means, and standard deviations for the independent, dependent, and control variables of the study population. The results show significant correlations between transformational leadership, job satisfaction, self-rated health, and well-being.
Intercorrelations, means and SD
Note. n=2,562–2,924.
*p<.05; **p<.01.
Perceptions of transformational leadership
Concerning the perceptions of transformational leadership, the independent sample T-test reveals no significant difference between the native Danish and the immigrant group in the reported mean level of transformational leadership as a general construct. In addition, a nonsignificant difference is found for each of the seven transformational leadership items (see Table 2). Furthermore, the native Danish group reports significantly better self-rated health and well-being than the immigrant group, while there is no significant difference in the reported mean level of job satisfaction.
Independent samples test for native Danes and immigrants
Note. SD appear in parentheses beside means.
**p<.01.
The principal component analysis (see Table 3) demonstrates that all of the seven transformational leadership items load highly (0.81−0.91) on one factor with a Cronbach’s α of 0.95. When performing the analysis for the two groups separately, we find analogous factor structures and loadings. The native Danish group has loadings from 0.81 to 0.91 with a Cronbach’s α of 0.95, while the immigrant group has loadings from 0.81 to 0.90 with a Cronbach’s α of 0.94. Table 4 displays the intercorrelations for the seven transformational leadership items for native Danes and immigrants separately. We find that item correlations are similar for the two groups. Hence, we find no evidence that native and immigrant employees differ in their perception of transformational leadership. Overall, Hypothesis 1 was thus not supported by our results.
Component matrix
Note. n=2,836/Danes and 111/immigrants. Varimax rotation.
Intercorrelations, means and SD for scores on the seven transformational leadership items for Danes and immigrants
Note. Intercorrelations between transformational leadership items for Danes (n=2,746–2,782) presented below the diagonal and immigrants (n=107–109) presented above the diagonal.
**p<.01.
Outcomes of transformational leadership
Relevant to the outcomes of transformational leadership, the analyses in Model 1 show for native Danes positive, significant relations between T1 transformational leadership and T2 job satisfaction, self-rated health, and well-being. In Model 2 (with baseline control), the relations of T1 transformational leadership with T2 job satisfaction and well-being maintain their significance, while the relation with self-rated health ceases to be significant (see Table 5). For immigrants, Model 1 demonstrates that T1 transformational leadership has a positive, significant relation with T2 well-being, and a trend towards a positive, significant relation with T2 job satisfaction. The model also shows a nonsignificant relation with T2 self-rated health. In Model 2 (with baseline control), all relations cease to be significant (Table 6).
General linear model: Danes
General linear model: immigrants
At item level, some items contribute more than others to the three outcomes in the native Danish group (see Table 7). While all seven transformational leadership items relate significantly to job satisfaction, only the items of vision, staff development, and empowerment relate significantly to well-being. Despite the overall nonsignificant relation between transformational leadership and self-rated health, the one item of vision relates significantly to the self-rated health of Danish employees. For the immigrant group, none of the items relate significantly to the three outcomes. For these reasons, our results offer overall support to Hypothesis 2 suggesting that transformational leadership affects individual outcomes differently for employees with native Danish and immigrant backgrounds.
The seven transformational leadership item loadings on three outcomes: Danes and immigrants
DISCUSSION
Challenging the assumption that transformational leadership is perceived and works similarly across cultures and nationalities, we set out to perform a within-country test of the assumption of universalistic transformational leadership.
Leadership universality or not?
We found no difference in the perception of transformational leadership between native Danish and immigrant respondents – neither in reported mean level, in pattern and magnitude of item loadings, nor in pattern of association by item intercorrelations. While these findings correspond with some current research suggesting a trend towards cultural convergence in employees’ perception of successful managers (Lankau & Chung, Reference Lankau and Chung2009), they do not support the contingency proposition that cultural differences determine different perception processes of leadership (Yan & Hunt, Reference Yan and Hunt2005). Although our findings could indeed be interpreted to lend support to the universality theory of transformational leadership, the fact that we find similar perceptions across native and immigrant groups could reflect an assimilation process rather than transformational leadership universality. According to the Merriam-Webster Dictionary, the word assimilation means: ‘To adopt the ways of another culture: to fully become part of a different society, country, etc.’ (http://www.merriam-webster.com/). The average current workplace seniority of the immigrants in our sample is 7 years, suggesting that there is indeed a probability that such an assimilation mechanism could have influenced the leadership perception of the participating immigrant group. At first glance, our findings relative to the perception of transformational leadership do therefore not offer an unequivocal answer to the question of universality.
When shifting focus to outcomes, our results cease to suggest any support for the universality of transformational leadership. In Model 1 (without baseline control), we find analogous patterns of association across the two employee groups with regard to job satisfaction and well-being. In Model 2 (with baseline control thus referring to the change in outcomes from baseline to follow-up), the differences between employee groups increase. For native Danes, transformational leadership remains a significant predictor of change in job satisfaction and well-being, but not in self-rated health. For immigrants, transformational leadership predicts change in none of the three outcomes. Therefore, transformational leadership seems more important for the development of job satisfaction and well-being of Danish employees than of their immigrant counterparts. The seven transformational leadership items’ impact on outcomes also differs between the native Danish and immigrant groups, which leads to questioning the universality of transformational leadership, also at component level. Our findings add to a more general picture suggesting that native and immigrant employee groups do experience different effects of leadership. A Swedish study found, for example, that dissatisfaction with current work and management was related to stress and turn-over intentions for natives, but not for immigrants (Rosmond, Lapidus, & Björntorp, Reference Rosmond, Lapidus and Björntorp1996).
These are important findings, suggesting that while the perception of transformational leadership is convergent for native Danes and immigrants, the outcomes of transformational leadership are not. The differentiated outcome relation of transformational leadership across the two employee groups suggests that, despite potential assimilation effects and possible convergence trends towards Western leadership concepts, transformational leadership effects are not universal.
Leadership and health
Tapping into the specific question of leadership and health, we found that transformational leadership was not related to change in employee self-rated health status in either of our groups. We offer three perspectives to this finding. First, we applied a measure of general self-rated health – not aiming specifically at either physiological or psychological health aspects. Kuoppala et al. (Reference Kuoppala, Lamminpää, Liira and Vainio2008) suggest that leadership practises may be more directly related to psychological health symptoms than to physiological. Whether transformational leadership relates differently to different aspects of health – and whether this holds true universally still remains unanswered. Second, the data collection time-lag of our study was 2 years. In the context of the job well-being pyramid (Kuoppala et al., Reference Kuoppala, Lamminpää, Liira and Vainio2008), the path from leadership to health goes through various intermediate outcomes. For such transitions to have a significant bearing on health may require an even longer time lapse. Theorell, Nyberg, Bernin, Oxenstierna, Romanowska, and Westerlund (Reference Theorell, Nyberg, Bernin, Oxenstierna, Romanowska and Westerlund2010) applied for example a mean time lapse of 9.7 years, finding an association between leadership and ischaemic heart disease. Third, the nonsignificant relation between transformational leadership and self-rated health for immigrants may relate to the generally lower immigrant health status. Exceptional life circumstances both before and after immigration may have contributed to their poorer health status, which therefore remains unrelated to transformational leadership. The exact explanatory mechanisms of this relation remain a question for future research to address.
Based on our findings, there is a need for understanding how to develop efficient leadership skills, which positively affect the job satisfaction, well-being, and self-rated health of both natives and immigrants. This need becomes even more pertinent as our study reveals a situation of worse self-rated health and well-being for the immigrant employee group. Lower self-rated health of immigrant employees has formerly been demonstrated in both quantitative (Borrell & Dallo, Reference Borrell and Dallo2008) and qualitative studies (Ortega, Carneiro, & Flyvholm, Reference Ortega, Carneiro and Flyvholm2010).
Limitations
In the interpretation of our findings, five main limitations must be taken into account. First, despite our relatively large sample size, the number of immigrants was not large enough to uncover potential within-group variance, which would have been possible with a larger immigrant sample size. Given that sample size affects the calculation of statistical significance, the relatively small sample size of the immigrant group could further have led to nonsignificant results. However, while small n is a precondition for any subgroup study and significant immigrant results were indeed identified prior to inclusion of baseline measurements, the impact of a small immigrant n on our study seems minor. Despite the smaller immigrant sample size, the composite nature of the immigrant group may have increased the generalizability of our findings. Second, as the immigrant sample had an average of 7 years of employment in their current workplace, the degree of assimilation may have been relatively high. Further, due to the categorization of employees, the native group may consist of some employees with one non-Danish parent. This situation may to some degree have influenced the responses given by a part of the native Danish employee group. While this categorization may be seen as a limitation to our study, it contrarily also suggests that the effects, we do find, are strong indicators of important between-group variance. Third, our survey was administered in Danish, wherefore our two participant groups did not have the same preconditions for responding. While the application of one questionnaire version ensures measurement equivalence, survey response patterns may differ slightly across languages for bilingual respondents (Marin, Triandis, Betancourt, & Kashima, Reference Marin, Triandis, Betancourt and Kashima1983). The length of service of our immigrant group suggests, however, that the overall language skills would be in a condition to respond both meaningfully and purposefully to the Danish language questionnaire. Fourth, the female domination of the elder care sector should be taken into account when generalizing results to other sectors with different gender distributions. While gender differences in transformational leadership have mostly been studied with regard to the gender of the manager (Eagly, Johannesen-Schmidt, & Van Engen, Reference Eagly, Johannesen-Schmidt and Van Engen2003), there is some indication that both male and female employees are more satisfied with transformational than with transactional leadership (Druskat, Reference Druskat1994). Despite the inherent limitation of the gender distribution of the sample, the similarity of gender preference for leadership style supports the generalizability of our findings. Fifth, the sample covers different educational levels (the time length of training programmes ranging from 7 weeks to 42 months). These differences in educational levels also represent differences in employee roles and responsibilities, which could influence the perception of leadership. However, as the distribution of educational level is similar for the Danish and immigrant respondents and the Danish culture is characterized by a very low power distance, any impact due to educational level is considered minor.
CONCLUSION
The universalistic paradigm holds the potential of simplifying the phenomenon of leadership in three ways: First, it implies that the same criteria may be applied to selecting people for positions of leadership in different contexts. Second, it implies that the same leadership training and development programmes may be given to leaders in different contexts. Third, it implies that leadership can be relatively easily transferred from one context to another (Shamir, Reference Shamir2012). Transformational leadership theory is a theory of general principles that only marginally attends to situational factors, as emphasized by Shamir (Reference Shamir2012). Newer theories such as authentic leadership theory (see e.g., Gardner, Avolio, Luthans, May, & Walumbwa, Reference Gardner, Avolio, Luthans, May and Walumbwa2005) seem to follow the same path, as do popular leadership models which have been written primarily for practitioners (Collins, Reference Collins2006; Bennis, Reference Bennis2009; Kouzes & Posner, Reference Kouzes and Posner2012). Although these approaches suggest that certain leadership principles apply across contexts, there are also indications that leadership can take different forms and that relations among leadership variables and outcomes are likely to vary from one situation/context to another (Shamir, Reference Shamir2012; Klimoski, Reference Klimoski2013). Frequent calls for integrating context in leadership research have indeed been made (see e.g., Shamir & Howell, Reference Shamir and Howell1999; Boal & Hooijberg, Reference Boal and Hooijberg2001; Osborn, Hunt, & Jauch, Reference Osborn, Hunt and Jauch2002).
The key findings of our study reveal that – in our study sample – there is no difference between native and immigrant employees’ perception of transformational leadership. We found no significant difference between the native Danish and the immigrant respondents in the reported mean level neither for the general construct, nor for the seven individual indicators. Item correlations, factor structure, and item loadings were analogous across groups. With regard to the outcomes of transformational leadership, we did find differences between the native Danish and immigrant employee groups. When controlled for baseline levels, transformational leadership predicted job satisfaction and well-being in the native group – but not in the immigrant group. Hence, these findings suggest that even when perceptions of leadership may be similar across national and immigrant employee groups (be that ascribed to universality or assimilation); national backgrounds do matter for employee-related outcomes of leadership, which therefore lends little support to the universalistic theory of transformational leadership.
Suggesting that national background matters for employee-related outcomes of leadership even within a shared national employment context, our study supports the importance of taking context into consideration when it comes to understanding organizational behaviour (Härtel & O’Connor, Reference Härtel and O’Connor2014) and leadership (Klimoski, Reference Klimoski2013). With the purpose of bringing both leadership research and practise forward, we suggest that leadership research orients itself away from the implicit absoluteness within both the universalistic and contingent paradigms. Instead, we should dedicate ourselves to defining and testing boundary conditions for leadership – across contexts (nations, cultures, employment relations, companies, sectors), individuals (preferences, characteristics, expectations, cognitions, and emotions), and implementation (situations, processes, interactions). We thereby argue for a combination of paradigms, proposing that leadership styles may hold universal potential – yet at the same time entail constraints concerning contexts, individuals, and implementation. Our study illustrates such a combination, finding on the one hand a culturally convergent or even (simple) universal perceptions of transformational leadership, while on the other hand demonstrating that transformational leadership does not work as a (functional or variform functional) universal agent for developing job satisfaction and well-being across employee groups. Keeping our study limitations in mind, the differences we find within a shared Danish workplace context do, however, increase the certainty with which we can identify them as related to the contextual elements of national and cultural backgrounds. Some of the opportunities for further research arising from our study suggest studies with larger immigrant group samples, and samples with more variation regarding types of professions and employee roles and responsibilities. Future research could also benefit from investigating the perception and effects of a variety of leadership behaviours.
Suggesting that positive developments in job satisfaction and well-being may hold aspects of both leadership universality and contingency, our findings also become relevant to management practise. In a globalized world where composite employee groups are becoming the rule rather than the exception, applying universal leadership principles with a contingency practise may be an appropriate approach for leaders who wish to exert cultural sensitivity and managerial actions reflecting global leadership skills. The demand for adaptability and flexibility in rapidly changing workplaces adds to the complexity of the questions of perception and impact of leadership. Nonetheless, more knowledge on which, how, why, and when a variety of leadership behaviours (beyond transformational leadership) work in nationally and culturally composite contexts is needed to support the development of leadership skills that are effective in complex organizations employing both native and immigrant groups.
Acknowledgements
The study was supported by a grant from the Danish government (SATS 2004) administered by the Ministry of Employment.
