Has presidential leadership become harder? The question itself raises numerous, and fundamental, boundary-defining questions. For example, how much time must elapse before we can say that a change is meaningful? The question clearly implies a change of condition. What changed? Why did it change? When did it change? The "what changed" question includes the notion of a change in outcomes or success and forces us to focus on the dependent variable, in this case, "what are the outcomes" presumably desired by the president, and what, therefore, is the president's level of success in achieving them? The "why did it change" question leads us to focus on independent variables and the structure of explanation. If presidential leadership has become harder, why? What are some of the possible lines of inquiry? The "when did it change" question implies not only that something did change (a matter open for debate) but also invites us to investigate the form of the change--linear? curvilinear? logarithmic? punctuated? and so forth. Has presidential leadership become harder is an easy question to ask but an extremely difficult one to answer.
It is safe to say that we will not answer those questions here. However, we need to put a searchlight into this thick fog to illuminate essential questions, a precondition to obtaining appropriate answers. We first note the problem of time--not as a philosophical problem but as a way of thinking about what we are trying to compare over what span of time and whether the comparison ought to take cyclical or chronological functions. Then, our next and, also, later our last stop is to try to clarify the "what" question; that is, just what is the dependent variable we should be focusing on when we talk about presidential leadership? Thus, we begin by asking what basic alternative models of presidential leadership are articulated--and, implicitly, therefore, how the role of the president in the political system is conceptualized. To what extent are the problems we detect regarding presidential leadership related to our choice of leadership model, and, consequently, how do we conceptualize the role the president plays in the American political system? Later in the article we come back to look at some standard measures of success and also some nonstandard ways of thinking about presidential success.
We look at some of the explanations offered for changes in the circumstances of presidential leadership, some of which (increased party cohesion, for example) were once proposed to strengthen the hand of the president in governing. We look both at broadscale notions of changing conditions, for example, a transformation of elite culture, and some more down-to-earth and precise ones, such as divided government. Again, we arrive at no definitive conclusion but do note the possibility of the compounded (interactive or multiplicative) effects of several key variables, for example, divided government, greater party cohesion, and greater average distance between party positions.
Always in the background of this exploration is the "when" question. When was some condition of presidential leadership said to have changed and in what form? If a function is nonlinear and also part of a longer cycle, a shorter time frame is likely to produce a falsely positive linear function. We are often tempted to compare the present unfavorably with a more distant past that we know only dimly, if at all (Ellis 1990). We often have gripes about the present (or at least concerns, even amid the balmy economic climate of the day) because we are living through it; we know the Republic survived the past, however, and being less aware of past blemishes by virtue of not directly experiencing them or being youthfully oblivious at the time, we tend to focus less on past problems.
At the end of this article, we again come back to the dependent variable, that is, success, and focus on the extent to which presidents have sought (and successfully implemented) new means of defining success in an environment of almost continuous divided government. …