**4. Discussion**

In summary, we found that moral leadership approaches were discussed quite a lot when compared to normalization responses. Twice as many quotes (350 vs. 171) illustrated moral leadership tactics in response to the dirty leadership images. When reflecting on moral leadership approaches, juniors and managers emphasized different options. Managers mentioned their active compassion and support of juniors twice as often as juniors, whereas juniors mentioned the approachability of their managers more often as important to them. Still, juniors and managers referred to the same kind of tactics. This, together with the ample groundedness of the codes, indicates good saturation.

Table 5 shows that the leadership style of managers is more often constructed as dirty (246 quotes) than the resulting effects (37 quotes). However, the number of quotes is not very conclusive regarding this dirtiness as something "essential". The mentioned effects like burnout are really problematic and not mentioning it might even be an avoidance or denial strategy. Still, there is a lot of talk about the dirtiness of the leadership style and this is a big issue in the construction of consultants' dirty leadership.

Moral leadership aims at neutralizing such dirtiness. For instance, the impressions of an output-oriented and pressuring leadership style resulting in long working hours and high workloads is countered by traditional moral leadership tactics. They counter the stigma that managers do not provide support and have no empathy, as these traditional tactics are targeting exactly these aspects of dirtiness in the leadership style. Institutionalized moral leadership tactics, like regular performance talks, training, selective hiring and various compensation tactics, are also discussed a lot, which indicates their social visibility, with the construction of a better leadership impression. These institutionalized approaches seem specific for our high-performance setting, and might be relevant to other high-pressure contexts as well, such as investment banks or law firms. The applicability of the individual moral leadership tactics must be much wider, as these tactics have been found in many other work contexts already [29,30,49,50].

The findings of our study contribute to the literature on moral taint and moral leadership. First, we found that moral problems that cause a dirty leadership image can be targeted with moral

leadership approaches [29,30,49]. Consulting managers engage in many forms of moral leadership to counterbalance what juniors and managers socially construct as a dirty leadership style. Especially the pressuring leadership style, like focusing on output only and lack of personal attention, is experienced as dirty in the image of consultant management. Traditional moral leadership tactics target such dirty aspects in the leadership style directly. Thus, they socially construct alternative and more ambiguous leadership image: pressuring yes, but also committed. As deeds can speak louder than words, this approach seems quite relevant for influencing a dirty leadership image. Except for confronting public opinion, normalizing mainly has a focus on individual stress reduction due to a dirty image. Our findings add a new repertoire of tactics to the literature on moral taint (cf. [3,24,26,35]). As a consequence, normalization tactics are just one way to respond to the experience of moral taint.

On a more critical note, we found that some of the propositions in Ashforth and Kreiner [31] did not hold very well in the context of consulting. For instance, proposition 4 that morally tainted professions are assumed to mostly engage in group-level defensive tactics, and proposition 10 that such professions create social buffers to normalize taint were not very prominent in our context. We found, similar to findings by Vaast and Levina [3], that the tactic of social buffering is not utilized much, whereas defensive tactics are most frequently applied. Considering defensive tactics Ashforth and Kreiner [31] assume in their proposition 11 that condemning condemners is the most common one to normalize moral taint, however, in our study, social comparison is the most reported normalizing tactic, a result that aligns by and large with the findings of Vaast and Levina [3] situated in the banking industry. Unlike the propositions in Ashforth and Kreiner [31], we found many instances of confronting public opinions about the dirtiness of consultant leadership. Probably, effectively normalizing moral taint largely depends on context, as both banking and consulting belong to the professional service sector. As professional service industries are a growing field of employment in today's societies, future research should address these high-performance sectors more specifically when studying morally tainted leadership and the potentially tainted health consequences for employees, like burnout. It is important to study such consequences, and possible gender, role or seniority differences based on a quantitative research design, to better tease out to what extent normalization attempts cover up such consequences.

Second, we contribute to moral leadership literature by finding that consultant managers not only apply traditional moral leadership approaches like being approachable, compassionate supportive and encouraging. Consultancies also support managers with institutionalized measures that aim at moral leadership, and that are new in this literature (cf. [29,30,49,50]). As the agency of managers is constrained by institutional pressures, we see that the organization also creates counter pressures with several specific HR practices. As these HR practices are quite visible, they help to clean up the dirty leadership image, again with deeds more than words. Therefore, we invite moral leadership literature to better include the organizational and institutional levels in its theorizing. By having frequent performance reviews planned, trainings available, and several non-monetary compensation policies in place, consultant managers mentioned how they are supported in taking responsibility for the wellbeing of juniors. These institutionalized measures might be specific to high-performance occupations with demanding top management, clients and projects, where it is very tempting to satisfy client needs first and think about your juniors second [11]. Therefore, similar institutional support might be relevant to other high-pressure professional service contexts, such as investment banking, law firms and marketing agencies, or even at the more competitive top universities.

Our research has also some practical implications beyond the fields of moral leadership theory and dirty work literature. We consider it a promising and innovative combination to facilitate moral leadership of middle managers with supportive HR practices. Such tailored institutions might help protecting employee health and reduce psychosocial risks at work, especially in settings of knowledge-intensive work. These are high-performance work contexts where employees are often ambitious and willing to give their best, but where burnout risks lure around the corner, especially when employees do not feel they are seen and cared for, or rewarded for taking initiative and for acting

responsible towards their organization, often at the cost of their own mental resources. Our research shows that HR institutions can be further developed to support middle management in taking care of their employees, inspired by moral leadership ideas. If organizations create such institutions to give attention, to show compassion and to take responsibility for workers that give their best, it could help to move away from the more bureaucratic, rule-based and one-size-fits-all HR institutions we are so familiar with today.
