Historians often note that Truman moved left in 1947–48. Somehow, they say, the joy of combat with the “do-nothing” 80th Congress liberated him to “become himself.” Historians also comment on the surprising way Eisenhower moved right in 1959–60. Contemporaries noted how Reagan lost his “flawless touch” for the deft legislative compromise after 1986, seemed to become more rigid, and lost more battles with Congress (New York Times, November 7, 1987:33; quoted in Nathan 1990). Finally, after the Republicans seized control of Congress in 1994, political commentators observed that President Bill Clinton emerged from a period of drift. But unlike Truman in 1947, Clinton seemed to move right rather than left. Are these incidents unrelated accidents of history? Or do they show the same causal mechanism at work?

Using a Model as an Interpretive Framework

Interpreting history is central to political science. In presidential studies, historical interpretation often takes the form of a narrative illustrating how presidents use power, manage the transition between administrations, handle crises, make appointments, and so on. When the narratives work as political science – and some work brilliantly – the selection of facts and their presentation flow almost inexorably from an underlying causal mechanism. An implicit model supplies the framework for interpreting history. It must be so, else the narrative remains only a chronology. But even though it is the model that scripts the drama, directs the actors, and moves the scenery, it rarely appears on stage. It remains behind the scenes, usually unexpressed and invariably informal. Concealing the model carries a cost.