INTRODUCTION
This study conducts an empirical analysis into whether presidential leadership qualities directly affect trust in government, which is typically a measure of the public’s perception of the government as well as its evaluation of government performance. Nowadays, not only political scandals involving the president’s staff and family members, but also poor management capabilities can affect the president’s approval rating. Of course, the president’s approval rating is not an exact measure of the public’s trust in government (Citrin & Luks, 2001). However, the lower the approval rating or the higher the degree of distrust in government, the more difficult it is to carry out state affairs. For example, when Choi Soon-sil’s abuse of authority was exposed in 2016, citizens held candlelight rallies in Gwanghwamun Square, which led to the impeachment of the president, Park Geun-hye, for the first time in Korean history. At this point, Park’s approval rating was only 5 percent (D. K. Lee, 2016); however, the public’s trust in the Korean government only declined to 24% in 2017 from 34% in 2015, according to the OECD trust in government survey (OECD, 2021). This study was conducted using survey data from mid-2019. President Moon Jae-in, elected in 2017 following the impeachment of former President Park, had significant popular support overall: he received an approval rating close to 80% one year into office and maintained a 45% approval rating into his second year. His second-year approval rating was the second-highest among those of previous Presidents (The Korea Economic Daily, 2019). There were no noteworthy political issues while the survey was conducted in 2019; the elections for the local government heads and the National Assembly, both of which resulted in landslide victories for the ruling party, took place in 2018 and 2020, respectively. This suggests that there is no correlation between a president’s approval rating and trust in government. Furthermore, this indicates that research is needed on the specific role of a president’s leadership pertaining to trust in government.
Although there is no clear definition of trust in government, it is commonly defined as the extent to which government performance meets the normative expectations of the public (Hetherington, 1998; Miller, 1974). Enhancing trust in government enables cooperation between the government and the public, which in turn leads to efficient policy making and implementation (J. M. Choi & Kim, 2014). Moreover, trust in government has gained ascendancy as social capital became recognized as a driving force in democratic systems and economic growth; this was only possible following the expansion of civil liberties and political freedom in post-democratization Korea (Cho & Nam, 2007). The more reliable the government, the more likely it is that citizens will comply with and support its policies (C. W. Lee & Cho, 2016). However, the level of trust in government has been declining steadily on a global scale (Kang & Park, 2018; B. K. Kim & Yi, 2013). According to the Pew Research Center of the United States, the proportion of U.S. citizens trusting the government declined from 77% in the 1960s to 54% in 2001, then further to 17% in 2019; similar patterns are observable in European and Asian countries. Possible consequences of this decline in confidence are a loss of essential resources needed for policy implementation and disruptions in government performance, both of which could undermine the quality and value of democracy (Nye et al., 1997). Furthermore, public resistance may incur additional costs as the government seeks to boost public support throughout the course of policy implementation, or even in the worst-case scenario, policy failure (Gamson, 1968).
Various studies on the determinants of trust in government have been carried out in the fields of public administration and political science. Easton (1965) developed the first theory pertaining to trust in government (1965); various factors have since been identified, including public policy satisfaction (King, 1997; Miller, 1974) and presidential performance (Citrin & Green, 1986). Drawing on Citrin and Green’s (1986) study, Cho and Nam (2007) explored the Korean public’s trust in its presidents, Kim and Yi (2013) undertook a comparative analysis of the differences between the Korean public’s trust specific to the president and trust towards the executive branch in general, and Jun, Kwon and Jung (2013) examined the relationship between the Korean public’s trust in the president and its trust as measured by its evaluation of government policy. In particular, Jun and colleagues suggested that in Korea, trust in the president is more influential than other factors, thus implying the importance of presidential leadership. This is because the authority of the president in making national policy decisions is paramount in Korea (Hahm, 2007).
However, current research regarding factors influencing trust in government in relation to presidents has a couple of limitations. First, the studies do not specify who or what the object of trust was; also, the nature and scope of the variables that affect assessments of trust are not clear (Bae & Lim, 2009b). This results in vague conclusions because the level of confidence can vary depending on not only the scope of trust in specific government personnel or departments, but also on the specific ways the government has either met or not met the expectations of its citizens (B. K. Kim & Yi, 2013). Second, although it has been argued that the relationship between presidential leadership traits and trust in government is significant in Korea, no clear metrics have been developed to gauge that relationship owing to the limitations of data and the absence of measuring tools (B. K. Kim & Yi, 2013). Furthermore, even in studies in which presidential leadership was used as a measure of Korean citizens’ trust in government, the statistical results were not significant (Cho & Nam, 2007).
To address the limitations of prior research, we limit our analysis of Korean citizens’ trust in the government to their trust in the central government. Second, we consider five presidential leadership qualities, which we draw from Lee and Kim’s (2017) situational-based qualities and the president’s basic needs, which are based on Greenstein’s (2000) criteria for successful presidential leadership: vision, communication, policy planning and implementation capacity, political power, and integrity.
LITERATURE REVIEW AND HYPOTHESIS
The Concept of Trust in Government
There is no common consensus among scholars regarding the definition of trust in government (M. H. Kim & Lee , 2019). In general, however, trust in government can be defined to the extent that the management of government affairs and performance meet the normative expectations of the public (Miller, 1974). In other words, the public’s trust in government is a subjective assessment of the government’s policy output, such as public goods, in connection with the government’s desired direction or policy outcomes (Hetherington, 1998). Thus, trust in government is at the heart of modern democratic theory (Cole, 1973). Also, as an indicator of the democratic system, cooperative interaction between the government and the public has been established, emphasizing its importance in efficient policy-making and implementation (Hetherington, 1998).
The meaning of trust here is specified into general trust and specific trust depending on whether it is trust for an unspecified number of people or trust for a specific target (Cho & Nam, 2007). Besides, trust has been discussed as an area of interest for the entire social sciences, with the reduction of transaction costs, the role of lubricants in information communication, and the concept of “Social Capital” functioning as the supply of a public good (Coleman, 1988; T. H. Park, 1999). The concept of trust has been conceptualized in various ways. It has been approached psychologically via personality theory and behavior theory, economically as a kind of public good, and politically as a positive assessment of the government affairs (Hetherington, 1998; Hwang et al., 2017).
Easton (1965, 1975) classified objects of political support as political communities, regimes, and public officials and defined political support as overt behavior that indicates a preference for one or more of these objects. Easton (1975) further divided political support into “specific support” for the achievements and capabilities of public officials and the performance of current government agencies and policies, on the one hand, and “diffuse support,” on the other, which refers to the public’s attitude toward government regardless of performance. A fundamental component of this political support is the legitimacy of the political system, which is deeply related to trust in government (Easton, 1975). The assessment of trust in government may thus vary depending on the public’s perception, and it is often difficult for researchers to evaluate these perceptions because it requires taking into account the natural cognitive limitations of human beings (Cho, 2008).
From this point of view, trust in government can vary depending on how one defines the government. Citizens might trust legislative, judicial, or administrative bodies, or they might trust the party in power or the leaders of that party (Jun et al., 2013; C. M. Park & Bae, 2011). Other studies on government trust have therefore considered whether to define trust in government (Bae & Lim, 2009b) as the public’s trust in the president and his or her party (Muller & Jukam, 1977), as trust in specific administrative agencies (Citrin & Green, 1986; Citrin & Luks, 2001), as trust of a parliament (Hetherington, 1998), or as trust in a political system that integrates the executive, legislative and judicial branches (Nye et al., 1997; Wong et al., 2009). Since we deem it important to define the scope of trust in government in our study, we define government here as the central government (Brewer & Sigelman, 2002; Cho & Nam, 2007; Citrin, 1974) that manages state affairs through the leadership of the president, who is an elected official. Since the results of our study confirmed that the relationship between presidential leadership and trust in government was significant, it is related to the evaluation of the ruling government under a democratic system as the trust of the central government (Cho & Nam, 2007).
Determinants of Trust in Government: Presidential Leadership
Three factors seem to be crucial in the determination of trust in government (Jun et al., 2013). First, the more positive the public’s perception of government performance, the higher the trust in government. Public satisfaction with policy outputs and outcomes, policy responsiveness, social safety, and economic performance affects trust in government (Hetherington, 1998; S. Kim, 2010; King, 1997; H. Y. Lee & Lee, 2017; Miller, 1974). Second, evaluation of the president’s management of state affairs and support for specific politicians’ influences trust in government (B. Barber, 1983; Citrin, 1974; Citrin & Green, 1986). The more capable and moral officials such as the president and political leaders are, the more trust they are likely to earn from citizens, effectively leading to more trust in government (B. Barber, 1983). Third, demographic factors such as age, gender, income level, education, and occupations affect levels of trust in government, although it is difficult to generate a consensus on this front because different researchers have arrived at different results (Christensen et al., 2020; Christensen & Lægreid, 2005).
The debate between Miller (1974), who argued that public’s perception of government policy was more important when it came to trust in government, and Citrin (1974), who argued that trust in individuals such as the president was more critical, led to many further studies on factors affecting trust in government (Erber & Lau, 1990). In other words, Miller (1974) approached the range of government with confidence in the national political system, which is the spread of public support, while Citrin (1974) studied public confidence in the president’s performance (Bae & Lim, 2009b). In this debate, there is discrimination against influencing factors depending on the scope and subject of confidence of the government, but not against the two-way relationship of the people (M. H. Kim & Lee , 2019; S. J. Lee & Kim, 2018).
Among them, Citrin (1974) and his follow-up studies investigated countries who had adopted the presidential system and concluded that the president is crucial. An empirical study by Citrin and Green (1986) described four factors that increase trust in government: awareness of political parties and presidents, good economic performance, policy satisfaction, and approval of personal characteristics of the president. The analysis also shows that trust follows from government policies that are consistent with national preferences and the achievement of policy objectives; however, individual characteristics such as presidential leadership and perception of presidential support emerge as chief determinations of trust in government. Both of these studies have had significant implications for studies on Korean citizens’ trust in government because the Korean president has so much authority and plays a significant role in national development (Bae & Lim, 2009b; Cho & Nam, 2007; Hahm, 2007; Jun et al., 2013; B. K. Kim & Yi, 2013). Even though there was an issue from reciprocal causation between the president’s thermometer and trust in government (Hetherington, 1998), others found that there was no reverse causality between them by controlling endogeneity with instrumental variables (Jun et al., 2013). Also, the previous study found that perceptions of presidential character independently impact the public’s political trust (Citrin & Green, 1986). In that sense, reciprocal relationships between presidential factors and trust in government would depend on the context and circumstances of the country.
The results of these studies indicate that the relationship between trust in the president’s leadership and trust in government in Korea requires further investigation. As we have noted, most prior studies do not explicitly explain the rationale as to why a given individual or entity was used to measure trust in government (Bae & Lim, 2009b). Further, conflating the president with the executive branch, as Jun and colleagues’ (2013) study does in measuring the level of public trust in the president through level of confidence in the Blue House (Cheong Wa Dae), is problematic (B. K. Kim & Yi, 2013). Empirical studies of the role of presidential qualities in trust in government are also few and far between when compared to those that consider other factors that play into trust in government (B. K. Kim & Yi, 2013; S. J. Lee & Yu, 2015); a majority of such studies are biased toward demographic background factors (Bae & Lim, 2009b). Although the distinction could affect assessments of trust in government, prior studies of trust in the president as an influential factor of the government’s confidence included presidential leadership factors as common sub-factors measuring presidential trust (Cho & Nam, 2007; Jun et al., 2013). Moreover, due to the limitations of measurement tools, few studies have yielded findings on the relationship between presidential leadership and trust in government (B. K. Kim & Yi, 2013). However, leadership in presidential personal traits and skills were not considered nor significant (Cho & Nam, 2007).
This study uses leadership qualities of the president as a measurement tool and a determinant of trust in government. Presidential leadership can be defined as a type of leadership that reflects the values and motivations that the president shares with his or her constituents (Burns, 1978) and that exerts influence to accomplish a purpose in a given government affair (Hersey & Blanchard, 1982). The president has the motivation and authority to coordinate government agencies as a single responsible actor by pursuing consistent national goals to ensure effective government operations (Kaufman, 1956, 1969). Presidential leadership theory is a type of traditional leadership theory in which leadership is posited not as being about domination and control but rather about qualities that affect a president’s constituents. It also draws upon trait theory that focuses on leaders’ innate qualities and interactions with leaders’ intrinsic personalities: trait theory focuses on leaders’ personal characteristics generally tied to their circumstances (Stogdill, 1974), while skill theory focuses on leaders’ competence or abilities that can be learned and are developable (Katz, 1955; Mumford, Zaccaro, Connelly, et al., 2000). Even though there is a distinction between the traits-based and skill-based theories, both are not mutually exclusive because they are leader-centered approaches about leadership; furthermore, skills appear to develop into a function of interaction between traits and experiences (Mumford, Zaccaro, Connelly, et al., 2000; Mumford, Zaccaro, Harding, et al., 2000; Northouse, 2016). Studies of presidential leadership have highlighted qualities of leaders, such as the personalities and abilities of the presidents and have evaluated the process of various leadership theories, either partially or overall (J. Barber, 1992; Hahm, 2018). Recent leadership theories have focused on leadership processes and situations based on the situational theory that argues effective leadership types may vary depending on circumstances (Northouse, 2016; Rost, 1991). Fiedler’s contingency model, for instance, combines the traits of a leader with the requirements of a situation to gauge the effectiveness of a leader’s characteristics in a given context (Fiedler, 1967; Vroom & Jago, 2007). Other studies have focused on common cross-situational traits that lead to the acquisition of specific skills and the development of expertise (Zaccaro, 2007). According to Greenstein (2000), who studied the factors contributing to successful US presidents, the leadership qualities of more successful presidents could be identified by considering their experience with failure. Through theoretical discussions such as these and the significance of the presidency as an impactful position on public and political groups, it can therefore be confirmed that studies related to presidential leadership focus on both acquired and innate qualities such as skills and traits of the president in a situational approach (Greenstein, 2000).
In our study, we rely the five conditions of success for Korean presidents as defined by Lee and Kim’s (2017) study, which develops a situational-trait leadership theory grounded in how presidents handle the political situation at home and abroad and how they draw on lessons learned from examples of failed leadership in previous domestic presidents. Moreover, this study is aimed at the five qualities based on Greenstein’s (2000) criteria for successful United States presidential leadership upon which the Korean government adopts a governance system similar to that of the U.S. (Horiuchi & Lee, 2008). Therefore, these five qualities based on Lee and Kim’s (2017) study provides great implications which call for wiser presidential leadership in Korea. The first condition is the ability to develop a vision that matches the spirit of the times, is realizable, and can be effectively shared with citizens. The second is communication skills that enable the government to actively share information with citizens and encourage citizen interaction. The third is empathy that the president uses to form a bond with citizens, conveying his or her sincere understanding of their difficulties (H. C. Lee & Kim, 2017; Rifkin, 2009). However, empathy can be seen as an aspect of communication skills (Pitson, 1996); therefore, we integrated empathy into the communication skills category in our study. Fourth is the ability to manage government affairs, or the ability to delegate, supervise, and implement planned means and ends. Fifth is political power, the ability to resolve political conflicts involving various stakeholders arising from wicked problems and to show the determination required in the event of a national crisis. In addition, a lack of principle has been identified as a characteristic of failed presidents, and integrity has been shown to be an important trait of elected officials (Laustsen & Bor, 2017; McAllister, 2016; Storey, 2004): therefore, it too needs to be taken into account as an influential factor in trust in government. In that sense, we would like to define these five presidential leadership qualities as determinants of trust in government that are measured by the public’s perceptions (i.e., presenting a national vision, communication skills, managing government affairs, problem-solving skills through political powers based on presidential leadership skills and, integrity based on presidential leadership trait).
Relationship between Vision and Trust in Government
Vision is based on predictive judgments such as the direction to move forward in state administration. By presenting them clearly, the government has a function to gain public support and trust (Hahm, 2007). It is mentioned by Katz’s (1955) skill-based leadership theory as a conceptual skill that expresses persuasively and concretely an organization’s goals as the direction of the government. Also, in the context of Kouzes and Posner’s (2002) transformational leadership theory, competent leaders present their ability to induce people’s behavior by expressing concrete and feasible results in the future as a vision. Take for example former South African president Nelson Mandela, who is commonly described as a leader who had a clear vision for his country during his tenure: this contributed to his success in transforming it (Northouse, 2016). Presidents who lack a clear vision often fail in the face of the uncertainties of national goals and government affairs (Greenstein, 2000; Hahm, 2007; H. C. Lee & Kim, 2017). Former U.S. president Jimmy Carter, for example, is considered a failed president on the account of his lack of experience in state administration and his weaknesses in showing the direction of the U.S. government at the time (Berggren & Rae, 2006). On the other hand, former U.S. president Ronald Reagan was named the “Teflon President” thanks to his ability to overcome the nationwide economic downturn by simply and repeatedly insisting on a vision of government affairs that motivated the public and eventually led most challengers to his re-election campaign of 1984 to change their minds (Greenstein, 2000). Among the former presidents of Korea, Kim Young-Sam amplified public distrust of the government by frequently changing state administration goals (H. C. Lee & Kim, 2017). Lee and Lee’s (2007) study of the public showed that citizens’ evaluations of all presidents following the democratization of Korea yielded statistically significant results on presenting a vision in the role of leadership. Therefore, we propose the following hypothesis.
H1: The president’s ability to effectively present a national vision to the public will have a positive impact on trust in government.
Relationship between Communication ability and Trust in Government
According to Fine and Waterman (2008), the president’s ability to communicate with the public is regarded as a personal skill, which he or she demonstrates through media appearances (e.g., speeches and press conferences) and which plays a role in the setting and execution of his or her policy agenda (Edwards & Wood, 1999). Although a number of studies show that the more media exposure a president has, the lower the approval rating (Y. J. Choi, 2008) and that none of the previous presidents communicated efficiently (Greenstein, 2000), recent advancements of information and communication technology such as social media, (e.g., Twitter) have enabled two-way communication channels between citizens and the president, thus enabling, at least in theory, a more transparent government and more efficient service delivery. Results from empirical studies do seem to indicate that there is a positive relationship between this expansion of communication channels and trust in government (M. J. Park et al., 2016). Moreover, confirming to Lee & Lee’s (2007) study, the relationship between presidential communicating skills and the role of leadership in the public’s evaluation of former presidents in Korea was statistically significant. Through this, a bond was formed by showing the sincerity of the president to share the suffering of the public (H. C. Lee & Kim, 2017; Rifkin, 2009). Furthermore, communication leadership is a factor mentioned in most recent leadership theories, such as servant leadership theory, which highlight qualities such as listening and empathy (Errol & Bruce, 2005; Spears, 1996), and ethical leadership theory, which focuses on interactive communication between leaders and constituents from a social perspective (Brown et al., 2005; Northouse, 2016). Greenstein (2000) argues that excellent public communication skills are crucial to the success or lack thereof of US presidents. Based on the above arguments, we set the following hypothesis.
H2: The president’s ability to communicate with the public will have a positive influence on trust in the government.
Relationship between Managing Government Affairs and Trust in Government
Studies have suggested that the president’s ability to appoint talented people with the necessary qualifications to engage in public service, empower agencies to carry out the duties of public institutions, and execute policy as well as his or her commitment to keeping the promises he or she made as a candidate are important in securing public trust (Hahm, 2002; H. C. Lee & Kim, 2017). It is a great approach to encourage subordinates to speak out their opinions rather than suggest the president’s dogmatic choice by forming a team and getting the most out of it as a presidential quality to create efficacious institutional systems (Greenstein, 2000). Moreover, the ability to form and manage policies in state administration has been used as a personal skill of the presidential leadership (Fine & Waterman, 2008). In this regard, for successful presidential leadership, it is appointed that focusing on administrative leadership or managerial leadership (Campbell, 1998; Hahm, 2002; H. C. Lee & Kim, 2017; Nathan, 1986). The emphasis on governance recalls Miller’s (1974) argument about the extent to which the effectiveness of relationships between government policies and trust in government, but the difference in our study is the focus on the president’s personal management capabilities. In addition, the idea that the effectiveness of government policies and individual qualities are mutually exclusive considerations in assessments of trust in government has been rejected (Bae & Lim, 2009a). Therefore, we assume the following hypothesis.
H3: The president’s ability to manage government affairs will have a positive impact on trust in government.
Relationship between Political Power and Trust in Government
The president’s leadership abilities should be complemented by political power that enables him or her to resolve political chaos (Neustadt, 1990), help government agencies function in an integrated fashion, reduce the costs incurred by political conflicts (Hahm, 2002), and ensure that the public serves as a strong support base for the government. In addition, leadership as a coordinator to resolve official or informal political participants’ conflicts in the management of government affairs, or the ability to lead political integration through determination to make a clear judgment on the crisis of the country, is a major characteristic in leadership theories (Hahm, 2002). Political power has also been cited as a major characteristic. In the context of Katz’s (1955) skill-based leadership theory as human skills, refers to the ability to harmonize and adapt one’s thoughts on pending issues by detecting other people’s perspectives. From the political psychology approach, the president’s political leadership is defined in the concept of bargaining ability and persuasive power found in prestige and reputation (J. Barber, 1992; Hahm, 2002; Neustadt, 1990). Through an empirical study by Lee & Lee (2007), the relationship between problem-solving skills through political power and citizens’ evaluation of presidential leadership in Korea was found. Greenstein (2000) offers up Lyndon Johnson as an example of a U.S. president who had significant political power. Owing to this power, the Senate and House of Representatives were aware of the political cost they would incur by ignoring his policy initiatives; this played a role in him gaining strong public support (Greenstein, 2000). We thus suggest the following hypothesis.
H4: The president’s political power will have a positive influence on trust in government.
Relationship between Integrity and Trust in Government
Scholars have not reached a consensus as to what constitutes integrity (Moon & Kang, 2012). In general, it is considered to be a quality that discourages political leaders from abusing their power or prioritizing their personal interests over those of the public (Oh & Chun, 2012) and encourages accountability and transparency alongside anticorruption (Langseth et al., 1997; S. B. Lee, 2012). According to Lee and Kim (2017), integrity as a quality of presidential leadership is a fundamental characteristic of the president: it is not only constant, but also relevant from a situational perspective. It has also been frequently discussed from a public policy perspective (Moon & Kang, 2012) and is typically identified as an important characteristic of elected officials such as presidents (Laustsen & Bor, 2017; McAllister, 2016; Storey, 2004). Even though integrity has conceptual overlap with ethical leadership theory, it is only one facet of moral behavior that refers to adhering to ethical essentials (Engelbrecht et al., 2017; Palanski & Yammarino, 2007). A study by Laustsen and Bor (2017) identifies the characteristics of candidates that influenced voters using data collected between and 2008 from surveys undertaken by the American National Elections Studies: it suggests that candidates’ integrity is a required criterion for senior executives and elected officials. Also, studies have proposed that there is a statistically significant positive correlation between integrity leadership and followers’ trust in the leader (Engelbrecht et al., 2017; Kalshoven et al., 2011). Although Citrin and Green (1986) conclude that while the relationship between presidential integrity and political trust is statistically significant, its impact is not as high as other leadership ability variables. However, Jun and colleagues’ (2013) study compared and analyzed the differences between trust in the president and trust in the executive branch as a whole: it suggests that immorality, including a lack of integrity and corruption, do affect assessments of Korean citizens’ trust in government. Taken together, both theoretical and public policy studies lead us to speculate the following hypothesis.
H5: Due to the presidential personal trait, integrity will have a positive relationship with trust in government.
METHODS
Our research model is summarized in Figure 1. We use the trust of the administration (central government), an organization that manages state affairs under the authority and power of the president, as a dependent variable. Presidential leadership qualities (vision, communication skills, ability in managing state affairs, political power, and integrity) as proposed by Lee and Kim (2017) serve as our independent variables. Control variables include demographic characteristics that could affect both trust in government and assessment of presidential leadership qualities such as gender, occupation, education level, age, happiness, and fairness.
Data
This study used “2019 Public Perception Survey on the Role of Government” data from the Center for Survey Research of the Graduate School of Public Administration at Seoul National University, which regularly conducts a survey that targets men and women over the age of nineteen across Koran on their perceptions of the role of the government in Korean society. The survey has been conducted since 2014 to generate basic data for effective policy making: it achieves this goal by accurately grasping the expectations and requirements through a survey on the public’s perception of the necessity and role of government work. We used the 2019 survey data as it included items about government trust and presidential leadership. A total of 6,997 (40%) out of 16,177 contacts were extracted through a multistage stratified sampling method in which the results of this survey was calculated using weights for representativeness and adjust the non-response bias (H. J. Lee & Kang, 2012). The specific demographic characteristics of the respondents are shown in Table 1.
Measures
Dependent variable
The survey asked the following: “How much do you trust the following institutions and social leaders?” and included a total of 14 types of institutions and leaders: central government, heads of local governments, presidents, lawmakers, prosecutors, police, courts, counties, public institutions, large corporations, small businesses, unions, and academic experts. The dependent variable of this study is the public’s level of confidence in the Korean central government. We defined “central government” as the executive branch that manages government affairs under the leadership of the president. To measure the public’s trust the survey used a single-scale question that asked respondents how much they trust the central government for which possible answers ranged from “I do not trust it at all” (=1), to “I trust it very much” (=4). Prior studies limited the scope of the dependent variable (i.e. trust in government), to the central government, which was measured using a single-scale statement, to confirm that it was theoretically reasonable and statistically significant (Jun et al., 2013; B. K. Kim & Yi, 2013; Yoon & Seo, 2016).
Independent variables
The independent variable of this study is the leadership qualities of the president. The survey asked respondents about the perception of leadership of national leaders in Korea such as the president with respect to vision, communication skills (in which we included empathy), management of government affairs (planning, execution), political skills, and integrity: the answers were measured using a 7-point Likert scale (very poor=1, neither poor nor good=4, very good=7). Thus, this scale determined established face-validity by asking about presidential leadership qualities, which allowed for more objective evaluations.
Control variables
Prior studies have shown that demographic factors such as gender, age, race, education level, occupation (public/private sector), income, political orientation and ideology affect trust in government (Brehm & Rahn, 1997; Christensen et al., 2020; Christensen & Lægreid, 2005; Hetherington, 1998). Therefore, in this study, demographic variables were used as control variables: gender (male=1, female=0), political orientation (conservative=1, moderate= 2, progressive= 3; criterion variable = moderate), education level (less than middle school=1, high school graduate=2, professional school/university graduate=3, masters/PhD graduate=4; criterion variable=less than middle school), occupation (public sector=1, private sector=0) measured in dummy variables, age (at least 19 years old), and households’ gross monthly income (an ordinal category variable of post-tax income; less than KRW 490,000= 1, more than KRW 10 million=15).
Another demographic factor that plays a role in trust in government is the extent of a region’s development, which empirical studies document in terms of the level of urbanization and the rate of participation in social organizations (H. B. Park et al., 2003; Putnam, 1993). Therefore, to control for the characteristics of the region, two dummy variables (nonmetropolitan area=0, metropolitan area=1; reference variable=nonmetropolitan area) were used as regional control variables in 17 questions asked by metropolitan governments. Furthermore, the emotional state of happiness affects the public’s perception of trust in government and has been used as a variable in a number of empirical studies (J. M. Choi & Kim, 2014; Hyun & Lee, 2007). Therefore, we use happiness (not happy at all=0, very happy=10) as a control variable. Finally, we explored the relationship between key variables, including Korean citizens’ perception of the degree of fairness (very unfair=0, very fair=10) of the government (Herian et al., 2012; Shin & Lee, 2016), that could affect the public’s satisfaction with policy making and its trust in government. The questionnaire used in this study is presented in Appendix A.
Analysis
To test the relationship between the presidential leadership qualities and trust in government, we set up a multiple regression model and estimated with the ordinary least squares (OLS) method. We also introduced standardized regression coefficients to determine the order of significance of each quality.
We used the OLS multiple regression method for the following reasons. First, the histogram of the dependent variable of this study shows that it is normally distributed. Furthermore, although understanding the relationship between the independent and dependent variables through the scatterplot was difficult because the survey we relied on for our data used a Likert scale, we found that hypothesis testing was possible with a multiple regression model: it showed a significant correlation between the independent and dependent variables. Second, the OLS method has the advantage of enabling researchers to estimate independent influence among variables by mutually controlling the significance between variables and reducing the omitted-variable bias to an extent (Wooldridge, 2018). However, regression analysis alone cannot identify a clear causality because there is a high correlation between variables in the field of social science (Wooldridge, 2018). Nevertheless, researchers can solve this problem by drawing on prior research (Wooldridge, 2018). Prior research has shown that regression models demonstrate no causal relationship between trust in government and confidence in the president (Jun et al., 2013), but do indicate that presidential leadership has a positive influence on trust in government (Cho & Nam, 2007; Citrin & Green, 1986). Therefore, we performed descriptive statistics with correlation analysis and used Breusch-Pagan tests to determine whether the basic assumptions of the regression model were met. We then tested the research hypothesis through OLS multiple regression analysis. All analysis was performed using STATA software (version 16.0).
RESULTS
Descriptive Statistics
Among the number of respondents residing in the Seoul Capital Area and the Honam region, the number of respondents with a progressive political orientation and the number of respondents with household income levels was relatively high. With respect to presidential leadership qualities, 45.1 percent gave the president an excellent rating on integrity, followed by communication (43.5 percent), vision (42.5 percent), management of government affairs (40.3 percent), and political power (39.8 percent). In other words, among the five qualities, integrity ranks relatively high.
Table 2 provides descriptive statistics and correlation coefficients for all variables. According to the results of the descriptive statistics, the dependent variable of this study, trust in central administration, was found to be slightly above ordinary (mean 2.49 out of 4 [SD=.72]). The mean was 4.27 (SD=1.35) for integrity, 4.24 (SD=1.36) for communication, 4.20 (SD=1.29) for vision, 4.16 (SD=1.13) for management of government affairs, and 4.12 (SD=1.32) for political power. We found that all five factors measured were somewhat higher than the average of 3.5 points.
In addition, we conducted a correlation analysis (Table 2) to examine the interrelationship between the dependent and independent variables and the control variables. If we look at the correlation between the major variables, we can see that the dependent variable, trust in government, used as a means of evaluation for the independent variable, presidential leadership, showed a statistically significant positive correlation (p<.001) for all five qualities (vision: r= .306, p= .000, communication: r= .289, p= .000, management of government affairs: r= .299, p= .000, political power: r= .302, p= .000, integrity: r= .281, p= .000).
On the other hand, the correlation coefficients between independent variables (.70-.77) were a little high, but the variance inflation factor (VIF) for each variable was 1.02-5.45 (mean VIF=2.25), so multicollinearity between variables was not found to be problematic (Agresti & Finlay, 2009). Furthermore, the Breusch-Pagan test showed that the error term was homoscedastic, since the chi-square value was 3.01 and the significance level was .083, which is not statistically significant (p<.05); that the variance of the error term was constant, fits the basic assumptions of the regression model (Wooldridge, 2018).
Hypothesis Testing
Table 3 presents the results of multiple regression analyses to test the hypothesis that presidential leadership is an influential factor in trust in government and shows the relationship between substantial independent variables that can affect the public’s trust. We also used a standardized regression coefficient to compare the relative importance of each of the presidential leadership qualities.
The overall model has an explanatory power of approximately 12% (R2 =.126, adjusted R2 =.124), and the resulting relationships between crucial variables are as follows. As predicted, the analysis showed that the presidential leadership qualities do affect trust in government.
Specifically, first, the coefficient of vision was positive and statistically significant (p<.000). The 95% confidence interval, which is relatively conservative in testing statistical significance and contains a lot of information (e.g. center of the confidence interval value), is suitable for the population mean and effect size of the regression coefficient (Cohen, 1994; D. K. Lee, 2016; Wooldridge, 2018). Since the 95% confidence interval does not include zero, it is a statistically significant positive relationship. Moreover, the effect size measured by the regression coefficient ranged from .035 to .078, which is relatively high compared to the coefficients of the other independent variables. Therefore, for one standard deviation increase in vision, trust in government is increased by 0.101, with all other variables held constant. Second, communication skills also show a positive and statistically significant (p<.006) result. The 95% confidence interval did not contain zero, and the effect size range from .008 to .051 is the smallest when compared to the coefficients of the other independent variables. Therefore, if the level of communication skills increases by one unit, people’s confidence in the government increases by 0.056. Third, the coefficient for management of government affairs is likewise positive and statistically significant (p<.001). Once again, the 95% confidence interval did not contain zero, and the effect size was found to be the third largest when compared to the coefficients of the independent variables, which was significant in the range of levels from .015 to .059. Thus, with a one standard deviation increase in managing government affairs, the public’s trust in government increases by 0.067. Fourth, the direction of the modification of the coefficient of political power was positive and statistically significant (p<.001). The 95% confidence interval did not contain zero, and the effect size is found to be the second largest among the other independent variables, ranging from .016 to .060. Therefore, with a one standard deviation increase in the assessment of political power, the public’s trust in government increased by 0.069. On the other hand, integrity did not produce statistically significant results. However, in the descriptive statistics, the mean of the integrity variable was the highest among the other presidential leadership qualities, although prior studies suggest that its impact on trust in government is low and also do not identify a direct relationship between integrity and trust in government (Cho & Nam, 2007; Citrin & Green, 1986). This result is addressed in the discussion and conclusions, where we consider the theoretical implications of our analysis.
Consequently, based on the multiple OLS regression analysis estimation, our hypotheses were supported only in the case of vision, communication skills, management of government affairs, and political power. In other words, the postulation that presidential leadership qualities will have a positive relationship with trust in government is only partly supported. Moreover, among the four presidential leadership qualities that showed statistically significant results, the relative magnitude of the relationship was the highest with respect to vision (β=.101, 95% CI: .035-.078), followed by political power (β=.069, 95% CI: .016-.060), management of government affairs (β=.067, 95% CI: .015-.059), and communication skills (β=.057, 95% CI: .008-.051).
DISCUSSION AND CONCLUSIONS
Our analysis of the relationship between trust in government and presidential leadership qualities confirmed that the ability to articulate a vision, communication skills, management of government affairs, and political power showed a significant and positive relationship with trust in government. Furthermore, the standardized regression coefficient confirmed that the order of importance of these qualities is vision, political power, management of government affairs, and communication skills. Therefore, it partially supported the research hypothesis.
The theoretical implications of our study are as follow. First, presidential leadership emerges as a relevant measurement tool in assessing citizens’ trust in government. Prior studies of determinants of trust in government were mainly focused on behavioral variables and demographic background factors (Bae & Lim, 2009b). Also, despite the relationship of crucial variables that needs to be verified theoretically and practically, few prior studies have been conducted due to the limitations of measurement tools in measuring the relationship between public awareness and evaluation of presidential leadership qualities. In a similar vein to this study, the particular study by Cho and Nam (2007), yielded results that the relationship between individual characteristics of the president and trust in government was not statistically significant. On the other hand, we narrowed down the scope to specify the trust in government variable and presented presidential leadership based on the situational-qualities theory. The analysis confirmed as a measurement tool and a clear indicator of new determinants of trust in government were presented.
Second, discussions about integrity leadership that was not statistically significant in our study. Previous studies have discussed that the integrity of elected officials such as the president can be applied as a crucial factor of trust in government (Laustsen & Bor, 2017; McAllister, 2016; Storey, 2004). Moreover, lack of integrity, or power-related irregularity for the president and government leaders, was pointed out as the cause of the decline in government confidence (Jun et al., 2013). However, determination of trust in government, in the presidential leadership aspect, our findings suggest that rather than ethical issues such as integrity, qualities associated with the ability to perform presidential duties were relatively more crucial. According to the Citrin and Green’s (1986) study, the president’s integrity (e.g. President Ford and Carter) had a relatively lower impact on government confidence than other leadership abilities. Moreover, President Reagan received unfavorable reviews from the public in terms of integrity and morality, but his strong leadership and inspirational characteristics enhanced government confidence (Cho & Nam, 2007; Citrin & Green, 1986). In addition, at the beginning of the president’s term, the integrity of President Roh Moo-hyun is applied as the most influential factor in the presidential trust, but there was no significant impact on trust in government (Cho & Nam, 2007).
Our study also confirms that the president’s leadership can increase the public’s confidence in government. As we mentioned before, trust in government is a comprehensive recognition and assessment of the government affairs management, which has a purely functional aspect that helps people gain more support in policy-making and implementing policies (Hetherington, 1998; C. W. Lee & Cho, 2016). Also, in countries with a presidential system such as South Korea, in which the president has a substantial amount of authority in policy making, presidential leadership is more important than any other factor in improving trust in government (Hahm, 2007; Jun et al., 2013). Furthermore, even though the qualities we investigated appear to have less impact than the others explored in various studies of trust in government, they have also been proven to be important in prior studies (Citrin, 1974; Citrin & Green, 1986; Jun et al., 2013; C. W. Lee & Cho, 2016). Our study further suggests that presidents may be able to enhance trust in government by paying attention to the ability to articulate a vision to the public because with respect to vision, the standardized regression coefficient (β=.101) is relatively high in the direction of a positive relationship with confidence in government compared to other leadership factors. That is, the ability to articulate a national vision is relatively important compared to other leadership qualities.
As studies on the relationship between trust in government and presidential leadership qualities are unusual in Korea, our study has the meaning of filling that gap. In addition, the fundamental reasons that led to the impeachment of former President Park Geun-hye in 2017 were the lack of communication and corruption of the entourage. However, based on the data surveyed in 2019, our study confirms that vision, political power, and management of government affairs were regarded as important qualities. While communication skills were seen as least important and integrity was not statistically significant, these results show slightly different from the cause of impeachment in 2017. Such results could be interpreted that issues such as lack of communication and corruption of close aides are inconsequential problems from the perspective of the public nowadays.
Our study does have limitations. First, we were not able to construct and measure multiple items of the questionnaire for the purpose of measuring presidential leadership variables: this means we were unable to fully evaluate the accuracy, reliability, and validity of the data in line with the goals of the study. Also, political factors that could have been used as control variables such as the increasing number of political scandals or media focus on political corruption were not included because of the secondary nature of the survey data. Empirical research using direct survey data that sets multiple items of the questionnaire would address this problem.
Second, there may be a reverse causality problem in the relationship this study suggests, particularly that between government trust and presidential leadership qualities. Even though our analysis satisfies the underlying assumptions in regression analysis and we tested the hypotheses through OLS multiple regression analysis, and prior scholarship indicates that presidential leadership is a determinant of confidence in government, endogeneity between trust in government and presidential personal leadership qualities is possible (S. K. Choi, 2018; Jun et al., 2013). Therefore, future research should take into account the fact that the estimated value of the regression coefficient can result in overestimation (Ko, 2019) and should reverse causation by setting and measuring the instrumental variable.
Third, the problem of common method bias in the data utilized in this study is inherent because it derives from questionnaires that ask abstract questions about the degree of trust in government agencies and social leaders, as well as the level of presidential leadership. In this study, Harman’s single-factor test (1976) was performed to check the common method variance (Podsakoff, 2003): the results are presented in Appendix B. According to the analysis results, the problem of common method variance was not that serious. The dominant general factors were not found as the four factors’ eigenvalues are greater than one and factor 1 accounts for 30.7% of the variance, less than the threshold of 50% (Harman, 1976). However, correlations may be the result of respondents’ social desirability bias (Baek, 2012). Also, studies measured as cross-sectional data show that dependent and independent variables are measured at once. Therefore, problems with a type 1 error caused by inflated correlation coefficients between variables or a type 2 error caused by reduced correlation coefficients can occur (Baek, 2012; Jakobsen & Jensen, 2015).