Personality Traits, Candidate Emergence, and Political Ambition: How Personality Affects Who Represents Us *

Transcription

1 Personality Traits, Candidate Emergence, and Political Ambition: How Personality Affects Who Represents Us * Adam M. Dynes Brigham Young University adamdynes@byu.edu Hans J.G. Hassell Cornell College hhassell@cornellcollege.edu Matthew R. Miles Brigham Young University-Idaho milesma@byui.edu Abstract Recent work has shown a strong influence of personality on the political ideologies, opinions, and involvement in politics. This work, however, has focused on individuals outside of government. Yet, personality should not become irrelevant to behavior once individuals enter government. In this paper we examine the associations between personality traits and nascent political and progressive ambition. Using two surveys, one a national representative survey and the other a survey of public officials, we find a strong influence of personality traits, as measured by the Five-factor Model, on nascent political ambition and progressive ambition. We find specifically that individuals with extraversion and openness are more likely to consider running for office. We also find that personality plays a similar role in the progressive ambition of individuals already serving in public office. We argue that democratic elections attract certain types of individuals to seek public office and become involved in government. Word Count: 9,223 Abstract Word Count: 149 * Paper prepared for presentation at the 2016 Annual Meeting of the American Political Science Association, Philadelphia, PA

2 1 Who we are affects how we act. Recent scholarship has shown that the heuristics that people use and the cognitive biases they are susceptible to are influenced by a number of factors, most notably age, gender, and personality traits (Kahneman, 2011; Sheffer et al. 2016). In political science, personality has been shown to influence an individual s opinions and political actions (e.g. Gerber et al. 2013; Gerber et al. 2011; Denny and Doyle 2008). While we know that personality has a strong effect on a number of political behaviors including ideology, public opinion, and political participation, this work has focused primarily on the influence of personality on the decisions and actions of the general public who have a much smaller influence on most policy outcomes compared to those in public office. Only recently, however, has work begun to show that these biases also carry over to politicians (Best 2011; Caprara et al. 2010; Dietrich et al. 2012; Sheffer et al. 2016). The personality traits of representatives affect the actions of those who represent us. This research has shown that the characteristics of those who seek public office and the personality traits that they have could have a strong influence on the functioning of government. Yet, in spite of these seemingly influential consequences, we know little about the relationship between personality and the political ambition that motivates individuals to seek public office. As a scholarly community we have largely taken ambition to be an unvarying constant that motivates individuals to seek public office and pursue higher positions of power. Rather than focusing on the uniqueness of those who ambitiously pursue political office, scholars have chosen instead to examine the structural factors that encourage or limit the nascent progressive ambition of ambitious politicians (Abramson, Aldrich, and Rohde 1987; Black 1972; Rohde 1979). Instead of seeking to understand what characteristics produce the ambition that is unique

3 2 to individuals who choose to run, ambition for political office has been taken as a natural response to the possibilities that lie before politicians (Schlesinger 1991, 38). By taking ambition as a largely unvarying constant in the political class, however, we perhaps have missed and downplayed the idea that there are varying levels of ambition in the general population that might affect the calculation to run for office and ultimately the behavior of those who represent us. Indeed, while the subsequent study of political ambition has largely focused on structural factors, early studies of the political class found that those in elected office often differed substantially from the general population in the qualities and traits they held (Lasswell 1930, 1948; McConaughy 1950). Only recently has scholarship returned to the formation of nascent political ambition which must proceed the decision to run for office when the structural opportunity arises (Fox and Lawless 2005; Lawless and Fox 2004; Maestas et al. 2006; Stone and Maisel 2003). Nascent ambition motivates both the initial run for public office (which we call nascent political ambition) and the eagerness with which public officials seek and take opportunities to run for higher office (which we call nascent progressive ambition). Most of the factors that have been used to explain both types of nascent ambition, however, have been those related to an individual s relationship with politics rather than underlying personality differences. Nascent political ambition has been explained primarily using non-personality related factors such as minority status, ideological beliefs, or familiarity with politics. In this paper, we argue that nascent political ambition is more than that and is related to individual personalities. Recent work has shown a strong influence of personality on other aspects of political behavior including ideology, public opinions, and involvement in politics. With few exceptions (Dietrich et al. 2012; Clifford et al. 2016; Sheffer et al. 2016), this work on the influence of personality on political behavior has focused on individual behaviors related to

4 3 citizen behavior rather than elite political behavior. Yet, personality should not become irrelevant to behavior for individuals in and around the elite political sphere. Specifically, we argue that personality also affects political ambition. The presence of certain individual personality traits make certain portions of the population more predisposed to enter the political arena and expose themselves to the political work environment. Using a nationally representative survey, we find that individual personality plays a strong role in the nascent ambition of individuals for public office. We show that individuals with certain personality traits as measured by the five-factor model are more likely to express an interest in running for higher office. Perhaps unsurprisingly, these traits are also the types of traits that would be well-suited for individuals working as a representative at a number of levels of government. In addition to affecting the general population s interest in running for office, personality traits also affect nascent progressive ambition for those who are already in low-level political offices and are in a primed position to seek higher public office. These findings show that the desire to run for political office is not merely the result of a set of underlying structural factors and extends beyond an individual s relationship to politics. We find that individuals motivation to run for office stems from personality traits inherent in who they are. These factors continue to exert a significant influence on political ambition even after controlling for other non-personality factors. Thus, it is a certain type of individual who is attracted to run for political office, and these individuals are fundamentally different from those who are not interested in serving in public office. Finally, our findings have real implications for governmental outcomes. Studies in psychology and economics have shown that personality traits affect the probability that individuals are subject to cognitive biases or utilize certain heuristics in the decision making

5 4 process (Gilovich et al., 2002; Kahneman, 2011). Indeed, there is growing evidence that differences in personality among legislators affect their behaviors and actions in office (Capara et al. 2010; Best 2011; Dietrich et al. 2012; Sheffer et al. 2016). The biases associated with the traits of those who are more likely to run for office could have real impacts on political outcomes. Political Ambition Rather than focus on individual characteristics that encourage ambition, contemporary studies of political ambition primarily build on a rational choice paradigm of individual utility maximization that is generally assumed to be standard across individuals. Joseph Schlessinger s (1966) Ambition in Politics set the stage for this approach to understanding political ambition. Rather than looking at individual qualities or traits that may predispose an individual to run for office, Schlessinger and those who followed him placed running for office as a strategic response to the institutional structure and political opportunities available to individuals (Abramson, Aldrich, and Rohde 1987; Black 1972; Rohde 1979; Schlesinger 1966). These factors included the ability to pursue higher office given one s current position (Canon 1993), the availability of open seats in higher offices (Copeland 1989; Stone and Maisel 2003), term-limits (Moncrief 1999), party majorities either in the office where the potential candidate currently was serving or in the office to which the candidate aspired (Maestas et al. 2006; Squire 1988), the partisan composition of the electorate (Maestas et al. 2006; Stone and Maisel 2003), the electoral environment (Jacobson and Kernell 1981), recruitment (Broockman 2014; Kazee and Thornberry 1990; Matthews 1984; Sanbonmatsu 2006) or negative recruitment (Hassell 2016; Niven 2006) by party actors or other major political organizations, or the position of the potential candidate within the party network (Desmarais, La Raja, and Kowal 2015; Dominguez 2011; Hassell

6 5 2016). In short, ambition has largely been considered to be a constant presence among politicians. In this framework, the decision to run for office then becomes merely a rational calculation of the benefits of winning multiplied by the chances of winning minus the costs associated with running. While individual characteristics may be a factor, they are only explicitly a part of the process when they limit the ability of individuals to run either due to the lack of resources, whether financial or experiential, or a lack of familiarity with the political process. While a large body of work focuses on the structural determinants of who runs for office among those who already have the political and progressive ambition to do so, we take a step back by examining the traits associated with having this ambition in the first place. Our argument that political ambition is rooted in individual personalities does not necessarily contradict the rational choice view on this topic. Rather, personality traits likely affect individuals calculations of the costs and benefits of running for office. In fact, the study of ambition in politics has not always been about the political environment and institutional structures that may encourage or deter candidates from running for public office. Early studies of individuals who were motivated to seek and hold political office centered on the characteristics and traits of those individuals rather than the institutional structures that might influence the decision to run (Barber 1965; Browning and Jacob 1964; Fishel 1971; Greenstein 1967; Lasswell 1930, 1948). These earlier works argued that political candidacy was not merely an extension of regular political activity by a randomly drawn subset of individuals, but rather that these individuals aspiring to office had distinct qualities and traits from the general public (Fishel 1971; Lasswell 1948). Limited tests of a small sample of South Carolina state legislators in the 1940s showed that political elites, along with other findings, were less neurotic and introverted as measured

7 6 using Bernreuter Personality Inventory and the Guilford-Martin Inventory of Factors GAMIN Test (McConaughy 1950). 1 At the same time, in many ways local politicians are similar to others in upper middle class society. 2 Politicians and those in middle class society held similar characteristics on such attributes as power-seeking, affiliation inclination, and achievement as measured using an individualized measure developed out of the Thematic Apperception Test (Browning and Jacob 1964). 3 Although scholars made significant progress in this line of inquiry, the lack of a means to sufficiently measure and isolate the configuration of personality variables among individuals in public office has limited the ability to continue this area of research (Greenstein 1967). While new tools to measure personality have been developed, getting politicians to take personality tests is difficult. The limited evidence we have, while suggestive, is dated, limited, and not entirely consistent in its findings. 1 The Guilford-Martin Inventory of Factors is a now infrequently used battery of questions created in the 1940s designed to measure personality (Martin 1945). The Bernreuter Personality Index was one of the first inventories created in the 1930 and designed to measure traits such as introversion, extroversion, and neuroticism. It was used widely in clinical settings and is a precursor to Minnesota Multiphasic Personality Inventory (MMPI) which is another now widely utilized personality index similar to the Big Five personality measure which we use here (Butcher 2009). 2 It is important to note that these tests did not measure the same personalities nor did they use the same comparison groups. In addition, Browning and Jacob (1964) note that there also were differences across different communities where the balance of power between the business communities and public officials was different. As a result they theorize that the political environment and the centralization of power in a governing body relative to other community organizations may also affect the type of individuals who choose to run for office. 3 The Thematic Apperception Test was developed in the 1930s. Respondents are shown a series of pictures and asked to explain the story in the pictures allowing the test giver to analyze respondents attitudes and underlying characteristics. It is a projective test based on psychodynamic theory which bases behavior on subconscious underlying factors. While this psychological test was more widely used in a previous era its merit (specifically because of its lack of validity and reliability) is now widely questioned in the psychological community (Cramer 1999; Lilienfeld, Wood, and Garb 2000)

8 7 In addition, these older efforts to understand the personalities of political elites have looked exclusively at the individuals who hold office rather than examining the desire to hold political office or a desire to seek higher office. While it may be that personality determines levels of nascent political ambition, it could also be the case that voters are more inclined to support candidates with certain personalities or that certain personality traits help run more effective campaigns. Thus it could be that personality does not influence nascent ambition, but rather that the electoral screen that filters the aspirants from the office holders favors a certain set of individuals with certain personality traits. Our research allows us to directly measure how personality affects both political and progressive ambition. Personality and Its Influence on Political Behaviors A substantial body of recent research finds that personality traits play a significant role in a myriad of individual political behaviors. We build on these findings to highlight how personality traits likely affect individuals nascent political ambitions and local officials nascent progressive ambition. Over the last several decades, students of personality have used the lexical approach to identify five facets of individual personality. This approach assumes that critical differences in personality are captured in everyday language and that people can identify the most salient differences in what people are like. Lexical analysis involves gathering lists of adjectives and asking respondents to self-report how well each of the words describes themselves. Over the years, the battery of questions has been tested using as many as 240 adjectives (Costa and McCrae 1992) and as few as ten (Rammstedt and John 2007). Regardless of culture (McCrae 2002) or the number of items included in the list of adjectives, five factors consistently emerge as dominant features of individual personality. Although adolescents and seniors may exhibit

9 8 sharp changes in response to life experiences, the five personality traits show remarkable stability over the course of one s life (Cobb-Clark and Schurer 2012; Specht, Egloff, and Schmukle 2011) and are genetically heritable (Bergeman et al. 1993; Van Gestel and Van Broeckhoven 2003; Jang, Livesley, and Vemon 1996; Jang et al. 2002, 2006; Vukasovic and Bratko 2015). These five factors do not represent an exhaustive, nor exclusive list of every aspect of human personality. Rather, the Big Five are seen as broad domains into which various subsidiary personality traits can be organized. After explaining each trait below and its relationship to other political behaviors, we develop a framework for understanding why personality influences nascent political ambition. Those high in extraversion are more interested in the breadth of activities than the depth. In groups, they tend to assert themselves, be noticed, and dominate conversation. Extraverts enjoy interacting with others and are usually perceived to be very energetic. Extraverts tend to have numerous friendships, enterprising vocational interests, numerous club memberships, participate in team sports, and have high interest in politics (Gerber, Huber, Doherty, and Dowling 2011a, McCrae and Costa 2008). Extraverts are also more likely to think that they can make a difference in their community and have higher levels of political efficacy (Mondak and Halperin 2008). The strong association between extraversion and political participation and efficacy suggest that individuals with higher levels of extraversion will be more likely to express interest in running for office. We also expect that most individuals will recognize that holding office entails regular interactions with constituents and other public officials, which is another activity that extraverts enjoy. Those already in office will have experienced the social demands associated with running for and holding an elected position. As such, we expect that current

10 9 office holders with higher levels of extraversion will also be more interested in running for higher office, 4 which leads us to our first set of hypotheses: H1a (Extraversion and Political Ambition): Individuals with higher levels of extraversion will be more likely to express nascent political ambition H1b (Extraversion and Progressive Ambition): Local officials with higher levels of extraversion will be more likely to express nascent progressive ambition People who are high in openness to experience tend to be creative. They are intellectually curious, more sensitive to beauty and appreciate art, emotion, adventure, unusual ideas, and variety of experience. Openness is associated with being interested in politics, knowing about politics, interest in travel, having many different hobbies, knowledge of foreign cuisine, and diverse vocational interests (Gerber et al. 2011b, McCrae and Costa 2008). Running for public office provides a range of new and unique experiences for those who have not run before. It could challenge the individual to respond to new situations and to process new information. An interest in politics and opportunities for new experiences might cause those high in openness to be more amenable to the idea of running for office. Conditional on already having obtained public office, the experience of running for a higher office may not seem as unique or new. Moreover, higher offices at the state and national levels are partisan, which may be less appealing to open individuals given their lower levels of partisanship. As such, we are uncertain as to whether to expect a relationship between openness and nascent progressive ambition among local officials. H2 (Openness and Political Ambition): Individuals with higher levels of openness to experience will be more likely to express nascent political ambition. 4 This is consistent with the findings from Dietrich et al. (2012) who measure the Big Five personality traits of 92 state legislators.

11 10 Agreeableness captures differences in concern for social harmony. Those high in this trait tend to work well in teams, trust other people, show kindness to others, and compromise. Agreeable people are more likely to be economically liberal, but socially conservative. As such, they are found in both major political parties and are strong partisans. So far, no direct effects have been found between agreeableness and political engagement (Gerber, Huber, Doherty, Dowling, et al. 2011; Mondak et al. 2010). Running for and holding office place individuals in situations where they are likely to have to interact with individuals who disagree with them. As such, discomfort with political conflict might motivate agreeable individuals to stay out of politics and not run for political office. A similar logic applies to agreeable individuals who already hold public office and are considering whether to run for higher office. Given increases in polarization and partisan discord in state and national politics, agreeable local officials may actually be less likely to want to seek higher office. Moreover, those who are lower in agreeableness will be more likely to see themselves as being better qualified to make decisions and less concerned if that offends others, be they constituents or other public officials. H3a (Agreeableness and Political Ambition): Individuals with higher levels of agreeableness will be less likely to express nascent political ambition. H3b (Agreeableness and Progressive Ambition): Local officials with higher levels of agreeableness will be less likely to express nascent progressive ambition. Conscientious people aim for achievement and strive to fulfill their duties. These individuals enjoy planning for the future and prefer to think carefully about their decisionmaking rather than taking spontaneous action (DeYoung, Peterson, and Higgins 2002; Thompson 2008). Those low in conscientiousness are impulsive, spontaneous, disorganized, and are more likely to engage in anti-social behavior (Ozer and Benet-Martinez 2006). Politically, conscientiousness is associated with political conservatism and being less likely to vote regularly

12 11 (Gerber, Huber, Doherty, and Dowling 2011a, 2011b). Our expectations for conscientiousness are mixed. McCrae and Costa (2008) argue that the basic tendency of conscientious individuals is striving for achievement. Conscientious people have a strong sense of purpose and high levels of aspiration. These strong ambitions are what motivate them to be organized and plan for the future. McCrae and Costa (2008) argue that this is why Richard Nixon ran for president. On the other hand, the tendency of conscientious individuals to think carefully about a course of action before making any decisions likely means they will more thoroughly consider the costs and benefits of running for higher office than someone who is less conscientious and, thus, more spontaneous and impulsive in her decision-making. As such, a conscientious individual may be more realistic about the costs of running for office and her chances of winning and, as a result, express less nascent political ambition. In addition, the unorganized and uncertain nature of political work and campaigns could repel conscientious individuals from running for office. The same logic applies to local officials who have already experienced the dynamics associated with public office. Conscientious individuals lower propensity to turnout to vote may also carry over to other forms of political participation, such as running for office. Because of these conflicting processes, we do not have a clear expectation for the effect of higher levels of conscientiousness on nascent political or progressive ambition. Finally, emotional stability is more often characterized by its opposite (neuroticism). Those with high levels of emotional stability are less likely to experience negative emotions such as anger, anxiety, or depression. Emotionally stable individuals are less likely to view ordinary situations as threatening, or have negative reactions to everyday experiences that last for long periods of time. In the United States, emotional stability is associated with having a strong interest in politics and expressing it (Gerber, Huber, Doherty, and Dowling 2011a), all of which

13 12 suggest that emotional stable individuals would have greater interest in running for office. However, emotionally stable individuals are also less likely to have a large social network or try to persuade others to vote for a particular candidate (Mondak et al. 2010), characteristics which are also essential to succeeding in a campaign. As such, we do not have strong priors on whether emotional stability will predict political ambition in the mass public. Concerning progressive ambition, however, Dietrich et al. (2012) find that emotionally stable state legislators are more likely to be interested in running for higher office. They argue that the risk of defeat does not spark anxiety among the emotional stable. Building on this, we expect that more emotionally stable local officials will also be more likely to want to run for higher office. H4 (Emotional Stability and Progressive Ambition): Local officials with higher levels of emotional stability will be more likely to express nascent progressive ambition. To test these hypotheses, we conducted two studies on the relationship between individuals ambition for running for office and their personality traits, as measured by the Big Five. The first study, which we describe below, employs data from a survey of US adults. The second is a survey of elected municipal officials from across the US. Identifying a credible pool of potential candidates has been a challenge in work on candidate emergence. Using a survey of local officials allows us to examine whether the findings from Study 1 are consistent in a sample of individuals who have experience in political office and have likely considered running for higher office at the local, state, and national levels. In fact, we found in our survey that 41% of local officials are interested in holding statewide office while 14% reported an interest in holding national office someday. Examining local officials also allows us to examine progressive ambition in addition to political ambition.

14 13 Study I: Personality and Political Ambition among Voters In June, 2015 Clear Voice Research recruited 2,118 participants in their online panel to participate in a research project about political attitudes, of those 1,935 completed the entire survey. Respondents were 50.7% male, 80.6% white, 33.7% Democrat, and 24.7% Republican. 5 Most of the respondents had completed at least some college (78.1%) and 86% of them reported an annual income of less than $100 thousand. Using language from Lawless and Fox (2010) we asked respondents to select Which of the following best characterizes [their] attitude toward running for office in the future. Only 1% of our respondents reported actively considering running for public office, 16% said that they were open to the possibility of holding elective office in the future, while 83% reported absolutely no interest in holding elective office at any time in the future. We followed up with an additional battery of questions from Sanbonmatsu et al. s (2008) survey of Mayors and State Legislators. Using a scale from 0-10, we asked members of these respondents to rate the importance of the following factors in their interest toward holding elective office: influence on policy, advancing a political career, increasing business contacts, increasing social contact, fulfilling their civic duty, their dedication to a candidate, the excitement of politics, their concern about a particular issue, their desire to support a political party, and their interest in serving the public. We use these as controls in our model. 6 5 More information on the survey demographics and recruitment information are available in the appendix. 6 While we use these 10 factors of interest in office as controls it could also be that these factors are mediators by which personality affects ambition. When we run the models without these controls, however, we find no substantive or significant differences. Those alternative models can be found in the appendix.

15 14 Finally, each respondent completed the personality scale from the National Survey of Midlife Development (Lachman and Weaver 1997). This scale consists of 31 adjectives which were selected from existing trait lists and inventories (Bem 1981; Goldberg 1992; John 1990; Trapnell and Wiggins 1990). Often, scholars will opt to utilize shorter scales to measure personality (Rammstedt and John 2007), however, since we expected to have a small percentage of respondents indicate an active interest in seeking higher office we chose to use a more comprehensive measure for the Big Five traits for greater precision (Bakker and Lelkes 2016). 7 Since the dependent variable has three distinct categorical options, we fit a multinomial probit regression model. As an alternative we also ran the model as an ordered probit model since the data has what could appear to be hierarchical properties. We included the items mentioned in previous paragraphs as well as controls for income, education, race, and gender. The results of both the multinomial logit and the ordered probit regression models are displayed in Table 1. 7 The distribution of personality traits among the full general public and those in the general public with higher levels of political ambition is available in the appendix.

16 15 Table 1: The Influence of Personality on the Attractiveness of Elective Office Open to Possibility of Public Office Actively Considering Running for Public Office Ordered Probit Regression Extraversion 0.364*** 0.560* 0.278*** (0.122) (0.297) (0.082) Openness to Experience 0.342** 0.538** 0.253*** (0.136) (0.243) (0.088) Agreeableness *** *** *** (0.119) (0.233) (0.081) Conscientiousness *** *** *** (0.124) (0.236) (0.083) Emotional Stability (0.087) (0.175) (0.060) Education 0.180*** ** (0.063) (0.119) (0.042) Income (0.028) (0.061) (0.019) Race * * ** (0.043) (0.079) (0.030) Gender (Male baseline) *** *** (0.119) (0.250) (0.083) Influence Policy 0.187*** *** (0.030) (0.054) (0.019) Political Career 0.095*** 0.128** 0.068*** (0.027) (0.050) (0.017) Business Contacts (0.031) (0.041) (0.019) Social Contacts (0.031) (0.055) (0.020) Civic Duty 0.064** ** (0.028) (0.049) (0.018) Dedication to Candidate (0.028) (0.063) (0.019) Excitement of Politics 0.052* 0.188*** 0.046** (0.028) (0.058) (0.018) Issue Concerns ** * (0.031) (0.052) (0.020) Support the Party ** ** *** (0.030) (0.057) (0.019) Serve The Public 0.126*** *** (0.027) (0.052) (0.018) Constant *** * (0.456) (0.933) (0.325) Cut *** (0.349) Observations 1,939 1,939 1,939

17 16 AIC Pseudo R-squared Source: 2015 Survey of US Adults Note: Entries are multinomial and ordered probit regression coefficients, robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1, two-tailed test. What about personality? The only personality trait that is not associated with a general interest in holding elective office is emotional stability. Those who are emotionally stable are equally likely to be interested in holding elective office as those who are not. Consistent with theoretical expectations, extraverts and those who are open to experience are significantly more likely to be interested in elective office. Recall that those high in extraversion are interested in breadth of experience. Since very few people in our study are likely to be office holders, the idea of holding elective office represents a new experience that would have strong appeal to extraverts. Openness is related to interest in politics and having stronger attachments to a political party, which is one possible reason that those high in openness are more interested in holding elective office. Yet, open individuals are also intellectually curious and seek a variety of experience which would make holding elective office an attractive experience. Conscientious and agreeable individuals are significantly less interested in holding elective office. Agreeable individuals seek both to avoid conflict and to please other people. The idea of holding elective office likely accentuates the impossibility of these two. Not only does campaigning for elective office require conflict, the duties of elected office require people to make decisions that will please some and displease others. Even if an agreeable individual could be appointed to an elected office without standing in an election, the business of elected officials involves considerable conflict as divergent interests seek to shape the policy agenda. It is a little surprising that conscientiousness is negatively associated with a desire to hold elective office. After all, conscientious people are motivated to achieve great things and accomplish their duties,

18 17 both of which seem like goals that could be accomplished through elective office (McCrea and Costa 2008). Yet, conscientious people also do not like spontaneity. They prefer an orderly, predictable life; which may not be perceived as consistent with holding elective office. It is equally interesting that personality traits are associated with actively seeking higher office. Consistent with our expectations, agreeable people are more interested in the idea of holding public office than they are in taking steps to become candidates for higher office. Yet, every other personality trait performs identically in each of the two models in Table 1, and the coefficients are larger. This suggests that personality has a consistent, predictable influence, both on the desirability of holding elective office generally and the probability that someone will actively seek to become a candidate for higher office. We now discuss the substantive impact of the Big Five personality traits on interest in holding elective office. Figure 1 plots the predicted probability of expressing an interest in holding elective office for people at each level of a particular personality trait. Of the four statistically significant personality traits, openness and extraversion have the weakest substantive effect on interest in running for office. Yet, even among these traits, the model predicts a, 11-12% change in interest as an individual move from the bottom of the scale to the top. In other words, someone high in extraversion is 11.8% less likely to report no interest in running for higher office than someone at the bottom of the scale. Openness follows a similar pattern. In short, this means that extraverts and those who are open to new experiences are much more likely to be interested in running for office than those who are not.

19 18 Figure 1: The Influence of Personality Traits on Interest in Elective Office. Source: 2015 Survey of US Adults Note: Lines are the predicted probabilities calculated using the Effects package in R Whiskers are the 95% confidence interval

20 19 The model predicts even stronger effects for agreeableness. Our model predicts that those who are at the bottom in agreeableness are 600% more likely to be interested in running for elective office in the future than those who are at the top. In fact, the probability that these kind, compassionate, cooperative people who score the highest in agreeableness will report no interest in running for elective office is.94. Likewise, those low in conscientiousness are also much more likely to be interested in running for elected office. Indeed, the model finds that the predicted probability of someone who scores at the bottom of conscientiousness will report an interest in holding elective office sometime in the future is 0.29, while the probability that someone at the top of the conscientious scale will be interested in running for office is In other words, the least conscientious are 480% more likely to be interested in running for elected office than the most. Study II: Personality and Political Ambition among Municipal Officials The preceding is sufficient to show that some personality traits influence the attractiveness of running for public office. However, as we discussed previously, nascent political ambition is different than nascent progressive ambition. The best candidates for higher office are those with past political experience. In addition, most people have never held nor sought elective office, which makes it difficult for them to assess whether or not they would be interested in running for higher office. That is, holding public office may or may not be appealing in the abstract, but could take on an entirely different meaning among those who are familiar with the duties of elective office and the processes through which one obtains an office. It is entirely possible that the relationship between certain personality traits and interest in running for higher office is based on differences in perceptions of what it means for elected office, and not the reality of running for and serving in elective office.

21 20 To scrutinize the robustness and generalizability of the findings in the previous study, we conducted an online survey of elected municipal officials from across the US. Our questions concerning political ambition and personality were one of several projects in the survey. The survey was conducted using Qualtrics, a popular platform for online surveys. Subjects were recruited via s with a link to the survey on Qualtrics website. We sent each potential subject three s one to two weeks apart, inviting them to participate. The survey was conducted in two waves sent to two different samples of municipal officials. Invitations to the first wave were sent in May and June of 2016 to a sample of 27,862 elected mayors and legislators (e.g., city councilors, aldermen, supervisors, etc.) from 4,187 cities. The sample was compiled by a for-profit organization that gathers contact information and addresses of public officials from municipalities that have a website and a population above 10,000. The organization uses webcrawler software to identify when information changes on the contact pages of each city s website and then has research assistants update its contact list of officials accordingly. Unfortunately, this approach has a high error rate. Based on Qualtrics tracking, only 18,567 (or 67%) of the invitations were delivered to an active address. In addition, we looked up a sample of 832 officials in the list and found that only 44% of the addresses were accurate. 2,165 officials answered questions on the first wave of the survey, resulting in a response rate of between 11.7 and 17.8 percent. 8 This rate is similar to those from other surveys of municipal officials (e.g., Butler and Dynes (2016) report a response rate of 23%). The second wave of the survey was conducted in June and July of The sample consisted of the addresses of elected mayors and city councilors (or equivalent) gathered 8 The 17.8% is calculated as follows: 2,165/(.4375*27,862).

22 21 by Daniel Butler and Adam Dynes for the 2012 and 2014 American Municipal Officials Survey (AMOS) (see Butler and Dynes (2016) for more details on the samples). Excluding the addresses from the first wave resulted in a list of 29,250 s. The addresses from the 2012 AMOS were gathered in January through March of 2012 by a team of undergraduate research assistants who searched for the website of 26,566 US municipalities. The addresses from the 2014 AMOS were gathered in a similar fashion in early 2014 but excluded municipalities with a population below 3,000 due to the low percentage of small towns with websites. Given that these addresses were gathered 2 to 4 years prior to this latest survey, we knew that a large percentage of the s and names of the officials (in the case of cities that use generic accounts for each office) would no longer be accurate. Indeed, 26% of the s sent through Qualtrics were undeliverable. It is likely that many more of the addresses are no longer monitored though they remain active. With 1,500 officials participating, the response rate for the second round of the survey was 6.9%. The dependent variable in the analysis is local officials progressive ambition, which we measured local officials progressive ambition with the same question used by Lawless and Fox (2010). 9 The survey asked elected officials, Which best characterizes your attitudes toward running for a higher office in the future? They had four options, which we list below beginning with the answer that indicates the least amount of progressive ambition and ending with the one that indicates the most. 1) It is something I would absolutely never do. 2) I would not rule it out forever, but I currently have no interest. 3) It is something I might undertake if the opportunity presented itself. 9 The distribution of the personality traits among public officials and public officials with progressive ambition is available in the appendix.

23 22 4) It is something I definitely would like to undertake in the future. Given the clear ordering to these responses, we estimate an ordered logistic regression model in addition to a multinomial logistic regression model. The results are largely the same though the multinomial regression provides some nuance to the results The key independent variables in our analysis are respondents Big Five personality traits, which we measured using the Big Five Iventory-10 (BFI-10) with one additional item to measure Agreeableness. 10 The BFI-10 uses two items per personality trait and has been shown to retain significant levels of reliability and validity compared to a 44 item measure of the Big Five (Rammstedt and John 2007, 203). However, Rammstedt and John (2007, 210) find the losses in reliability are greatest with the two item measure of Agreeableness. To mitigate this, we followed their recommendation of adding a third Agreeableness item. 11 Local officials were asked to let us know how well the following statements describe your personality. I see myself as someone who This was followed by the list of the eleven items from our modified BFI-10, presented in random order. Respondents indicated how much they agreed with each statement using a five-point Likert scale ranging from Agree Strongly to Disagree Strongly. We scaled the responses from 0 to 4 such that higher values indicate someone who rates high on each personality trait. We then calculated policymakers average response to the 2 (or 3) items measuring each personality trait. A major concern with self-reported personality tests taken by elected officials is social desirability bias in their responses (Remmel 2016). Elected officials may be reluctant to admit that they get nervous easily or are not kind to almost everyone, even in a confidential survey. 10 See Figure A1 in the Appendix for the list of items used to measure the Big Five. 11 The additional Agreeableness item is I see myself as someone who is considerate and kind to almost everyone.

24 23 Remmel (2016) examines this concern by conducting a Big Five personality test on Vermont state legislators and then comparing these self-reports to peer-reports of those same state legislators completed by their friends, spouses, and/or adult children. Remmel (2016) finds a strong correlation between the self-reports and peer-reports. Moreover, the results suggest that the self-reports are less prone to social-desirability bias than the peer-reports, as state legislators were more likely to agree with the more negative statements than their peers. These results support our use of local officials self-reported personality tests. As before, we included a number of control variables that might influence an elected official s willingness to seek higher office. Margin of victory is a dichotomous variable that is coded one if the elected official won their last election by less than five percent of the vote. Those who narrowly won a previous election might think that they are unlikely to win an election for higher office. Years in office is the number of years that the elected official has been in their current position. We asked the elected officials how long they expect to be in their current office. Anticipated length in current office is the number of years they said that they planned to remain in their current position. We also asked if their current office had term limits. Those who are forced by statute to leave their existing position might be more likely to seek higher office. We included a dichotomous variable coded one if the elected official was selected in an election in which their party affiliation appears on the ballot and zero if it does not. Finally, we expect that elected officials are concerned with who will fill their seat should they seek higher office. As a control variable we asked them to suppose that their current office was vacant and tell us the probability that an individual with similar views as themselves would be selected to fill their seat. Current seat filled with similar candidate is the numeric (0-100) percent chance that their current seat would be filled with someone like the elected official.

25 24 We also asked how likely someone like themselves would be to win an open state legislative seat. The variable labeled legislative spot filled with similar candidate is the numeric percent chance that someone sharing the elected official s policy views would be able to win the open seat in their state legislative district. Because this is a survey of municipal officials, this variable measures the respondent s perception of how likely someone like them would be to win an election to the next highest office. The first three columns in Table 2 display the results of the multinomial logistic regression model with the option I would never run for higher office as the baseline. The models proceed in order of progressive ambition. The final column displays the results from the ordered logistic regression model. The findings from the two models yield similar results. In general, the longer an elected official has been in office and the higher the perceived probability that someone similar to them would be elected to fill their position, the less likely they are to be interested in running for higher office. While the effect of someone similar winning office is counter to our predictions, the size of the effect on these estimates is so small that we do not think that the difference is meaningful. In addition, this effect is only significant in the multinomial logit on the likelihood of responding an individual would definitely like to undertake the pursuit for higher office. Given its negative impact on definitely seeking higher office, the higher likelihood of someone similar to themselves winning their current seat could be also be construed as a measure of job security. In contrast, those who plan to be in their elected office for a long time and those who think that there is a good chance someone like them will be elected to their state legislative office are significantly more likely to be interested in running for higher office. As expected, these findings show that elected officials who think that they can win are more likely to seek higher office.

26 25 Table 2: The Influence of Personality on the Attractiveness of Higher Elective Office Multinomial Ordered Multinomial Multinomial (No Current Probit (Possibility) (Definitely) Interest) Regression Extraversion * *** *** (0.072) (0.078) (0.092) (0.045) Openness to Experience (0.077) (0.082) (0.096) (0.047) Agreeableness (0.107) (0.114) (0.133) (0.065) Conscientiousness *** (0.105) (0.115) (0.132) (0.061) Emotional Stability ** * ** * (0.077) (0.084) (0.100) (0.049) Won previous election by 5% pts. or less (1=yes) (0.226) (0.238) (0.284) (0.135) Years in Office *** ** *** (0.003) (0.004) (0.004) (0.002) Anticipated Length in Current Office *** *** *** (0.002) (0.003) (0.003) (0.001) Term limits exist for current office (1=yes) (0.039) (0.039) (0.040) (0.005) Partisan elections (1=yes) (0.133) (0.133) (0.133) (0.009) Probability current seat filled by similar ** *** candidate (Scale from 0 to 100) (0.002) (0.002) (0.003) (0.001) Probability similar candidate could win state ** *** *** *** legislative seat (Scale from 0 to 100) (0.002) (0.002) (0.002) (0.001) Constant * (0.480) (0.521) (0.609) Observations 2,156 2,156 2,156 2,156 AIC Pseudo R-Square Source: 2016 American Municipal Official Survey Note: Entries are multinomial probit estimates (Never baseline) and ordered logistic regression coefficients, robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1, two-tailed test.