Presidency scholars have long debated whether a president's personal traits or the historical circumstances of their presidency have a bigger effect on presidential performance. Stephen Skowronek (1993, 2008) argues that the political conditions a president faces, what he calls "political time," largely defines the president's opportunities and challenges. In contrast, others argue that presidents' personal traits (e.g., personality, political skill, vision, communication skills) profoundly shape presidential behavior (e.g., Barber 1972; Greenstein 2009; Renshon 1996; Wayne 2012).
We contribute to this debate by examining whether historical conditions (political time) and two important personal traits (brilliance and character) relate to president's standing in polls of "presidential greatness." We find that both historical conditions and personal traits significantly predict historical evaluations of presidents. We provide novel evidence that presidential character affects presidential greatness. Although many have found that scandals tend to lower presidents' standing in the historical polls (e.g., Cohen 2003; Nichols 2012; Simonton 1987, 1991, 1992, 1993, 2001, 2006), we show that more nuanced differences in character predict historical judgments. Although many have claimed that character is central to effective presidential leadership (e.g., Gergen 2000; Renshon 1996; Wayne 2012; Wilson 1995), Wayne (2012, 3) argues that "unfortunately, the study of character is one of the least analyzed aspects of presidential behavior," partly because such study often precludes the use of quantitative analyses favored by most political scientists. Our analysis is far from a complete statement on the relationship between character and presidential behavior, but it provides some systematic evidence supporting the claim that character (or at least perceptions of it) matters for the presidency. At the same time, the historical context in which presidents serve powerfully shapes their current standing. Our results provide insight into one of the field's core debates and contribute to the literature examining greatness ratings (in addition to Simoton's works cited above, see, e.g., Balz 2010; Cohen 2003; Curry and Morris 2010; McCann 1992, 1995, 2005; Nichols 2012; Simon and Uscinski 2012). These ratings provide one measure of presidents' legacies. Because "speculation about how presidential actions will be viewed by future generations weighs heavily on the minds of chief executives," knowing what predicts historical views of greatness may affect current presidential behavior (Panagaopoulos 2012, 719).
In the next section, we review the literature on presidential greatness polls and explain why character, brilliance, and political time may shape presidents' scores in these polls. We then describe our data and methods, both of which build strongly on previous work. We present our results and then draw conclusions.
The Roots of Greatness
Presidency scholars, indeed students of leadership in general, have long tried to identify what makes a great leader. (1) Most of these studies have focused on biographical and historical materials, the best relying on careful qualitative analysis to marshal detailed, nuanced accounts of the great and not so great presidents (e.g., Barber 1972; Landy and Milkis 2001; Felzenberg 2008). Other studies examined the collected evaluations of various groups of "experts" asked to assess the presidents in different ways. Since Arthur Schlesinger (1948) first polled 55 experts (mostly historians), several individuals and organizations have followed suit, asking different collections of experts to evaluate presidents.
Scholars have long discussed, critiqued, and analyzed these ratings. Critics argue that the ratings are biased by the raters' ideology or partisanship (Felzenberg 2008; Lindgren and Calabresi 2000; Uscinski and Simon 2011; but see Bose 2003, 8-9, and Murray and Blessing 1994), that the ratings tell us more about the raters than presidents, and that the ratings are generally "not very rigorous" (Pfiffner 2003, 23; see also Felzenberg 2008). However, systematic analysis of these ratings has concluded that they "may tell us more than critics admit" (Nichols 2012, 272). Indeed, Dean Keith Simonton (2001, 294), who has analyzed presidential greatness polls for over two decades, concludes "there is a strong prima facie case that these greatness assessments reveal how U.S. presidents varied in their effectiveness as the nation's highest political leader." Thus, we take the ratings as reflective of two things: the raters' sense of how we should evaluate presidents and "consistent and unbiased measures for a somewhat nebulous concept," namely, presidential greatness (Curry and Morris 2010, 522). (2)
The study of these ratings has been dominated by Simonton's work, which has culminated in what is now known as the Simonton model. This model is the starting point for recent research on presidential greatness (e.g., Cohen 2003; Curry and Morris 2010; Nichols 2012; Simon and Uscinski 2012) and consists of six variables: (1) number of years in office, (2) whether the president was a war hero, (3) whether the president was embroiled in a significant scandal, (4) whether the president was assassinated, (5) the number of years the country was at war during the president's tenure, and (6) the president's intellectual brilliance.
Although the Simonton model has held up well, one recent study argues that taking Skowronek's theory of political time into account eliminates the importance of intellectual brilliance in presidential greatness (Nichols 2012). However, there are good reasons to believe brilliance, along with another personal attribute, character, should consistently affect presidents' standing in greatness polls. Greenstein (2009, 5) argues that "six qualities ... relate to presidential job performance." Several of them relate to intellectual brilliance. Simonton (2006, 511) argues that intellectual brilliance "is closely associated with the cognitive complexity necessary for meeting the demands of modern life." Greenstein (2009, 6) argues that successful presidents must adopt a "cognitive style" that allows them to "process the Niagara of advice and information that comes [their] way." Intellectual brilliance may well lead a president to an effective cognitive style. In addition, Simonton (2006, 512) argues that intellectual brilliance "is associated with other advantageous attributes, such as charisma and creativity," which, for presidents, seems likely to translate into successful political communication, organizational capacity, political skill, and vision, four more of Greenstein's six qualities.
In addition, character may be related to presidential greatness. Character is a notoriously difficult concept to define. Felzenberg (2008, 11) defines character as integrity, honesty, and courage. In his book, The Character Factor, Pfiffner (2004) argues that a commonsense definition of character encompasses traits like trustworthiness, integrity, reliability, loyalty, compassion, self-restraint, consistency, and prudence. However, he notes that discussions of presidential character often employ different definitions of the term (see Pfiffner 2004, 172n.22). As we discuss below, our measurement of presidential character is inherently nebulous, consisting of experts' assessments of presidents' "integrity," "character," and "moral authority." We do not know what exactly the experts have in mind when they use these terms, but as we will see, the experts tend to exhibit a fair bit of consistency when they evaluate presidents along these lines, suggesting that there is a strong systematic component to the way experts think about character.
Character may shape greatness ratings for two main reasons. First, public expectations for great presidents include strong character. In his study on public expectations of the president, Edwards (1983, 189-91) found that the public had "high expectations for the president's official performance, but also had lofty expectations for his private behavior." Textbooks and the standard readings on the presidency that inform much of the public's learning about the presidency tend to exaggerate the president's powers and virtues, leading to idealized views of what presidents should be like: exhibiting "qualities [of] honesty, knowledge, and open mindedness along with not being power hungry, unstable, or weak" (Simon 2009, 140; see also Cronin 1975, 1980). Waterman, Jenkins-Smith, and Silva (1999, 949) report on two surveys that asked respondents to rank how important "high ethical standards" are for being an excellent president. On a 0 to 10 scale, ratings averaged about 8.5. When people evaluate presidents, their views of the president's character consistently influence those evaluations. Views of the president's integrity and competence consistently shape the mass public's evaluations of sitting presidents (Greene 2001; Newman 2003, 2004; see also Goren 2002). Along the same lines, when evaluating presidential candidates, the public relies on its view of candidates' character (Rahn et al. 1990; Funk 1999). If the public generally expects presidents to have strong character and rewards and punishes current presidents on the basis of their character, experts may evaluate historical presidencies the same way.
Second, many argue that character shapes presidential behavior itself. Stanley Renshon (1996, 184) describes character as "the basic foundation upon which personality structures develop and operate." He continues, "character shapes beliefs, information processing, and, ultimately, styles of behavior" (Renshon, 1996, 184). Moreover, Wayne (2012, 1) argues simply that "character affects what presidents say and do and how they relate to others. It is a dimension of any behavioral explanation." He continues, "knowing a president's character provides a guide to...