Ethical leadership in the Netherlands: testing the broader conceptualization and measurement scale Free

Recent scandals associated with moral lapses of leaders have stirred widespread discussions. Examples include the non-declaration of assets by a former Icelandic prime minister which led to his resignation after revelations in the Panama papers (Henley, 2016). Similarly, in the private sector, fraud and embezzlement by the former chairman of Hyundai Motors led to a jail sentence (Tharoor, 2010). Although such controversies are not new, they can have adverse effects. In the public sector, ethical issues can lead to a lack of trust among the general public in public institutions. It is for these reasons that ethical leadership has remained a topic of importance within academia (Shakeel et al., 2019).

However, the literature presents multiple conceptualizations of ethical leadership, such as the seminal construct by Brown et al. (2005) and subsequent work by Kalshoven (2011), Eisenbeiss (2012), Voegtlin (2016), and Van Wart (2014). Shakeel et al. (2020a) divided these studies into classical and contemporary conceptualizations. The classical conceptualization pertains to studies that include and uphold the construct of Brown et al. (2005), which portrays ethical leadership as a set of attributes including the personal values of leaders and their efforts to instill these in followers. Contemporary conceptualizations refer to studies that include attributes such as empowerment, attention for diversity, attention for the capacity for leader learning, and care for the well-being of followers.

The multitude of conceptualizations has given rise to numerous measurement tools as well, such as the ethical leadership scale (ELS) by Brown et al. (2005), the ethical leadership at work (ELW) by Kalshoven (2011) and the ethical leadership questionnaire (ELQ) by Yukl et al. (2013). The difference in their focus mirrors the variance in approach outlined above. For example, the responsibilities of the leader have a narrower focus in the classical conceptualization than in the contemporary conceptualization. Classical conceptualizations stress the moral values of a leader and followers with some non-explicit emphasis on external stakeholders, whereas contemporary studies stress leaders’ attention for internal and external stakeholders with a particular focus on the wellbeing of employees, the environment and its sustainability. Critics of the classical conceptualization also highlight a number of other limitations that we will explain in the upcoming section.

The existence of different definitions and tools hinders generalization and replication of findings. It is important that a common ethical leadership conceptualization and scale are developed. That could also address the ambiguity between ethical leadership as a leadership style and related leadership styles. At the moment, the classical and contemporary conceptualizations present different answers to the question about the association between ethical leadership and other styles. Classical conceptualizations deem ethical leadership to be distinct from relating styles although, according to these studies, ethical leadership has links to styles like transformational and transactional leadership (Brown et al., 2005). Some contemporary studies maintain that numerous ethical leadership styles exist as varieties of other styles (Van Wart, 2014).

Shakeel et al. (2020b) have proposed an integrated conceptualization and definition of ethical leadership. This conceptualization was based on the categorization of ethical leadership by Van Wart (2014). They systematically synthesized existing studies to include both the ideas of earlier classical as well as recent contemporary schools of thought. The integration of both schools of thought led to a broader conceptualization and definition, and a new integrated measurement tool, the broader ethical leadership scale (BELS). Subject to empirical testing, the study by Shakeel et al. (2020b) will help in bringing forward a new conceptualization, definition and measurement tool of ethical leadership. This will also help in determining the link of ethical leadership and associated leadership styles.

The need for the broader ethical leadership scale (BELS) stem from the lack of a survey scale that covers a broader conceptualization. The existing scales were either based on the classical conceptualization, thus, predominantly based on a two-dimensional constructs or based on more contemporary studies and ignored some basic sub-values of classical conceptualization. Hence a scale that covers both the essentials of the classical conceptualization and contemporary values as well as reflecting the interconnectedness of the associated styles is missing. The broader ethical leadership scale (BELS) aims to address these missing links and provide a scale that is inclusive of all major sources of ethical values.

The goal of this study is to perform a first test of the broader ethical leadership scale which can be used to corroborate the conceptualization of Shakeel et al. (2020b). This is achieved in a number of steps, creating the most parsimonious measurement scale. In doing so it aims to find preliminary answers to three overarching questions; (1) Is ethical leadership broader than the construct put forward by Brown et al. (2005)? (2) Is ethical leadership a uni-dimensional or a multi-dimensional construct? (3) How does ethical leadership relate to other leadership styles? This test will help to determine whether ethical leadership should be studied from a broader lens (including more contemporary attributes) and it will aid in providing an authentic tool that can be used for replication purposes. In answering the third question, this study will also establish whether ethical components of various leadership styles are all dimensions of one style of ethical leadership as proposed by Van Wart (2014).

We first shortly review the literature on ethical leadership and discuss the development of the broader ethical leadership scale. The method section presents a brief introduction about the nature of our sample size, which was based in the public sector in the Netherlands (909 public servants in leadership positions) and elaborates our analytical approach which is carried out in four different steps. The article then proceeds to the results of the analysis. A discussion regarding the role of findings in theoretical development and considerations for a future research agenda will conclude the chapter.

Ethical leadership has been defined by Brown et al. as “the demonstration of normatively appropriate conduct through personal action and interpersonal relationships, and the promotion of such conduct to followers through two-way communication, reinforcement, and decision-making” (Brown et al., 2005, p. 120). This definition implies two dimensions of ethical leadership: the moral person (referring to the personal qualities of the leader) and the moral manager (referring to the activities undertaken by leaders to guide followers towards ethical activities).

These dimensions were apparent in the definition provided and upheld by studies in the classical conceptualizaton. The researchers in this tradition viewed ethical leadership as a separate construct from relating other leadership styles, although they did not deny strong overlap. The contemporary studies identified new attributes stretching the responsibility of the leader from the limited role of managing the self and followers’ ethics to taking into account the responsibility of stakeholders outside the organization such as external stakeholders, society and environment.

Critics of the classical school of thought highlight a number of limitations of the original construct. These limitations revolve around the incapacity to address (a) a number of vital and contemporary leadership attributes such as leader learning and ethical competency, (b) the lack of addressing stakeholders outside the organization, and (c) the lack of establishing linkages to ethical components of related leadership styles. The link to other leadership styles, such as transformational leadership, has remained a puzzle avoided by classical and contemporary ethical leadership scholars alike. However, ethical leaders have been acknowledged to use multiple styles (Brown et al., 2005) depending upon the context of a situation (Heres and Lasthuizen, 2012).

A study by Shakeel et al. (2020b) presents an overview of the limitations of the classical conceptualization and proposes a broader conceptualization to overcome these problems. They present an integrated, broader conceptualization and associating scale, the Broader Ethical Leadership Scale (BELS). The BELS has increased the scope of ethical leadership using two core assumptions; that contemporary attributes form part of ethical leadership and that related leadership styles constitute an integral part of ethical leadership (cf. Van Wart, 2014). Important contemporary attributes such as role clarification and sustainability were already identified as components of ethical leadership by some contemporary scholars (Kalshoven et al., 2011; Eisenbeiss, 2012) but Van Wart’s (2014) claim that other styles are sub dimensions of ethical leadership has not yet been studied empirically.

The broader definition of ethical leadership is: “Ethical leadership is the implicit and explicit pursuit of desired ethical behavior for self and followers through efforts governed by rules and principles that advocate learning motivation, healthy optimism and clarity of purpose to uphold the values of empowerment, service to others, concern for human rights, change for betterment and fulfilling duty towards society, future generations, environment and its sustainability” (Shakeel et al., 2020b, p. 16). With this broader construct in mind, existing assessment scales measuring ethical leadership fell short of measuring key attributes of the new conceptualization. Hence a new survey tool encompassing all the newly added values was needed.

The BELS is an amalgam of items from numerous existing scales and some new ones based on the literature (Shakeel et al., 2020b). The items included in this scale are related to each of the numerous ethical leadership varieties listed by Van Wart (2014). These styles include virtuous, authentic, positive, moral manager, servant, spiritual, CSR and transformational leadership styles. Table 1 shows the ethical concerns of the various leadership styles and their focus on internal and/or external stakeholders, which have all been included in the broader conceptualization, and consequently in the BELS. Each of these styles is treated as a sub dimension of the broader concept of ethical leadership that is measured by the BELS. The original version of the BELS is included in the  Appendix A. A more detailed description of the composition of the BELS follows below.

The BELS consists of 44 items pertaining to 9 ethical leadership sub dimensions. Most items were taken from existing scales: the ethical leadership questionnaire ELQ (Yukl et al., 2013), virtuous leadership scale (Wang and Hackett, 2016), servant leadership scale (Dennis and Bocarnea, 2005), and items measuring CSR, transformational and other leadership attributes (Avolio et al., 1999; Turker, 2009; Kalshoven et al., 2011). New items were developed to measure Spiritual leadership as there is no scale for this style yet. For example, the item “I as leader, encourage bereavement programs” has been self-developed based on Van Wart (2014). Four negatively coded items based on the existing scales were added to counter respondent bias. All items were translated to the Dutch language by the two Dutch co-authors and then translated back by an independent scholar. This led to some minor editorial adjustments. Respondents were asked to which extent they agreed that the propositions fitted with their style as a leader, on a 5-point Likert scale. As a pilot study the questionnaire was tested for structure, clarity and wording by an expert panel of five leadership scholars. After discussing the results, some small amendments were made, leading to the final version of the measurement scale.

To check for bias relating to self-assessment, we also incorporated a social desirability scale (Hays et al., 1989). This is important because self-assessment surveys run the risk of self-favoring tendencies by respondents suffering from low self-esteem (Wells and Sweeney, 1986). Other items that inquired into demographic details included questions pertaining age, gender, sector, education and number of employees under the leadership of the respondent.

The data for this study were collected from public servants in a leadership position in The Netherlands using Flits panel ^[1]. The BELS questionnaire was sent via a link by email between 25th May and 18th June, 2018 to 3645 public officials in leadership positions in all types of public organizations represented in the panel. After one reminder, it received a response of 25% by a total of 909 respondents. Of these respondents, 653 (71.8%) were male and 256 (28.2%) female respondents. This distribution is representative of the population of public sector employees in a leadership position, as evidenced in the annual report on the Dutch national public service (Ministerie van Binnenlandse Zaken en Koninkrijksrelaties, 2019) The respondents represented public officials working in education (30.1%), semi-autonomous government agencies (26.3%), the police force (9.6%), municipalities (12.5%), judiciary (0.8%), army (2.8%), water boards (3.5%), research institutions (4.7%), university medical centers (6.2%), provinces (2.5%) and regional government bodies (1.0%). Respondents varied in age (mean: 55, sd = 7.39 range: 32–73 years) and background qualifications; most respondents had a higher education diploma (polytechnic 39%, university 28.3% and doctoral degree 13.8%) while the others mostly had technical educational diplomas. All respondents received a small report with descriptive findings, in return for their voluntary participation (which is the standard Flitspanel way of working).

The analysis was carried out in four distinct steps. In the first step ‘Test of the proposed factor structure”, we determined the authenticity of the BELS scale as designed by Shakeel et al. (2020b). To this end, we used confirmatory factor analysis. The associating model Figure 1 depicts this scheme using AMOS (version 25) SPSS package to determine model fit. This fit, however, did not yield the desired results. To check if and which attributes and dimensions of the broader conceptualization hold true, an exploratory factor analysis was needed. Therefore, in the second step ‘The new one factor model of ethical leadership”, we carried out an exploratory factor analysis in SPSS version 25. A parallel analysis test for Eigenvalues was used to determine the significant Eigenvalues, and a reliability analysis was used to determine the reliability of items from each factor, allowing us to drop redundant items and achieve a parsimonious end-result. The findings from this confirmatory analysis were retested in the third step, ‘Test of one factor model of ethical leadership” using Model fit in AMOS. This led to a small improvement, reducing the number of items in the final scale to 33 instead of 35. Step 4 of the process discusses rooting out social desirability bias. We now turn to the results, followed by a discussion on their impact on theory development and an agenda for future research.

• Step 1: Test of the proposed factor structure

Figure 1 shows the results of the AMOS analysis for the nine sub dimensions of ethical leadership: Virtuous, Positive, Authentic, Moral Manager, professionally grounded, spiritual, servant, CSR and Transformational leadership. The values depict a weak fit of the proposed model although there are multiple high factor loadings for some items that represent contemporary factors. We believe that these results can be interpreted such that even though not all dimensions of ethical leadership are confirmed, there are sufficient elements that appear to be of significance, and which point to the existence of a broader concept than the original EL scales. This will be tested in the next step, but first we discuss the findings from step 1.

There are three measures to express the model fit values of the 9-dimensional ethical leadership construct. CMIN refers to the minimum discrepancy and according to literature its value should be closer to 1 and less than 3 for a good fit (Hu and Bentler, 1999 as cited in Gaskin, 2016). The 9-dimensional model yields a CMIN value (3.267) which does not indicate a good fit. We turn to the other indices, RMSEA and CFI. RMSEA shows the fit of the hypothesized model to a perfect model (Xia and Yang, 2019) and its value should lie below or be equal to 0.08 (Brown and Cudeck, 1993) and less than 0.05 for a good fit (Xia and Yang, 2019). The value for our 9-dimensional model (0.050) again does not show a good fit. PCLOSE is the p value of test on RMSEA (Hox and Bechger, 1998) and its desired value is greater than 0.05 (Hu and Bentler, 1999) whereas value for our model is (0.507). The CFI shows the fit of the hypothesized model with baseline model (Xia and Yang, 2019, p. 409) and its acceptable value is greater or equal to. 90 (Bagozzi and Yi (1988). Some researchers recommend a value greater than 0.95 for a better fit, however, deem value above 0.80 as permissible (Hu and Bentler, 1999 as cited in Gaskin, 2016). The 9-dimensional model has a CFI of (0.756) which is not acceptable. In sum, the full 47-item BELS cannot be corroborated, but our next step will show that there is support for a smaller, more parsimonious version of the BELS.

• Step 2: The new one factor model of ethical leadership

Given the bad model fit in step 1, we proceeded with an exploratory factor analysis (with no fixed factors) to investigate the factor structure of these items using SPSS, based on Eigenvalues and scree plots (see Figure 2). Although it was apparent from the Eigenvalues that a 11-factor solution was possible, the scree plots indicated that a 7-factor solution is more likely. We repeated the EFA multiple times keeping the number of factors fixed from 1 till 7. Table 2 shows the number of items, the nature of these factors and reliability scores for these solutions.

These analyses did however not reveal very strong individual factors. Three factors stood out. We labeled these factors as (1) overall moral values (which included mostly items pertaining to virtuous leadership), (2) a factor with items pertaining to principles and human rights, and (3) a factor consisting of items on sustainability. Table 2 illustrates the details of this factor analysis. These three factors had however small numbers of items and low reliability. Therefore, we decided that there was insufficient evidence for a multiple factorial solution and concluded that a one-factorial solution offers the strongest results. The total number of items of this one-factor solution was 43 when we began the analysis. Then we decided to suppress items with factor loading below 0.3, which led to a 35 items factors (reliability alpha = 8,85). This overall high reliability could be caused by the high number of included items. Therefore, we decided to run a model fit for the 35 items using confirmatory factor Analysis (CFA), see step 3.

Based on the analysis so far, we can conclude that ethical leadership does include all the variables from the contemporary perspective that were identified by Shakeel et al. (2020b). It also upholds the assumption that instead of a separate style of its own (as indicated by the classical studies), ethical leadership is a uni-dimensional construct. Ethical leadership can be explained as the sum of the ethical components of various leadership styles, as items pertaining to these multiple leadership styles are present in the one-factor construct of ethical leadership.

• Step 3: Test of one factor model of ethical leadership

Next, we checked the model fit of the one factor model with 35 items using CFA in AMOS. The value of loading for these items ranged from 0.66 to 1.11. We encountered two high cases of covariance between the error terms. These values were 66,900 between the error terms e8 and e9, and a value of 96,939 between the error terms e24 and e30. The error terms e8 and e9 correspond with the items ‘Am resilient in nature” and ‘Am an optimist”. Although using a single factor model, we are able to covary any of these error terms, however, it is interesting to know that both these items represent a similar underlying theme which can be referred to as ‘positive leadership”. The error terms e24 and e30 corresponds to the items, ‘make investment to create a better life for future generations” and ‘Show concern for sustainability issues”. A clear underlying theme behind these two items is ‘Sustainability”. We co-vary these error terms and Table 3 provides model fit values before and after co-varying.

The model fit for the uni-dimensional ethical leadership construct shows a strong and better fit than the 9-dimensional model. The RMSEA for the model fit is 0.044 which depicts a good fit and is lower than the desired 0.05 and the 0.049 result of the 9-dimensional model. The CMIN which ought to be closer to 1 (Hu and Bentler, 1999) is 2.760, which is also better than the 3.212 of the model we dismissed in step 1. Finally, the CFI of the one factor model is 0.849 and thus higher than the permissible (>0.80) value according to Hu and Bentler (1999). These findings strengthen the results of the exploratory factor analysis in step 2.

To reach an even more parsimonious solution, we tested the effects of removing the co-varying items mentioned before, see Table 3 for results. A 33-item solution, taking out two covarying items, renders the best model fit values of all. The removed items are ‘Am an optimist” (e9) and ‘Show concern for sustainability issues” (e30). Although concern for sustainability is thus removed from the final scale, concern for the environment – which is an almost confounding idea – is represented through items 38 ‘Am also responsible for society and environment of my organization” and item 32 ‘Make investments to create a better life for future generations’.

Compared to the reliability of the 35-item scale (Cronbach alpha 0.889), removal of the items 14 and 34 does not change the reliability of the scale (same alpha). Achieving a robust and parsimonious measurement instrument is of primary importance for this validation. So, while, sustainability is a vital attribute of ethical leadership, already identified by Kalshoven (2011), we feel that other items can cover it. “Sustainable development is a process of change in which the exploitation of resources, the direction of investments, the orientation of technological development, and institutional change are all in harmony and enhance both current and future potential to meet human needs and aspirations” (Wheeler and Beatley, 2014, 57 as cited in Zeemering, 2009). Taking into account this definition, item 32 ‘Make investment to create a better life for future generations’ and item 38 ‘Am also responsible for society and the environment of my organization’ can together ensure that sustainability remains an integral component of the BELS. Hence, we opt for the 33-item scale as the best representation of the broader conceptualization of ethical leadership.

• STEP 4: Social desirability bias

The BELS included a social desirability scale (Hays et al., 1989). This was used to check for potential bias that could arise due to self-assessment of their leadership by the respondents. To check for social desirability bias, we check the model fitness for the items of the social desirability scale and correlate this with our current model to achieve a good fit. The correlation matrix between the unidimensional BELS and the social desirability scale is negative (−0.051), and not statistically significant. Thus, there is no evidence that results are confounded by social desirability bias.

The model fit analysis has supported a 33 item solution for the construct of ethical leadership as a single factor conceptualization. This is indicative of a broader conceptualization of ethical leadership with the presence of additional elements missed by the classical conceptualization, as argued by Shakeel et al. (2020b). Essentially ethical leadership now does not only include the display of ethical behavior and its promotion inside the organization as the classical theory suggests, but also includes a focus on all direct and indirect external stakeholders, as well as values such as empowerment, leadership learning, and human rights. These elements increase the scope of responsibility for ethical leaders and make it possible for leaders to navigate carefully while considering basic as well as evolving values. The 9-factor assumption adopted by the study initially did not hold true, however, our explanation for an one factor conceptualization supports the concept of ethical leadership as a singular phenomenon (Shakeel et al., 2019), and hints at a common ethical denominator for all leadership styles instead of ethical leadership being a distinct style of its own.

The testing of the BELS confirms the existence of an unidimensional scale of ethical leadership, instead of a multidimensional scale, and so we can conclude that the different elements of the broader conceptualization are highly integrated. The results from this study have multiple implications not only for our proposed model but also for our current understanding of ethical leadership. They indicate that the classical conceptualization of ethical leadership is indeed an incomplete model of ethical leadership. This confirms the limitations of the classical conceptualization as outlined by Shakeel et al. (2020b). Empowerment, need for leader learning, and responsibility of the organization towards the environment are essential parts of ethical leadership.

In light of these results, we can now answer the three questions posed at the start of this study, before delving into an in-depth explanation of each component of the new definition. The first question concerned the scope of ethical leadership in relation to the classical conceptualization. Our findings suggest that the different elements can be theoretically distinguished but are empirically closely related, that is to imply a leader who advocates ethical conduct within an organization is also more likely to advocate diversity, human rights and concern for environment. Hence, one element does not seem to occur without the other. This supports a broader conceptualization than presented in the classical approach.

“The new conceptualization of ethical leadership includes additional factors such as role clarification (factor validated in ELW by Kalshoven et al., 2011), leadership learning, ethical competence, resilience, openness (transparency), human rights, and consideration for environment, society and future generations. A lack of attention to these factors can result in not recognizing sources of moral lapses. Since leaders find themselves increasingly in unprecedented dilemmas.

This enhances the scope of ethical leadership considerably since factors such as human rights have not been a part of ethical leadership conceptualization earlier, even though it has been a source of problems for companies facing such allegations, for instance Chevron, which faced backlash for human rights and environmental abuses (Global Exchange, 2019). Similarly, leaders’ lack of learning, whether be it reluctance to inform themselves about new environmental friendly practices or not anticipating how products and services might be misused in consumption, can lead to untoward circumstances and possible moral red flags. Such considerations put the onus of responsibility not only to supply chains (for transparency purposes) and production processes but also to responsible use and consumption, and forces leaders to contemplate the possibility of preventing or mitigating the misuse of products and services. Such foresight cannot be guided by policies and regulations only (moral management) but will increasingly require the agency of committing and upholding principles and values of organizations identified and upheld by leaders in key decisions, both in daily practice and during crises (He, 2023).

Hence encompassing these factors leads to a conceptualization which leaves no room for unethical pro-organizational behavior. This also negates the conception of ethical leadership as separate from other leadership styles and maintains it as a central consideration of all leadership conceptualizations. This is in line with our understanding of certain leadership styles which holds moral and immoral (pseudo component) sides like transformational leadership (Bass and Steidlmeier, 1999 as cited in Barling et al., 2008; Shakeel et al., 2019). Hence all leadership styles have an embedded ethical component which greatly enhances the need to include factors from outside of the organizations into the ethical leadership conceptualization. To address varying allegations and dilemmas faced by corporations today, the broader conceptualization (by including all these factors) can help leaders and companies identify and minimize potential ethical lapses.

Furthermore, the confirmation that ethical leadership encompasses all other styles makes clear that ethical leadership as measured by the BELS is an unidimensional construct, which answers the second question and puts at rest other theories such as the two-dimensional classical conceptualization and the nine-dimensional vision of Van Wart (2014). In this way ethical leadership neither distinguishes between different tendencies or forms, but points towards one phenomenon which can be enhanced and perhaps studied better as a process (Shakeel et al., 2019). Another plausible argument for concluding that the BELS measures a single factor construct is that elements such as leader learning cannot be attributed to any particular leadership form or style, but is an integral component of the overall phenomenon which binds all elements towards a singular skill present in leaders in varying levels, as practical behavioral ethics is not binary but present as a spectrum in leaders.

The findings from this study also provide an answer to the third question about the link between ethical leadership and other leadership styles. The inclusion of items representing all multiple leadership styles in the final one-factor model suggests that ethical leadership is an amalgam of ethical components of all associating leadership styles and not a separate style of its own. However more research is needed to further test and support this claim.

To structure the further discussion of our findings, we will use the different elements of the broader definition and then subsequently go into the relation of ethical leadership and associated leadership styles. The broader definition of the ethical leadership includes the following elements: “Ethical leadership is the implicit and explicit pursuit of desired ethical behavior for self and followers through efforts governed by rules and principles that advocate learning motivation, healthy optimism and clarity of purpose to uphold the values of empowerment, service to others, concern for human rights, change for betterment and fulfilling duty towards society, future generations, environment and its sustainability” (Shakeel et al., 2020b, p. 16).

One of the most important shortcomings of the classical definition of ethical leadership related to the vagueness of the phrase: ‘demonstration of normatively appropriate conduct’. This warrants a two-point explanation for readers. It is true that on the one hand ethics are relative and tend to vary across cultural boundaries, and it could be for this very reason that this term is left vague for scholarship intentionally. However, on the other hand, normatively appropriate conduct could also refer to the activities that have been developed over time within organizations and have become embedded within organizational culture. Although beneficial for the organization, such activities may not be good for stakeholders outside the organization. Studies relating to unethical pro-organizational behavior (Miao et al., 2013; Graham et al., 2015; Zhang, 2020) specifically discuss instances when commitment towards the own organization surpasses the ethical boundaries for concern of other societal and environmental factors. Such activities may still be part of a present ‘norm”, especially in the absence of laws that take into account environmental concerns.

The inclusion of numerous attributes of the broader conceptualization serves to put an end to this vagueness by highlighting the boundaries of concern for an ethical leader, which are not limited to the organization but also take external stakeholders into account, which is missing in the classical conceptualization of ethical leadership (Den Hartog, 2015; Frishand and Huppenbauer, 2014; Voegtlin, 2016 as cited in Shakeel et al., 2020b). This clearly extends the responsibility of leader from previously ‘organizational focus” to broader concerns. All the items of the new scale that advocate concern for society and environment prove useful to establish this narrative. Associated items of the BELS are, ‘Am also responsible for society and the environment of my organization’ and ‘Oppose the use of unethical practices to increase performance’.

Numerous studies over time (e.g., Eisenbeiss, 2012; Den Hartog, 2015) have questioned the scope of the responsibility of an ethical leader. It has not been clear whether the role of an ethical leader is limited to her/his own organization or if that jurisdiction exceeds immediate environment. This study has now established and redefined the scope of ethical leadership. The responsibility of ethical leaders is not limited to the organization alone but outside the organization as well towards the society and the environment in which the organization operates. This responsibility is not time-bound but addresses the needs of future and upcoming generations as well.

Classical theories of ethical leadership have failed to identify external stakeholders as part of a leader’s responsibilities. The studies by Eisenbeiss (2012), Kalshoven et al. (2011) and Voegtlin (2016) stand at contrast with the classical conceptualization in terms of attention for external stakeholders. The broader conceptualization makes it clear in concept and definition that the responsibilities of ethical leaders are not limited to the organization itself but also includes external stake holders including the society in which the organization operates, and the environment. The associated item with this attribute is ‘Am also responsible for society and the environment of my organization’ .

Now that we have established that ethical leaders are also responsible for activities outside the organization, it becomes vital to understand how the ethical leader operates within the organization. Classical conceptualizations defined ethical leaders as the sum of two main dimensions: the moral person and the moral manager. The moral person includes the virtues of the leader her/himself whereas the moral manager refers to the activities that ethical leaders undertake to promote ethical behavior in their followers. These activities include role modeling, reinforcement and communication regarding ethics. Our results show that besides these, it is also important for an ethical leader to strongly advocate the empowerment of employees.

Although empowerment has been a point of discussion for scholars in the classical tradition, it has been missing in the definition put forward by Brown et al. (2005) and the survey scales associated with it. Later studies in the classical tradition already mention empowerment as an integral part of ethical leadership (Heres and Lasthuizen, 2012). Empowerment is part of the broader definition and included in the validated BELS scale through the item ‘Empower others with opportunities so that they develop their skills’.

Ethical leadership studies, both classical and contemporary schools, have ignored the importance of learning for ethical leaders whereas some leadership studies attribute learning as one of the major functions of being a leader. According to Voegtlin (2016) it is an important part of leadership responsibility. This attribute has been associated with a professionally grounded leadership style. To be able to choose between competing values and tackling ethical dilemmas, leaders need expert skills. While experience can be of good help, to keep track of changes in this fast-paced environment learning becomes important not just for followers but more so for leaders to learn and apprise followers of the information needed to perform their jobs more effectively. Learning by leaders was assessed using one item in BELS: ‘Am committed to lifelong learning’.

Some contemporary studies of ethical leadership identified and addressed role clarification as part of ethical leadership (De Hoogh and Den Hartog, 2009; Kalshoven et al., 2011). Our study has shown that role clarification (in terms of expectation of leaders from followers relating outcomes) is an integral component of ethical leadership. However, we believe that there can be two possible interpretations/components of role clarification, which Shakeel et al. (2020b) failed to distinguish in their original proposition. These include (1) communication or expectations of ethical standards from followers and (2) expectation in terms of output or performance. Although Shakeel et al. (2020b) criticized the classical studies for ignoring this aspect, we believe that the classical studies do partly address the concerns in terms of communication of ethical standards. However, both views have been represented in two distinct items. These items are ‘Communicate clear ethical standards for members’ and ‘Clarify who is responsible for what’.

The classical understanding of ethical leaders was based on a moral manager who strategized and managed ethical values in followers. However, our results show that the relationship of ethical leaders and followers is not transactional but based on concern. Ethical leaders are identified as individuals who focus on the strength of employees, show concern for others, and promote wellness, assistance and bereavement programs for staff and followers. Such behavior marks a shift from the role of traditional leader towards a role which is similar to mentorship and coach. Items associated with such behavior include ‘Show concern for others’ and ‘Encourage wellness and assistance programs’.

Rewards and punishment to generate needed behavior of employees and in return favorable outcomes represent a transactional style of leadership and are an important component of the classical conceptualization of ethical leadership, both in the definition and the survey scales based on it. However, such a mode of disciplining employees can be unhealthy (Covin et al., 1992; Pierce and Gardner, 2004). Our study does not support either point of view; the item ‘Think that rewards and punishments are not useful in the long run’ was not included in the final scale. Based on these findings, it is however not immediately clear whether rewards and punishments are used or considered useful in the short term, and therefore part of the ethical leader’s repertoire. That would require more research.

Several studies in the literature have predicted a link between ethical leadership and relating leadership roles (Treviño et al., 2003; Brown et al., 2005; Van Wart, 2014). Besides the earlier studies of Treviño et al. (2003) and Brown et al. (2005) who initially proposed and acknowledged the links between ethical leadership and

Related leadership styles, contemporary studies by Van Wart (2014) and Shakeel et al. (2020a, b) strongly advocated for ethical leadership to imbibe these styles (cf. Table 1). Our results show that ethical leaders imbibe servant, positive, professionally grounded, CSR and transformational leadership styles. We briefly discuss each of these individual component styles below.

All elements of virtuous leadership (moral person and moral manager) are included in the BELS. This shows the importance of virtuous leadership as a fundamental component of ethical leadership, both in the classical and broader conceptualizations.

Following our broader conceptualization, we added an item about wisdom, as part of virtuous leadership: ‘Exercise sound reasoning in deciding on the optimal course of action’. This was inspired by Van Wart (2014). The analysis confirmed the importance of this item. Wisdom may however also be linked to ethical competency and professionally grounded leadership. More research is therefore needed to disentangle this.

Positive leadership is represented with three distinct attributes including resilience, optimism and transparency. All these attributes play a vital role in ethical leadership. Resilience for instance is the ability to bounce back from adversity (Hartley, 2018, p. 211) which is relevant to ethical dilemmas or ability of leader to help recover an organization from an ethical lapse. Two items on positive leadership are included in the BELS: ‘Am resilient in nature’ and ‘Prefer openness in all situations’. The item on optimism was eliminated because of covariation and the need to reach a parsimonious scale.

Spiritual leadership is characterized by attributes relating to the welfare of employees and diversity. As ethical leadership is expected to result in positive organizational changes including higher motivation of employees and ethical

Leaders are expected to be an advocate against discriminatory policies, these attributes are important to the broader conceptualization. All three items related to this style have been included in the final BELS: ‘Encourage wellness and assistance programs’, ‘Encourage diversity practices’, and ‘Encourage bereavement programs’. These items are in line with the concern for others that was discussed above, and signal a switch of responsibility for ethical leaders from a more managing style (as propagated in the classical studies) towards more serving behavior. Diversity, which remained a key critique on earlier studies, has now also become part of ethical leadership.

Servant leadership is represented by three items in the BELS. Only two items were included, but these are among the most important attributes according to both classical and contemporary authors. These items include ‘Show concern for others’ and ‘Empower others with opportunities so that they develop their skills’. Survey scales based on classical studies also included an item ‘giving voice to employees’ which makes for a plausible connection with servant leadership but was never part of the classical definition.

Out of the three items associated with CSR, two items were included in the BELS: ‘Make investment to create a better life for future generations’ and ‘Respects human rights beyond the legal requirements’. The inclusion of an item pertaining to human rights will make the BELS the only ethical leadership scale that considers human rights as an assessment criterion for ethical leaders. The item about concern for future generations transcends the responsibility of ethical leader to a wider degree than previously considered and fits with the boundary transcending definition of the broader conceptualization.

Out of the three items originally made part of BELS as representing grounded leadership, two items were included: ‘Am committed to lifelong learning’ and ‘Am able to distinguish between competing values’. A connection with grounded leadership

Helps to broaden the view of ethical leadership by introducing ethical competency which we believe is the product of learning, as discussed above.

Classical studies of ethical leadership have already speculated on the interlinkages of transformational and ethical leadership. Shakeel and colleagues incorporated 6 items (positively worded) relating to transformational leadership in the BELS. Four of these have been included after the analysis, which confirms the presence and linkages of transformational leadership as a vital part of broader ethical leadership construct. These items are: ‘Emphasize the collective mission’, ‘Suggest different angles’, ‘Focus on strength of employees’ and ‘Am also responsible for society and the environment of my organization’.

Other items in the BELS connect to characteristics of transformational leadership such as the tendency to prefer openness in all situations and acknowledging and taking responsibility for mistakes and acquiring attributes that can be learned such as the ability to choose between competing values. This shows the competency of the ethical (transformational) leader to steer out of complex ethical dilemmas.

The importance for a more integrating conceptualization and measurement scale is also in accordance with consideration for both morality and ethics (Cooper, 2012). In the case for an ethical leadership construct and measurement scale, the earlier conceptualizations focused on the morality of organizational norms with disregard for any assessment on part of managers (falling short of engagement with ethics), whereas the BELS not only brings an updated code of morality for organizations but also leaves room for consideration on part of leaders to go beyond the rules, if needed, to make decisions based on their principles. This examination of logic of morality, when rules are insufficient, is considered engagement with ethics. It is vital that ethical leadership goes beyond rules since Cooper (2012) asserts that law specifies the moral minimum. This helps leaders discern between competing values instead of staying focused on morality of organizational rules alone. Ethics also keeps the morality of leadership scholarship and practice more fluid, which in turn help leaders interact better with changing contexts and evolving values hence invoking more informed reasoning with decision making. If fostering such an interacting agency is important for ethical leaders, it is equally important to measure it as such.

This study aimed to test the BELS. The results achieved are the first and by no means conclusive steps towards achieving a validated scale. More research and replications are necessary, but the results show that the BELS can help to bridge the gap between classical and contemporary studies. This offers new insights to the existing literature on ethical leadership.

Our results suggest that the ethical leadership contains more sub-values than in the seminal definition. This study has therefore corroborated a broader definition of ethical leadership: “Ethical leadership is the implicit and explicit pursuit of desired ethical behavior for self and followers through efforts governed by rules and principles that advocate learning motivation, healthy optimism and clarity of purpose to uphold the values of empowerment, service to others, concern for human rights, change for betterment and fulfilling duty towards society, future generations, environment and its sustainability” (Shakeel et al., 2020a, b, p. 16).

This study performed the first steps in validating the ethical leadership measurement scale (BELS). Using exploratory factor analysis and AMOS, we found that the one-factor solution with 33 items fitted best. The resultant ethical leadership scale is a comprehensive and parsimonious tool to assess ethical leadership. It includes all contemporary attributes that were identified as lacking in earlier classical conceptualizations and scales, such as empowerment, leader learning, concern for others, but also diversity and advocacy for human rights. Our study has also established links between ethical leadership and other leadership styles, including virtuous, positive, professionally grounded, servant, spiritual, CSR, and transformational leadership. These styles are part of the new construct, which indicates that ethical leadership is not separate from other leadership styles but is in fact an amalgam of ethical components of associated leadership styles.

This study has a few limitations that merit attention. The foremost limitation is the lack of a comprehensive pilot study to assess the tool, although we did make use of expert panel to provide feedback on editorial points and check the structure of items when they were translated to the Dutch language and translated back in English. Other possible limitations include a focus on the public sector and the fact that the findings are limited to one country. The final version of the BELS is an extensive tool that uses 33 items. Some items could perhaps be more relevant to a particular sector than another sector. We used a self-assessment scale and while there are no indications of social desirability bias, it is important that future studies also focus on follower assessment of their leaders. Future researchers can focus on further improving the definition of ethical leadership as the present definition might have been built around only positively valence themes (Alvesson and Einola, 2019; Hughes et al., 2018; Sidani and Rowe, 2018).

This study made use of same dataset to perform CFA, EFA and repeated CFA. Although it may have not provided stronger grounds since it runs the risk of capitalization on chance but presently serves as a good point of direction for future research studies and to progress academic dialog. Upcoming studies can overcome these limitations by aiming for distinct and independent samples or splitting sample at the least. Predictive studies and testing nomological networks can also be used to achieve a more robust scale.

We invite scholars to replicate the BELS, preferably in different sectors, countries and cultures. As the broader construct includes elements which are equally linked to all sectors and disciplines, it is possible to use it for political leadership to study different narratives and their ethical denominators including populist leadership (cf. Mirza and Azhar, 2021) and other political leadership types. Similarly other less focused public sector institutions like Judiciary, where popularity for celebrity judges is rising (Starr, 2023), merits equal attention. Since there is a greater tendency for shorter scales in literature (Yukl et al., 2013), hence the length of BELS can be further explored by validating its accuracy for different sectors and disciplines. Increasingly, leaders turn to social media to connect with their followers, this can serve as a viable channel for ethical behavioral analysis of leaders using BELS since self-assessments are prone to biases and contemporary methods like video analysis and social media posts may help study leader’s behavior in a natural setting. It should also be adapted to other languages to verify its authenticity.

The appendix displays all items of Bels with selected items denoted in Italics by an asterisk. It shows standardized factor loadings of 35 and 33 items scale.

Table A1