Daniel Goleman's 2000 Harvard Business Review article on leadership styles remains one of the most widely read pieces of management research ever published. Working with researchers from the Hay Group, Goleman identified six distinct leadership styles and measured their impact on organisational climate and financial results. The findings were striking: leadership style accounted for 30% of variance in business results.
That research is now 25 years old. The core findings have been substantially replicated and extended. What has changed is the context in which leaders operate. In an AI-augmented, rapidly changing, high-uncertainty environment, the question of which leadership styles are most effective has become more urgent and more nuanced than ever.
The Six Styles
Goleman's six styles are worth understanding both individually and in combination, because the most effective leaders use different styles in different situations rather than defaulting to a single approach.
The Coercive style demands compliance. "Do what I say." It is characterised by tight control, short-term focus, and close supervision. Goleman's research found that it has the most negative impact on organisational climate of any leadership style. It kills initiative, creativity, and intrinsic motivation. The only situation where it is appropriate is a genuine crisis requiring immediate coordinated action, or a turnaround situation where established patterns must be broken quickly.
The Authoritative style mobilises people toward a vision. "Come with me." It combines clarity of direction with genuine confidence that people can get there. Goleman's research found this to be the most broadly effective leadership style, producing the most positive climate impact across the widest range of situations. It works because it combines purpose (where we are going and why) with autonomy (how you get there is largely up to you).
The Affiliative style creates harmony and builds relationships. "People first." It prioritises emotional connection, encouragement, and team wellbeing. It is highly effective for healing conflict, building trust after a difficult period, or motivating during high-stress transitions. Its limitation is that it tends to avoid difficult performance conversations, which means underperformance can go unaddressed.
The Democratic style builds consensus through participation. "What do you think?" It slows decision-making but produces higher commitment from the team and surfaces insights that a directive leader would miss. It is particularly effective when the leader genuinely does not have the best answer and the team's combined knowledge and perspective will produce a better decision. It becomes counterproductive when the situation requires speed or when the team lacks the expertise to contribute meaningfully.
The Pacesetting style sets high standards and expects them to be met. "Do as I do, now." This is the style of many high-performing individual contributors who become managers. It produces impressive short-term results with high-capability teams who are already intrinsically motivated. It creates anxiety, burnout, and disengagement in teams that are not already performing at that level. It is one of the styles most likely to be overused by high-performing managers who have not made the transition to leading through others.
The Coaching style focuses on long-term personal development. "Try this." It is characterised by genuine interest in each person's development, willingness to let people try and fail in order to learn, and focus on building capability over time rather than immediate task performance. Goleman's research found it to be one of the most rarely used styles despite having a consistently positive climate impact. The primary reason is time: coaching takes longer than directing, and most managers feel they do not have it.
What the Research Says About Context
The most important finding from Goleman's research is not which style is best but that the most effective leaders use multiple styles fluidly, choosing the right approach for the specific person, situation, and objective. Leaders who rely on a single style, regardless of which one, are less effective than those who adapt.
This insight has been extensively supported by subsequent research. The Centre for Creative Leadership's work on leadership agility consistently finds that the ability to read a situation and adapt one's approach is a stronger predictor of leadership effectiveness than any single leadership style.
Research by Edmondson (2023) found that in high-uncertainty environments, the combination of psychological safety (affiliative and coaching elements) with high performance standards (authoritative and pacesetting elements) produces the conditions most associated with team innovation and resilience. Neither safety alone nor standards alone is sufficient.
Leadership Style in an AI Age
The leadership context has shifted significantly in ways that affect which styles are most effective and when. Three shifts are particularly important.
The shift from information asymmetry to information parity. Leaders used to be effective partly because they had more information than their teams. AI changes this: a team member with good AI tools may have access to more relevant information on a specific question than their manager does. The authoritative and pacesetting styles, which depend partly on the leader's superior knowledge and judgment, become less reliable as defaults.
The shift from stability to continuous change. In a rapidly changing environment, the coaching style becomes more valuable because the ability to learn continuously is more important than any fixed set of skills. Leaders who develop their team members' learning capability are building assets that compound over time. Leaders who direct and control are creating dependencies that become liabilities as contexts shift.
The shift from intrinsic to psychological safety as the foundation. Research by Bao et al. (2025) found that AI adoption can reduce team cohesion and psychological safety when not managed carefully. Leaders whose default style includes affiliative and coaching elements are better positioned to maintain the team conditions that support effective AI adoption.
Building Style Flexibility
Leadership style flexibility is a learnable capability, not a fixed personality trait. The most effective development approach combines self-awareness work (understanding one's default style and its impact), structured feedback (understanding how the team experiences one's leadership approach), and deliberate practice in styles that do not come naturally.
For most managers, the styles most worth developing are coaching (because it is underused despite its positive impact) and authoritative (because it combines direction with autonomy in a way that sustains both engagement and performance). The coercive and pacesetting styles are typically overdeveloped in high-achieving managers and benefit from conscious restraint.
Try This
Reflect on your leadership conversations over the last week. Which style were you using most frequently? Which style were you using least? Think of one specific upcoming situation where your default style is not likely to be the most effective, and plan deliberately to use a different approach. Notice what happens to the conversation.
References
Bao, Y. et al. (2025) 'The impact of AI adoption on employee well-being', Nature Human Behaviour, 9(2), pp. 312--324.
Centre for Creative Leadership (2023) Leadership Agility: The Complete Framework. Greensboro, NC: CCL Press.
Edmondson, A.C. (2023) Right Kind of Wrong: The Science of Failing Well. London: Atria/One Signal Publishers.
Goleman, D. (2000) 'Leadership that gets results', Harvard Business Review, 78(2), pp. 78--90.
Hay Group (2000) Research into Leadership Effectiveness and Organizational Climate. Boston, MA: Hay Group.