The paths of glory : structure, selection, and leaders

Abstract

Practitioners, journalists, and historians believe that leaders play a crucial role in determining events. Social science theories, however, generally argue that leaders are unimportant. This dissertation proposes a two-stage theory of leader impact called Leader Filtration Theory (LFT). In the first stage, leaders are usually chosen from among a pool of candidates by a filtration process that homogenizes the pool. This makes actual leaders highly similar to those who almost got the job, rendering individuals fungible. Sometimes filtration is bypassed. When this occurs, leaders can gain power who are very different from potential alternates. In the second stage, leaders face constraints from within their organization and outside it. If these constraints are sufficiently weak an unfiltered leader can have a very large impact on outcomes. Such leaders are likely to display a high degree of variance in their performance. The theory is tested quantitatively using historians' rankings of U.S. Presidents. Unfiltered leaders should be disproportionately represented among the best and worst Presidents. LFT's prediction of higher variance in performance is very strongly supported by this test. The theory is also tested by examining three American Presidents and two British Prime Ministers. Jefferson, a Filtered President, made decisions with regards to the Louisiana Purchase very similar to those that alternative Presidents would have made. Lincoln and Wilson, by contrast, both Unfiltered Presidents, made decisions that were radically different from those that would have been made by alternative Presidents. Chamberlain, a Filtered Prime Minister (PM), made decisions with regards to appeasement of Germany before the Second World War very similar to those of alternative PMs. When his preferences diverged from those of potential alternate PMs, British policy followed the alternates' preferences, not his. When Churchill, an Unfiltered PM, gained office, his unique preferences determined British policy. The theory should have power outside democratic politics as well. Eight leaders from other domains - dictatorships, the military, business, and scientific research - who are commonly believed to have been highly consequential are examined. All eight would have been classified as Unfiltered by LFT. Finally, five approaches to choosing better leaders suggested by LFT are described.