Motivated to Share? Using the Person–Environment Fit Theory to Explain the Link between Public Service Motivation and Knowledge Sharing

: Despite the growing evidence that public service motivation (PSM) is an important source of employee outcomes, there is still a dearth of empirical evidence on whether it stirs one’s willingness to share learned knowledge with other members of an organization. The authors explore the mediating role of workers’ perceived ﬁt in a given setting as a causal mechanism through which PSM promotes knowledge-sharing indirectly. Findings based on the primary data of 1048 occupationally diverse employees working in 33 local governments in South Korea show that PSM has a positive impact on knowledge sharing, person–group (P–G) ﬁt, and person–job (P–J) ﬁt, and P–G ﬁt and P–J ﬁt plays a mediating role in improving knowledge sharing, respectively, while person–supervisor (P–S) ﬁt does not. This study interprets these results as suggesting that P-S it should be considered independently from environmental factors in the public organization, as it is derived from the interrelationship between employees and supervisors that began in the hierarchy of the organization, and thus is difﬁcult to be related to other factors. Based on literature review and empirical analysis, this study presents theoretical and policy implications.


Introduction
In the long term, the sustained competitiveness of an organization is related to its ability to create, share, and utilize innovative knowledge. Considering the long-term perspective, knowledge sharing (KS) is an indispensable prerequisite for the creation and application of intellectual resources in an organization. In recent studies, KS has been found to be a key factor having a direct effect on organizational performance [1] and the diffusion of innovative ideas and creativity [2]. In addition, knowledge sharing, such as knowledge-centered activities, is a vital resource in an organization because it enables employees to maximize the organization's capability and competitive advantage in the public sector and to generate solutions and efficiencies [3].
Despite the importance of knowledge-sharing activities, it has not been thoroughly studied in public administration. Most previous studies about KS focused on antecedents for improving knowledge-sharing activities in the public sector [4][5][6]. Recently, some studies tried to understand the public sector's motivation to share knowledge based on its unique form of intrinsic motivation, Public Service Motivation (PSM) [7][8][9]. However, considering that the studies focused only on the direct relationship between PSM and knowledge sharing, it is still worthwhile for PSM to understand the mechanisms that influence knowledge sharing.
This study considers person-environment (P-E) fit as a mechanism through which PSM promotes knowledge-sharing behavior. Since P-E fit is defined as the degree of compatibility between an individual and environmental characteristics [10], P-E fit itself involves the concept of the work environment. Therefore, it is meaningful to investigate KS through the lens of P-E fit by examining the work environment and the relationship between P-E fit and KS. Considering the relationships between PSM, P-E fit, and KS, the research questions were developed as follows: Do individuals with higher levels of PSM have a higher propensity toward KS? Does the congruence between employees and their work environment increase employees' knowledge-sharing behavior? Does P-E fit theory help explain the causal relationship between PSM and KS?
The majority of studies about P-E fit have focused on its effect on various work outcomes [10,11]. Research on P-E fit generally supports the idea that a high level of P-E fit is related to a number of positive individual and organizational outcomes. By and large, P-E fit theory states that workers who feel fitted to an organization will show a higher level of individual outcomes, such as higher job satisfaction and organizational commitment because they share values, personality, or goals with others [12,13]. Therefore, in light of the significant effects of P-E fit on organizational outcomes, considering P-E fit in this study enables us to have broad perspectives on understanding the relationship between PSM and KS. As a theoretical background, the researchers apply self-determination theory as an overarching framework of the current study.

Conceptualization of KS
KS is defined as the activities by which employees "share their work-related experience, expertise, know-how, and contextual information with other employees through informal and formal interactions within or across teams or work units" [4] (p. 371). It means that KS provides a frame for the combination and evaluation of information and new experiences. Therefore, employees can exchange knowledge with other members and convert it into an asset and resource in their organizations.
KS can be understood in an organizational context as the provision of task information and the knowledge of how to help and collaborate with others, solve problems, develop new ideas, and implement policies or procedures [14]. It also can be viewed as one part of a socially interactive culture involving the exchange of employees' knowledge, experiences, and skills throughout a whole department or organization. It comprises capturing, organizing, and transferring experience-based knowledge that resides within the organization and making that knowledge available to others in an organization [15]. In this context, since KS is a process among people, it is considered to consist in mutual understandings among members [16]. As reviewed in previous studies, KS is closely associated with employees in the public sector workplace. In other words, KS can occur when an employee is willing to share his or her own knowledge with other members in the organization [1]. Therefore, this study understands KS on a personal level by considering employees' perspectives.
Consequently, an organization that knows how to share individual knowledge within itself is more likely to succeed, and employees who share their knowledge and experiences are treated as more valuable members of that organization [17]. Moreover, considering organizational outcomes, KS improves employees' performance so that they can accomplish their tasks better, more quickly, or more efficiently [15]. It also enables organizations to enhance innovation performance and to reduce redundant learning efforts [18]. In summary, KS enhances the innovation of an organization, facilitates information exchange within an organization, improves the quality of individual work, and reduces the time required for problem-solving.

PSM and KS: Application of Self-Determination Theory
Self-determination theory (SDT) fundamentally presumes that people have three basic needs, namely autonomy, belongingness, and competence, and that they have selfmotivated behavior to satisfy these needs [19]. Self-determination is "the process of utilizing one's will" [20] (p. 26), and self-determined behavior is defined as "the attitudes and abilities required to act as the primary causal agent in one's life and to make choices regarding one's actions free from undue external influence or interference" [21] (p. 305). That is, depending upon the extent to which people are self-determined to satisfy these needs, they could have three types of motivation, amotivation, extrinsic motivation, and intrinsic motivation [19].
SDT provides a broad perspective for understanding the relationship between PSM and employees' knowledge-sharing behaviors. Among the three types of motivations in SDT, intrinsic motivation is regarded as the most important variable for predicting employees' knowledge-sharing behavior [22,23]. When public employees find meaning in their jobs, their intrinsic motivation can be enhanced to help others, for example, by engaging in knowledge-sharing behavior, because the behavior is enjoyable, personally meaningful, and fits their value system [24]. They can maintain their motivation in the workplace without any external sources such as reward or support [25].
PSM is defined as "an individual's predisposition to respond to motives grounded primarily or uniquely in public institutions and organizations" [26] (p. 368). PSM can be considered as the intrinsic motivation of public employees [27,28]. Knowledge-sharing behavior is less likely to be motivated by external sources, because it is considered to be a prosocial behavior based on voluntary contribution [24]. In other words, intrinsic motivation plays a key role in enhancing knowledge-sharing behavior in an organization. Intrinsic rewards are derived from the satisfaction achieved by performing meaningful work well. In light of this perspective, public employees place a high value on helping other people, serving the public, and doing good for society. In fact, they are more intrinsically motivated and committed to serving the public than others [26,29], so PSM has a positive effect on their prosocial behaviors, such as organizational citizenship behavior [30]. KS is considered as one of such prosocial behaviors, and intrinsic motivation is positively associated with KS [24]. Thus, the following hypothesis is explored: Hypothesis 1 (H1). PSM has a direct and positive effect on KS.

PSM and P-E Fit
As noted above, P-E fit is generally defined as "the compatibility between an individual and a work environment that occurs when their characteristics are well matched" [10] (p. 281). The concept of P-E fit has five sub-dimensions such as person-vocation fit (P-V fit) person-job, (P-J fit), person-organization fit (P-O fit), (4) person-group (or team) fit (P-G fit), and person-supervisor fit (P-S fit), and P-V fit and P-J fit are from the job perspective whereas P-G fit and P-S fit concern interpersonal relations [10]. The current research focuses on P-J fit, P-G fit, and P-S fit.
Several PSM studies have dealt with P-E fit theory. However, they have only focused on P-O fit and P-J fit as playing a mediating or moderating role between PSM and organizational outcomes [31][32][33]. That is, direct relationships between PSM and those fit frameworks have not been investigated. Although both theoretical background and empirical research on P-E fit are rare in PSM studies, this study took some cues from previous studies to understand the relationship between PSM and P-E fit. Employees with high levels of PSM regard their values as congruent with the values of the organizations they work for [26,34]. In addition, PSM-driven employees are not only willing to risk personal loss to help others and to work for the public [35] but also are more likely to engage in prosocial behaviors [30], which is based on performers' altruism and generalized compliance. In that sense, although employees may perceive their fit with their work environment as not suitable, they might try to work actively to match themselves to that work environment. Therefore, based on employees' internal motivation and willingness to fit into the work environment, we can assume that PSM appears to be positively related to P-E fit.
Moreover, considering both SDT and PSM theory, since highly PSM-driven employees in the public sector are less motivated by extrinsic rewards, such as promotion and high monetary rewards, this study considers internal factors, such as relationships or the compatibility of employees and their work environment-as in some previous studies [26,28,34]. Indeed, Perry and Wise [26] insisted that "the greater an individual's public service motivation, the more likely the individual will seek membership in a public organization" (p. 370). Moreover, from a broader perspective, as Perry and Wise's [26] statement includes sub-factors of the work environment such as coworkers, job tasks, and organizations, PSMoriented employees prefer and pursue congruence with their work groups, coworkers, job tasks, and supervisors. In light of this, we hypothesized a positive relationship between PSM and the sub-dimensions of P-E fit, which are P-G fit, P-J fit, and P-S fit.
Hypothesis 2a (H2a). PSM has a direct and positive effect on P-G fit.
Hypothesis 2b (H2b). PSM has a direct and positive effect on P-J fit.
Hypothesis 2c (H2c). PSM has a direct and positive effect on P-S fit.

P-E Fit and KS
Knowledge is essential to organizations, but the management of knowledge receives little attention in public administration. Instead, in information studies and business administration, KS is considered one of the most important activities, and many factors affect KS, such as leadership, organization structure, and trust [36]; vision, goals, and social networks [4]; and leadership and information [5]. All of these factors have been studied as potential ways to increase employees' willingness to share knowledge with other members within an organization.
Despite the lack of research, a few scholars focused on the importance of P-E fit in knowledge management. For example, Argote et al. [37] recognized the importance of P-E fit in organizations and insisted that P-E fit can influence knowledge management outcomes. They found that "when properties of units, properties of relationships, and properties of knowledge fit or are congruent with each other, knowledge retention, and transfer increase" (p. 580). Namely, P-E fit theory dealt with the organizational culture by focusing on the suitability of properties with a work environment that a good match between employees and their work environment increases KS. Based on this theory, the author can assume that P-E fit is positively associated with KS. However, it only recognized that P-E fit influences knowledge-related behaviors from a general perspective, and research on the relationship between P-E fit and KS is still relatively rare. Therefore, it is worthwhile to empirically examine the relationship between those two factors.
In this study, P-E fit was categorized into three sub-dimensions: person-group (P-G) fit, person-job (P-J) fit, and person-supervisor (P-S) fit. P-G fit occurs when an employee is matched well with his or her work group [10]. Since P-G fit focuses on interpersonal compatibility with coworkers in a work group [11], it is distinguished with the concept of a demographic relationship with coworkers [38]. Therefore, P-G fit is related to multiple factors, such as beliefs, characteristics, and abilities. KS can be sensitive and rare because employees are reluctant to share information they have. However, employees are likely to share knowledge when they think it is necessary for completing tasks assigned to them in their organizations [39]. Furthermore, when employees perceive with their work groups or coworkers and when they need to share abilities in order to achieve common goals, they become willing to share their knowledge with others [40]. Therefore, based on the perspective of P-G fit, when employees feel that their characteristics fit their work groups' task demands, they are willing to share knowledge in those groups. P-J fit has been considered to be important factor in knowledge transfer research [37] and is the most widely studied type of P-E fit [41]. It focuses on the congruence between an employee's characteristics and his or her job or tasks. Within the context of a career, P-J fit is conceptualized based on two perspectives: demands-abilities fit and needs-supplies fit [42]. The former perspective considers the fit between an employee's abilities, such as knowledge and skills, and job demands [43]. The latter perspective refers to the fit between the needs and desires that an employee has and what is provided by the job [44]. Considering these perspectives, P-J fit provides a useful measure for determining how satisfied employees are with their tasks and job demands. Indeed, individuals who are satisfied with their jobs are more likely to engage in prosocial behaviors, such as organizational citizenship behavior and KS [45]. In particular, public employees with high levels of P-J fit are more likely to share their knowledge with co-workers because they may learn something from other co-workers, resulting in the feeling of competency (demands-abilities fit). Moreover, sharing information behaviors may satisfy their needs for job competency since they can provide and get the information on job-related skills when they need it (needs-supplies fit).
A number of P-S fit studies have demonstrated that when subordinates perceive their supervisors as similar in terms of demographics or personalities, it has positive effects on work outcomes [10,46]. However, like P-J fit, the congruence between subordinates and supervisors has not been extensively discussed in knowledge sharing studies [1]. P-S fit is based on employees' perception of similarities with their supervisors and on the quality of the relationship between subordinates and supervisors [46]. Employees' perception is positively associated with leader-member exchange (LMX) theory, which focuses on dyadic relationship between leaders and followers [47]. Moreover, P-S fit is more interpersonal and less task-oriented than other types of P-E fit. Considering the significant relationship between P-S fit and LMX, mutual respect and trust are important factors for understanding P-S fit. Indeed, the employees' perception of a good fit with their supervisors is related to how much those employees trust and respect them [46], and trust between subordinates and supervisors positively influences employees' knowledge-sharing behavior [48]. Considering the theoretical linkages for understanding the relationship between the three sub-dimensions of P-E fit and KS, the following hypotheses were proposed: Hypothesis 3a (H3a). P-G fit has a direct and positive effect on KS.
Hypothesis 3b (H3b). P-J fit has a direct and positive effect on KS.
Hypothesis 3c (H3c). P-S fit has a direct and positive effect on KS.

The Mediating Role of P-E Fit on a Relationship between PSM and KS
Some studies have seen P-O fit theory as having a mediating effect on the relationship between PSM and job satisfaction [28,49]; between PSM and work attitudes, such as organizational commitment and job satisfaction [31]; and between PSM and job performance [32]. On the other hand, other researchers have examined the moderating effect of P-J fit and P-O fit on the relationship between PSM and willingness to work in the public sector [33]. As reviewed, such researchers have a limited perspective of P-E fit in that they primarily have considered P-O fit and P-J fit separately in different research. For understanding the indirect effects of P-G, P-J, and P-S fits between PSM and KS, we focused on PSM theory and comprehensive perspective of P-E fit so that the current study can cover the sub-dimensions of P-E fit.
According to previous empirical evidence, employees with high levels of PSM are attracted to sharing values with their organization and coworkers [33], and they are more likely to be compatible with their organizations because they complete their job tasks and achieve better outcomes [26,50]. Specifically, employees with higher PSM levels are willing to fit into their organizations and work environments because they put a high value on matching themselves with their organizations and on better organizational outcomes. Al-though employees might feel that the interests and characteristics of coworkers, demands and goals of their organizations, and values of supervisors differ from their own, individuals with high levels of PSM are likely to take pains or losses to achieve better compatibility with the different types of objects in the workplace. One way they might do this is by voluntary behavior intended to help others. Considering theoretical frameworks for the indirect relationship between PSM, the sub-dimensions of P-E fit, and KS, the following hypotheses were developed: Hypothesis 4a (H4a). PSM has an indirect and positive effect on KS through a positive influence on P-G fit.

Data and Measures
This study used the self-administered survey method to empirically verify the research hypotheses. A survey was conducted on 33 local government agencies in South Korea. Data were collected from May to August 2017. While survey-based studies commonly collect data by using emails or mails, researchers visited local government agencies to distribute survey questionnaires and explained the purpose and anonymity of data collection to participants before participants responded to the survey. Initially, 1420 questionnaires were distributed to those public agencies, and a total of 1094 of them were returned, yielding a response rate of 77%. Considering missing data or insincere answers of some respondents, 46 were excluded, and 1048 questionnaires were used for the final analysis.
The survey comprised questions for measuring employees' behaviors and perceptions of their work environments and demographic information. Except for demographic variables, most variables were measured on a five-point Likert-type scale ranging from 1 to 5, where a value of 1 corresponded with "strongly disagree", and a value of 5 cor-responded with "strongly agree." Table 1 presents the descriptive statistics for all the variables measured in the survey.

Dependent Variable
In order to measure KS, this study used four items adapted from Ford and Staples [51], such as "I share my work reports and official documents with members of my organization"; "I provide my manuals, methodologies, and models for members of my organization" (Cronbach's α = 0.78).

Independent Variable
To measure PSM, this study adopted a 16-item measure, which is appropriate for international contexts [52], including "I admire people who initiate or are involved in activities to aid my community"; "I think equal opportunities for citizens are very important"; "I feel sympathetic to the plight of the underprivileged"; "I am prepared to make sacrifices for the good of society" (Cronbach's α = 0.93). In Kim et al.'s [52] research, PSM was classified into four sub-dimensions: attraction to public service, commitment to public values, self-sacrifice, and compassion. While PSM is typically measured by considering sub-dimensions in several studies [32,52], it is measured as one factor by calculating the overall PSM score [53,54]. Considering that the latter uses the sum scores of PSM and also gives equivalent results, it is a much simpler way than using all sub-dimensions of PSM. Moreover, because we were interested in examining PSM as a whole, in this study, the four observed indicators were considered as a single factor, and responses were not treated separately by each of the four categories.

Mediating Variables
For the current study, P-E fit is divided into three sub-dimensions, which are P-G fit, P-J fit, and P-S fit. Each fit is also defined as the compatibility between an employee and coworkers, his or her job demands, and a supervisor, respectively. To measure the sub-dimensions of P-E fit, a total of 13 items established in previous studies were used. Specifically, this study measured P-G fit with the five items of Kristof-Brown et al. [10], such as "My approach to work fits in with that of my work unit members", "My professional interests are the same as those of my work unit members", and "I identify with my work unit members" (Cronbach's α = 0.88); P-J fit with the four items of Lauver and Kristof-Brown [55], such as "My abilities fit the demands of this job" and "There is a good match between the requirements of this job and my skills" (Cronbach's α = 0.91); and P-S fit with the four items of Chuang et al. [56], such as "The things I value in life and the things my supervisor values are matched" and "My personality and my supervisor's personality are matched" (Cronbach's α = 0.88).
Since only the items derived from Chuang et al. [56] are provided as an interrogative sentence with the response format from "no match" to "complete match", for preventing any confusion among respondents regarding different formats of survey questions, this study changed the form of these items to a declarative sentence with the same response form as the others, which use a scale from "strongly disagree" to "strongly agree." For example, one of the items for measuring P-S fit, "How would you describe the match between the things you value in life and the things your supervisor values?", was changed to "The things I value in life and the things my supervisor values are matched."

Control Variables
This study considered work environment variables such as job autonomy and organizational learning culture and demographic variables as control variables. Previous studies examined the relationships between job autonomy and KS and between organizational learning culture and KS. In one study, a high level of employees' perception of autonomy had a positive effect on their willingness to share knowledge with coworkers in their organization [57]. In another, an organizational learning culture had a positive effect on KS [58]. In light of those findings, job autonomy and an organizational learning culture were included as variables in the proposed research model. In order to measure those variables, this study used four items adapted from Beehr [59], such as "I have a lot of say over what happens on my job" and "My job allows me to make a lot of decisions on my own", and four items from Marsick and Watkins [60] and Watkins and Marsick [61], such as "My organization creates continuous learning opportunities" and "My organization promotes inquiry and dialogue", respectively. In addition, survey respondents were also asked to provide their demographic information, including age (1 = under 29, 2 = 30-39, 3 = 40-49, 4 = 50-59, and 5 = over 60), length of service in the current organization (1 = 10 years or less, 2 = 11-20 years, 3 = 21-30 years, 4 = more than 30 years), gender (0 = male, 1 = female), and annual salary (1 = less than $10,000, 2 = $10,000-$20,000, 3 = $20,001-$30,000, 4 = $30,001-$40,000, and 5 = over $40,000).

Confirmatory Factor Analysis and Model Fit Diagnostics
Confirmatory factor analysis (CFA) was conducted in order to determine whether the proposed measurement model in this study fits the data collected. To determine the goodness of fit, scholars suggest particular cut-off values for the fit index. For a good fit, the p-value must not be statistically significant (p > 0.05) in the chi-square test [62]. However, the chi-square statistic is highly sensitive to sample size and is only used as a basis for acceptance or rejection in certain situations [63,64]. Due to the sensitivity of the chi-square, researchers considered an alternative index, the relative/normed chi-square (x 2 /df), to assess model fit. The relative/normed chi-square must be less than 5.0 for an acceptable fit [65]. Other relevant values for a good or an acceptable fit are that CFI and NNFI must be larger than 0.9 [66], that SRMR must be less than 0.08 [67], and that RMSEA must be smaller than 0.08 [62].
Considering the thresholds mentioned above and the results of the CFA, values of indices (except for the chi-square test) present a good fit, so the proposed model fits the data well (x 2 (754) = 2785.210 (p < 0.001), x 2 /df = 3.694, CFI = 0.930, NNFI = 0.920, SRMR = 0.024, RMSEA = 0.050). Based on these results, this study confirmed that the measurement model met the desired standards of reliability and validity. Table 2 presents the bivariate correlations between key variables used in the current study. PSM, P-G fit, P-J fit, and P-S fit have positive and significant correlations with KS. As the researcher proposed in the research hypotheses, the correlation coefficients between PSM and the sub-dimensions of P-E fit are statistically significant. In addition, PSM, P-G fit, P-J fit, and P-S fit are positively correlated with job autonomy and organizational learning culture. Among demographic variables, age and length of service are positively correlated with KS, while gender is negatively correlated with KS. Therefore, the correlations between the key variables in this section support the research hypotheses. As a result, the researcher tested the direct effects of PSM and the mediating effects of P-G fit, P-J fit, and P-S fit further by running a regression analysis.  Table 3 provides the results of multiple regressions to check the direct causal relationships between key variables. Model 1 tests the relationship between PSM and KS. As a result shows, the independent variable is significantly related to KS. The standardized coefficient of PSM is 0.203, and it is positively associated with KS (p < 0.001). The result is consistent with previous research [7] and supports hypothesis 1 in this study. Model 2 tests the relationship between sub-dimensions of P-E fit and KS. Model 2 can provide preliminary evidence that P-E fit subscales work as mediators in the research model. In the results for Model 2, P-G fit and P-J fit show a standardized regression coefficient of 0.216 and 0.186 (p < 0.001), respectively, whereas P-S fit is not statistically significant with KS. Therefore, hypothesis 3 is supported partially due to the rejection of hypothesis 2c. Model 3 shows that all the presumptions between PSM and sub-dimensions of P-E fit in this study are not proved. Specifically, PSM is positively associated with P-G fit (β = 0.216, p < 0.001) and P-J fit (β = 0.186, p < 0.001), while it is not significantly related to P-S fit. Therefore, unlike the relationship between PSM and P-S fit, PSM is significantly and positively related to P-G fit and P-J fit, supporting hypotheses 3a and 3b.

Regression Analysis with Testing Hypotheses
Hypothesis 4 focused on the indirect effect of the independent variable, PSM, namely the mediating roles of P-E fit sub-scales on the relationship between PSM and KS. To test the hypothesis, the researcher used the PROCESS macro, which utilizes the bootstrapping method to calculate the mediation effect.   Table 4 shows the indirect effect of PSM on KS through the sub-dimensions of P-E fit. Regarding the indirect effect of PSM on KS through P-G fit and P-J fit, two individuals that differ by one unit on PSM are estimated to differ by 0.040 and 0.018 on KS as a result of the effect of PSM on P-G fit, and P-J fit, in turn, affects KS. The columns of the bootstrap confidence intervals make it possible to check an indirect effect of PSM on KS by considering each mediator based on the statistical significance between variables at p < 0.05. Overall, since the confidence intervals for the indirect effect of PSM on KS through P-G fit and P-J fit are entirely above zero ([0.022, 0.065] and [0.006, 0.034] respectively), it is clear that the relationship between PSM and KS is mediated by P-G fit and P-J fit. Unlike P-G fit and P-J fit, the confidence intervals for the indirect effect of PSM on KS through P-S fit include zero ([−0.010, 0.002]). That is, the relationship between PSM and KS is not mediated by P-S fit. Therefore, given the results, hypotheses 4a and 4b are accepted, while hypothesis 4c is rejected. Work environment variables (i.e., job autonomy and organizational learning culture) and demographic variables (i.e., age, length of service, gender, and annual salary) were controlled. IV = independent variable, MV = mediating variable, DV = dependent variable, IE = indirect effect, Boot SE = bootstrap standard error, Boot LLCI = bootstrap lower limit confidence interval, Boot ULCI = bootstrap upper limit confidence interval, P M = ratio of indirect to total effect.
In addition, Table 4 provides the ratio of the indirect effect of each mediator to the total effect of PSM on KS (P_M). In the proposed model, considering P_M through P-G fit (P_M1 = 0.203) and P-J fit (P_M2 = 0.089), total indirect effects account for almost 30% of the total effect of PSM on KS. Finally, P_M through P-S fit (P_M3 = 0.006) is relatively smaller than P-G fit and P-J fit because P-S fit does not mediate the relationship between PSM and KS.

Discussion and Conclusions
Using self-determination theory and P-E fit theory, this study investigated the role of perceived fit in the organization as a causal mechanism linking PSM to KS behavior in the public sector. Although prior to this study a direct connection between PSM, KS, and sub-dimensions of P-E fit had not been considered in public administration, the current study fills that research gap by developing a logical framework based on the literature review and by examining the relationships empirically. However, unlike the expectation of the current study regarding the research hypotheses proposed, some of them were not supported by the results of the statistical analyses. Based on the regression results, the relationship between PSM and KS was statistically significant. PSM was positively related to P-G fit and P-J fit and not significantly associated with P-S fit. P-G fit and P-J fit had a positive effect on KS, while P-S fit was not significantly related to KS. Finally, mediation analysis results showed that P-G fit and P-J fit have a mediating effect on the relationship between PSM and KS, but P-S fit does not. That is, research models based on predictions regarding P-S fit were not accepted, while other models considering P-G fit and P-J fit were supported. As this study looked over a different facet of P-S fit with other dimensions of P-E fit, some points need to be discussed here.
The literature review reveals PSM is positively related to P-E fit [26,34], and P-E fit plays a significant mediating role [31,32,49]. Indeed, employees with high levels of PSM are more likely to be compatible with their organizations, causing them to be more likely to complete their tasks and to achieve better outcomes [26,50]. Therefore, this study assumed that PSM-oriented employees would pursue congruence with their work groups, coworkers, job tasks, and supervisors, and then they have more willingness to share learned knowledge with other members in the organization. However, P-S fit should be distinguished from other types of P-E fit, and the relationship between PSM and P-S fit should be understood differently.
P-E fit theory states that workers who feel fitted to an organization will show a higher level of individual outcomes, such as higher job satisfaction and organizational commitment, because they share values, personality, or goals with others [12,13]. However, building or maintaining high congruence between subordinates and their supervisors differs with other types of P-E fit. In other words, subordinates and leaders do not have to be similar in their personality, personal values, and goals, and they do not have to share those factors, although they still must want to achieve better performance [68] because they can complement each other when necessary [11]. Moreover, because P-S fit is based on interpersonal relationships [69], it can be directly derived from mutual relationships between subordinates and their supervisors. In this context, therefore, PSM may not be a reliable factor for predicting P-S fit, and P-S fit may not explain employees' knowledgesharing behavior.
It is interesting to note that this study used data collected in South Korea, whose culture is primarily based on Confucianism. Confucianism emphasizes hierarchical order in human relationships [70], a belief that may increase the rigidity of organizations. Therefore, this belief system could hinder active exchange and communication between subordinates and supervisors and also result in subordinates blindly obeying authority [71]. Considering the top-down culture of Confucianism, the subordinate-supervisor relationship can be formed with a different mechanism. Therefore, P-S fit in the Korean context can be distinguished from other types of P-E fit. This feature of P-S fit was proved empirically by the primary analyses, which revealed that PSM did not predict P-S fit.

Theoretical Implications
This study has significant implications from a theoretical perspective. First, this study considered multiple dimensions of P-E fit relating to several different facets, such as coworkers, job demands, and supervisors in the work environment. While previous studies have typically depicted P-E fit as a singular concept [28,72] or have dealt with only one dimension of P-E fit [31,32,73], the current study examined the differential effects relating to three dimensions of P-E fit: P-G fit, P-J fit, and P-S fit. In addition, considering that very few studies in public administration have discussed P-S fit [74], the empirical evidence provided by this study regarding P-S fit in the public sector context is worthwhile.
Second, this study applied particular theories to understand work environments in the public sector. Specifically, it tried not only to understand the knowledge sharing motivation process in the public sector but also to examine the relationships between PSM, the sub-dimensions of P-E fit, and KS by applying both SDT and P-E fit theory as theoretical backgrounds. Considering the naturally high intrinsic motivation of public employees to serve the public [26,29], SDT was helpful for understanding the relationship between intrinsic motivation and employees' prosocial behavior. In addition, considering that work environment and organizational culture promotes KS [16], and P-E fit itself involves the concept of the work environment, P-E fit perspective was valuable to explain the knowledge sharing motivation mechanism by focusing on compatibility between an employee and their work environment.
Third, this is the first study that applied P-E fit theory to the knowledge sharing motivation process in the public sector. Although some studies have investigated the relationship between PSM and knowledge sharing behavior in public sector organizations [7][8][9], they were limited to examine a direct relationship and did not explain how PSM affects KS. Specifically, the present study disentangled the relationship based on P-E fit theory. Furthermore, considering sub-dimensions of P-E fit, this study provided theoretical insights into individual and organizational mechanisms to determine how and why PSM leads to KS.

Practical Implications
From a practical perspective, the current study provides several implications for managers in the public sector. Since KS is understood as a prosocial behavior [24], it cannot be mandated or required by external force. Therefore, organizations should hire people who meet the organizational requirements and support current employees steadily to promote KS.
First, given the positive and direct relationship between PSM and KS, this study recommends that managers focus on employees' PSM levels as the crucial factor for increasing KS. Government agencies or managers need to hire employees who have high levels of PSM. Specifically, they can assess the PSM levels of job applicants through in-depth interviews or tests that use the scale developed in PSM studies. In addition, managers also need to conduct strategic and tactical actions such as appropriate training for enhancing public employees' PSM levels.
Second, based on the results indicating a relationship between P-G fit and P-J fit and KS, organizations should focus on work group and job compatibility to promote employees' knowledge sharing intentions. To do so, organizations need to assess the degree to which job candidates are compatible with values or norms of a particular group and with the job demands. Moreover, since KS can be enhanced by improving employees' perception of compatibility with their group and job tasks, managers should strive to increase employees' satisfaction with their work groups and to meet other job-related needs.
Third, in addition to appropriate training and education for promoting employees' PSM levels, managers should provide them with additional opportunities, such as informal events and workshops, to improve their identification with their organization. Employees with high levels of PSM believe their values are more likely to be compatible with the public organization for achieving organizational outcomes [26,50]. In line with those findings, managers need to support public employees' identification with their organizations by creating work environments conducive to compatibility so that employees are more likely to engage in KS.

Limitations and Future Research
While the empirical nature of this study has strong theoretical and practical implications, it also has several limitations to be considered. First, since the data were collected by a self-report measure, this study is not free from the common method bias (CMB). Since CMB is associated with single-source and survey-based research, it is a major issue in organizational behavior research [75]. Future research should use a different data source for variables. Specifically, the dependent variable of this study, KS, can be measured by surveying managers or supervisors, while other variables can be reported by employees themselves.
Second, while the data were collected from local government agencies across wide areas, many respondents were in junior roles and not in managerial positions. Although this study controlled the influence of service years in the organization for the primary analyses, it might be impossible to control all respondents' characteristics relating to rank and service years. Therefore, the respondents might not represent the propensity and work behaviors of senior-level local government employees.
Third, this is a cross-sectional study, and data were collected at a given point in time to examine the relationships between main variables. Although cross-sectional research is normally connected and matched with survey research [76], it has an inherent limitation in that the results of data analyses might be changed with time, thus becoming uncertain [77]. Considering the necessity of collecting data at multiple time points, a longitudinal study would be an alternative way to verify whether the relationships between variables in this study are maintained over time.
Fourth, this study controlled a relatively small number of variables relating to work environment, such as job autonomy and organizational learning culture. Despite the close relationship between P-S fit and LMX [47], the author did not consider it in the proposed research model. In addition to the main variables of this study, future research should investigate how other variables (e.g., voluntary learning behavior and fairness perception) might influence KS based on a more in-depth literature review.
Lastly, this study considered PSM as an intrinsic motivational factor that can improve knowledge-sharing behavior in organizations. Although public employees have higher levels of PSM than private-sector workers [26] and are more intrinsically motivated and committed to serving the public [26,29], extrinsic motivational factors still do exist in the public sector workplace. Considering that extrinsic motivation significantly influences knowledge-sharing behavior [15,78], future research should include extrinsic motivational factors to understand knowledge-sharing motivation in the public sector.  Institutional Review Board Statement: Ethical review and approval were waived for this study by the Institutional Review Board at Virginia Commonwealth University in accordance with the national legislation and the institutional requirements.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data are not publicly available due to restrictions of privacy, though the data may be made available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.