Principal agent theory implies that legislators will delegate power to a leader only when they need the leader's help and the leader can be expected to provide satisfactory help if granted power. This study is the first to evaluate the implied interaction between legislators' need for help and the degree to which legislators and leaders have similar preferences. By analyzing the Speaker's powers in the U.S. states, I arrived at three key conclusions. First, institutional leadership power responds to the interaction between preference alignment and policymaking challenges. Traditionally expected effects only appear when both alignment and challenges are relatively high. Second, professionalization causes weaker leadership powers. Finally, electoral competition correlates with stronger appointment, committee, and resource powers, but weaker procedural powers.