The requirement for a leader to be a charismatic superman was one of the mythologies debunked in a major survey on qualities needed at the top. Beverly Alimo-Metcalfe and Robert Alban-Metcalfe report

We are seeing major initiatives on developing leadership in the NHS including those related to clinical governance leadership, primary care group leadership, and of course, the chief executives' programme.

The NHS Executive's paper Leadership for Health: the health authority role (NHSE 1999) extols the virtues, and common sense, of adopting a transformational leadership approach to achieving a new NHS. We need to know the behaviours, skills, qualities, and attitudes that reflect this form of leadership.

In 1997 we were commissioned by the local government management board to undertake a two-year exploratory investigation into the nature of leadership in local government. The Nuffield Institute for Health at Leeds University was keen to co-sponsor the project, and in parallel to conduct an investigation of its own on the NHS. The end-point of the project was to be the creation of a new 360-degree feedback instrument for measuring the dimensions of leadership identified in the research.

Unlike previous, US studies of leadership, we designed a project which would start with no pre-determined notion of what it meant.

It was evident that as this was an exploratory study, a qualitative methodology was essential for the first stage. The technique we selected was the repertory grid interviewing process. We decided to interview staff in a dozen organisations in both the NHS and local government. In each case, at least six of the organisations were selected randomly from a national list, and four were organisations which had a reputation of being cutting edge in their respective sector.

We spent a day in each of these organisations conducting one-hour interviews with the chief executive, a director, an assistant director, and a middle-level manager. We asked that both sexes be represented, and that the group included at least one person from a minority ethnic group. A final sample of 92 individuals was interviewed. The table opposite shows the breakdown of managers by sector, level, and gender.

The process of eliciting constructs in the interview comprised asking each interviewee to identify (using initials) on a piece of paper two individuals with whom they currently worked, or had worked in the NHS, who they believed had outstanding leadership qualities. They were also asked to identify two individuals who were average, and a further two who were poor with respect to their leadership. These individuals are referred to as the 'elements'. Finally, a seventh element of 'self 'is added to the group. The pieces of paper were shuffled, and three were randomly selected. The interviewee was asked to think about the three individuals identified on the pieces of paper, and consider in what way two of them were similar to each other, but different from the third, with respect to their leadership qualities, or lack of leadership. What the person said was recorded as their 'construct', or perception, of leadership.

For example, the individual might say: 'Two of these people are outstanding at developing the confidence of their staff, whereas the third person is totally self-interested. 'The perception 'develops staff 'versus its opposite pole, 'self-interested', provides us with a construct this particular person has of one aspect of leadership.

The interview continued, using different combinations of elements.

Typically, we elicited around 15-20 major constructs from each person interviewed.

To ensure that we had obtained the constructs that doctors held of leadership, we also gathered data from six focus groups of doctors, in which a total of 51 male and female doctors participated.

Content analysis of the 2,000+ constructs, conducted independently by the two authors, led to the emergence of 48 'themes', which formed the basis of items for a pilot leadership instrument.

This list was augmented by a small number of items based on a review of the leadership literature. No significant differences were detected between the range of constructs from NHS and local government managers, nor between males and females, nor between managers at different levels in the organisation.

The range of constructs held of leadership was complex. But as fascinating as this data appeared to be, it would have been impossible to generalise from the views of around 150 individuals, of whom around 100 were staff in the NHS, to the wider NHS and local government.

So an 171-item questionnaire was distributed, initially to a sample of 100 managers, who were asked to comment on the clarity and relevance of the questions, and then to two samples of managers, one in the NHS, the other in local government. Participants were asked to rate their current manager, or a previous manager with whom they had worked, on a sixpoint scale from 'Strongly agree' to 'Strongly disagree'; two additional responses were available: 'Don't know'and 'Not relevant'.

In the NHS, around 300 organisations were selected at random and a letter explaining the research was sent to the chief executive, with a request that they return a slip offering their support. Packs of 45 to 60 questionnaires, depending on size, were distributed to trusts. An explanatory letter, and stamped addressed envelope, was included with each questionnaire. The request was again that the manager receiving the questionnaire anonymously rated their current boss, or a previous boss with whom they had worked in the NHS.

By mid-1999, after a second wave of requests to both organisations, a total of 1,464 usable responses were received from local government, and 2,013 from the NHS. The results were analysed separately for the two public sector organisations.

Among the NHS managers, six factors emerged in the research model, and a further eight factors emerged in a second additional analysis of the data. These were turned into 14 robust leadership scales, with high internal reliability co-efficients, each of which measures a different aspect of leadership. These dimensions will form the basis of a 360-degree feedback instrument, 'The transformational leadership questionnaire copyright (NHS version)', to be launched this month.

The most obvious implication of these findings is the staggering complexity of the role of leadership in the NHS. Another lesson is that the transactional competencies of management, while crucial, are simply not sufficient on their own.

What is clear is that existing US models of leadership do not encapsulate this complexity. Typically, they place an overwhelming emphasis on charisma and vision, that is, on the leader as primarily acting as the role model for his/her followers. Is this the product of adopting research methodologies which focus solely on the views of top managers, and/or on those researchers who develop models from their observations?

In contrast, the results which emerge from asking the recipients and ultimate arbitrators of leadership effectiveness - namely, the staff who work in the NHS - how they perceive leadership, presents a very different model.

They are clearly stating that the most important prerequisite role for the leader is what they can do for their staff. This is far more reminiscent of the model of leader as servant. But they are not simply stating that leadership is about meeting staff 's needs; it is much more than that.

The 2,000 staff who participated in this research project are also saying that leadership is fundamentally about engaging others as partners in developing and achieving the shared vision and enabling us to lead. It is also about creating a fertile, supportive environment for creative thinking, for challenging assumptions about how healthcare should be delivered.

And it is about much closer sensitivity to the needs of a range of internal and external stakeholders, inside and outside healthcare. It is about 'connectedness'- joined-up thinking, even.

Another very positive feature of the findings is that the model significantly reflects aspects of the modernisation agenda, including partnership working, valuing staff, aiming for best practice, removing the traditional barriers between agencies working together within the community.

It would appear that there is a high degree of congruency between what those who work in the NHS believe to be leadership and the espoused leadership tenets of the centre. This can only be very good news.

What needs to be heeded, however, is the nature, and degree, of support from the centre to enable this to happen locally.

To make most effective use of the findings, to make a significant impact, it will require NHS organisations to develop a transformational culture which reflects the dimensions of leadership emerging from this research.

US scholar Ed Schein believes that probably the most important responsibility of leadership is to create the most appropriate culture for the organisation. We also know from substantial research evidence that there is a highly significant relationship between the leadership style adopted by the top manager and the culture of the organisation.

Here we face, perhaps, the greatest challenge for leadership reform in the NHS, since it is highly likely that the current top and senior managers were not selected into these posts on the basis of possessing a transformational style of leadership. What, then, will happen if the model is not adopted by the top NHS managers? Clearly the answer is that there will be no cultural infrastructure supporting, and nurturing its growth. The noble attempts of managers further down the organisation to practice a transformational approach will be blocked, or even 'punished', by the influence, and perhaps game-playing of more senior managers.

Will managers who adopt a predominantly controlling, or laissez-faire style see the need to change? It is likely that these very managers will lack the insight or willingness and commitment to develop transformational leadership. Who holds the responsibility to deal with these powerful blocks to progress? This is a key consideration for the NHS centre, since the government's modernisation agenda will not be achieved unless the NHS recognises that cultures of blame, authoritarianism, narrow-mindedness and reckless disregard for staff are not to be tolerated. Can those managers who wish to strengthen their transformational leadership change? This oft-asked question is still a matter of some debate among organisational psychologists. However, the findings from recent research are heartening.

There is evidence that the use of 360-degree feedback, plus tailored individual support, can lead to sustained positive changes in leadership style.

1But, like quitting smoking, the all-important prerequisite is the managers' willingness and openness to change.

Another vitally important aspect of the findings is that the model of transformational leadership which emerged does not simply apply to managers, it applies equally to other NHS professionals, including clinicians.

Leadership, like communication skills, must be incorporated into doctors' appraisals (both developmental and summative), and for those of other staff, including the posts of primary care trust chief executives, and chairs.

NHS organisations can no longer rely on the transactional competencies of management to run the NHS. They must adopt, and be truly committed to incorporating these dimensions in their recruitment, selection, promotion, performance management, and development processes, for all staff whatever the area of work or level.

The human resources professionals will have to think creatively about how these qualities can be measured, and developed, alongside the transactional competencies of management. It will also require organisations to consider whom they use as judges/assessors in such situations. Why should this be important?

The reason is that the chances are very high that those who will occupy the gate-keeping role of assessors have been promoted to those positions on the basis of a track record on the transactional competencies. Given the difficulty HR professionals often say they have in influencing these decisions, it will be beholden on the centre to advise NHS organisations of the importance of ensuring that best practice is exercised here.

As the findings show the local government and NHS models of leadership are virtually identical, it is time to seriously consider joint development - and perhaps recruitment - since inter-agency working can be strengthened considerably by this approach.

Finally, this study has served to debunk and demythologise the 'heroic' image of leadership which has been so prevalent. The model which emerges here is one of down-to-earth decency, humanity, humility, sensitivity and respect for others, but this is no soft option.

It is fired with a passionate commitment to living the values, to engaging all in sculpting the vision, to creating an environment where challenging, questioning, and turning mistakes into catalysts for learning, are regarded as the norm.

The notion of connection and partnership is a backbone theme of this UK model, and appears to be driving a very new image of leadership for the new millennium.