Effective leadership across economic contexts

We use a laboratory experiment to study how leaders affect workers' productivity across economic incentive contexts. In four-person groups, three group members work on a production task, with a fourth member potentially serving as a leader. We vary the economic context by changing how worker pay is determined as a function of worker outputs, comparing Revenue Sharing, Weak Link or Tournament incentives while holding constant the activity performed by workers and the incentives for leaders. A second treatment varies whether groups have Active Leaders who can exert influence through messages to workers or Passive Supervisors who exert no influence. The average effect of having an Active Leader on group output is large only under Weak Link incentives. Across all incentive contexts, we find a positive correlation between the productivity increase in output produced by an Active Leader and independent ratings of leader quality based on measures from leadership research. The nature of leaders' communication varies across incentive contexts, with comparisons between workers most common under Tournament incentives and messages about group earnings, which speak to social considerations, most common with Weak Link incentives.