How does transformational leadership influence on the performance of public service organizations in a developing country? The interventional roles of NPM cultural orientations

Abstract By drawing on New Public Management (NPM) framework and transformational leadership theory, this study examined how transformational leadership impacts the performance of public organizations via NPM cultural orientations. Research data is collected from 205 public service organizations in Vietnam. The results of analysis with the support of Partial Least Square-Structural Equation Modeling (PLSSEM) technique. The research results indicate (1) Transformational leadership has a significant positive impact on the performance of public organizations; (2) Results-oriented culture mediates the relationship between transformational leadership and organizational performance; (3) Innovation-oriented culture doesn’t mediate the relationship between transformational leadership and organizational performance. This is one of the few studies examine the interrelationship between transformational leadership, NPM culture orientations on public organization performance. The message of the findings to practitioner is that public managers should apply transformational skills at fostering results-oriented culture to lead higher organizational performance. Thus, the findings contribute to a more nuanced understanding of how transformational leadership and organizational culture play a vital role in performance management practices.


Introduction
In the public sector, public service organizations account for the largest number, which are public organizations that provide services essential to the socio-economic life of the country, the community or ensure national security (Tran et al., 2021;Võ & Löfgren, 2019). For example, public service organizations provide services in education, health, culture, physical training and sports, social security, etc. Public service organizations have been under great pressure from structural, governance and financial reform policies according to the theory of new public management (NPM) in OECD countries and beyond (Tetteh et al., 2021;Wynen et al., 2016). To adapt to this change of environment, these organizations have been interested in reform elements such as autonomy, innovation and performance management (Van Der Wal & Demircioglu, 2020). Accordingly, leaders need to change leadership mindset to promote staff to perform better and enhance organizational performance (Dartey-Baah et al., 2021). Leadership is a popular concept in management literature, which is the process of influencing members in an organization to achieve common goals (Yukl et al., 2019). Leadership theory in the public sector is also identified to be quite diverse, including: traditional leadership, collaborative leadership, transactional leadership, transformational leadership, and ethical leadership (Van Wart, 2013). Among that transformational leadership is considered the most popular in management research for the past three decades because of its importance to organizational productivity (Campbell, 2018). Still, many scholars have supported evidence of a positive influence of transformational leadership on individual, group, and organizational performance (García-Morales et al., 2012;Nam & Park, 2019;Wang et al., 2011).
Scholars also agree that transformational leadership creates organizational change by changing organizational culture (Agolla & Van Lill, 2016;Parry & Proctor-Thomson, 2002;Sun & Henderson, 2017). As Bass and Avolio (1994) have noted, transformational leaders inform new visions, values, assumptions and norms, therefore, they could reform organizational culture. In addition, Al-Husseini and Elbeltagi (2016), Sun and Henderson (2017) indicated that transformational leadership plays a vital role on innovation because this style help enhancing subordinates's goal-oriented behavior and positive motivation, promoting organizational change. These arguments reinforce that transformational leadership is a significant predictor of various performance through organizational culture. In addition, Sun and Henderson (2017) found the mediating role of collaborative culture in the relationship between transformational leadership and school performance. Xenikou (2017) provided evidence that innovation and goal culture orientations play as mediators for leadership and organizational identification. By previous studies, we realized that transformational leadership seems to have an indirect influence on organizational performance through organizational culture; however, the studies interested in the mediating role of results-oriented culture and innovation-oriented culture is still limited.
Organizational culture impacts performance by integrating values, beliefs, and norms to shape the method which employees interact and engage with each other (Jacobs et al., 2013). Organizational culture is received much attention of public managers because it plays a core role in changing in organizational activities to achieve goals, especially NPM cultural orientations (Nitzl et al., 2019), such as a result-oriented culture (Verbeeten & Speklé, 2015) or innovationoriented culture (Bendak et al., 2020). First of all, a results-oriented culture from the point of view of NPM theory stresses promoting accountability of managers, emphasizing commitment to organizational goals of employees, and recognize personal achievements (Verbeeten & Speklé, 2015). Transformational leadership style creates a workplace in which employees always try their best to go beyond their self-interest by connecting an inspiring mission to core value, always feel recognition and encouragement, accept innovation, thereby improving individual and organizational performance (Bass, 1999). We therefore have reason to believe that adopting transformational leadership in public organizations will have the potential to enhance a results-oriented culture, and ultimately, better organizational performance. However, empirical evidence for the mediating role of results-oriented culture is still rarely.
Moreover, in the NPM reforms context, along with the emergence of the Covid epidemic, leading to large budget deficits of countries, these have affected the ability of public organizations to provide public services (Mitchell et al., 2021). To deal with these challenges, some authors argue that innovation is important because of its benefit for performance (Clausen et al., 2020). Since then, many scholars have more focused on public sector innovation (S. Kim & Yoon, 2015;De Vries et al., 2018). Previous studies have highlighted that transformational leadership has significant effect on organizational innovation because this leadership style supports culture for innovation, in which employees are encouraged freely discussing and trying out creative approach and ideas (Al-Husseini et al., 2021;Lei et al., 2020). When leaders have transformational characteristics, they will persuade their follower engage with their task activities and also attend to decision-making, these could enhance innovation among follower as well as among organization (Owusu-Agyeman, 2021). At the same view, Elmasry and Bakri (2019) stated that transformational leaders always pay attention to creating a working environment that encourages employees' creativity, promotes flexibility to change and willing to provide resources for organizational innovation. From the characteristics of innovation culture, Gupta and Gupta (2019) suggested that innovation culture has a positive influence on the organization performance. However, Moussa et al. (2018) found that the influencing of leadership behavior on increasing public organization's performance through innovation culture remains ambiguous. Therefore, the interrelationship between leadership style, innovation-oriented culture and performance needs to be further studied.
This study aims at closed the mentioned gap by investigating the relationships among transformational leadership, NPM cultural orientations (results-oriented culture and innovation-oriented culture) and organizational performance in public service organizations in Vietnam. Two research questions guiding the study are (1) Does transformational leadership have a direct impact on organizational performance? and (2) Do results-oriented culture and innovation-oriented culture mediate the relationship between transformational leadership and organizational performance? The findings will add to the academic literature by clarifying how leadership behavior influences organizational performance by emphasizing NPM culture orientations.
Vietnam is still trying to reform financial and accounting policies in the public sector to meet the new requirements of the market economy. For example, Decree no. 60/2021/ND-CP was issued by the Government which demonstrated regulations in increasing the autonomy of public service organizations in using assets, financial and human resources according to the market mechanism (Government, 2021). Additionally, the Ministry of Finance in Vietnam has promulgated five public accounting standards on the basis of approaching international public accounting standards (Ministry of Finance, 2021). In Vietnam, public service organizations are a key part in providing public services and implementing social security policies (Central Executive Committee, 2017). The implementation of the mechanism of autonomy and self-responsibility by public service organizations has achieved initial results, however, these organizations still have many weaknesses in terms of internal governance, and the quality and efficiency of public services are still low (Central Executive Committee, 2017). One of reason for the above limitations is the application of an inappropriate leadership style (Pham, 2018). Specifically, many leaders of public organizations are not fully aware of the reform process, lack determination, backward thinking, have not comprehensively innovated and actively applied the autonomy mechanism (Central Executive Committee, 2017). Like other developing countries (P. S. Kim, 2009), public sector organizations in Vietnam still face with limited institutional capacity, this lead to a weaknesses in regulatory practice, administrative inefficiencies, and low level of transparency; thus, the volume and quality of public service are inadequate (Võ & Löfgren, 2019). Therefore, the recommendations from this study will be a useful source for public sector governance implications in Vietnam and other developing countries. At the same time, it contributes to enriching the academic literature on the effects of transformational leadership on organizational culture and performance in the public sector context.

Transformational leadership
Transformational leadership concept was introduced by Downton (1973), and later extended by Burns (1978), which is described as a process in which both leaders and followers discuss and share to create higher motivation (Bass & Avolio, 1994). Transformational leaders are those who want to develop their organization's full potential, create good value systems, and motivate their followers (Jensen & Bro, 2018). The structure of transformational leadership has four components including idealized influence, intellectual stimulation, motivational motivation and individualized consideration (Bass & Avolio, 1994). Idealized influence refers to a leader's ability to attract employees towards the achievement of organizational goals (Antonakis et al., 2003). Intellectual stimulation includes behaviors that encourage employees to develop creative thinking and innovation in problem solving (Bass et al., 2003). Individualized consideration regarding a leader's behavior in creating a supportive environment for employee growth (Bass & Avolio, 1993). Finally, inspirational motivational demonstrates a leader's ability to inspire employees through symbolic behaviors (Yammarino & Dubinsky, 1994). Transformational leadership style has been receiving special attention from leaders in the public sector (Orazi et al., 2013), especially in emerging economies that are implementing public governance reforms like Vietnam (Tran et al., 2021).

NPM cultural orientations
NPM theory has its roots in a combination of the new institutional economics movement and a set of managerialism-style business waves in the public sector (Hood, 1991). Most of scholars found NPM theory to be one of the most powerful reform doctrines to improve performance of public organizations (Nitzl et al., 2019). Basically, NPM theory is intended to create an environment that stimulates employees to increase flexibility, risk taking, creativity, and strong commitment to the goal of improving organizational performance, similar to the private sector (Hood, 1995;Wynen & Verhoest, 2013). In addition, the goal of NPM is towards getting things done better, in one way or another, the trend towards privatization of public services, increasing the efficiency of public sector activities (Lapsley, 2009;Wynen et al., 2016). The main idea of NPM is that organization should change culture, more detail, the traditional culture of adherence to rules and procedures needs to be replaced by a culture of results and innovation; then, every employee will focus on improving performance (Verbeeten & Speklé, 2015). Obviously, the principle of NPM is even more suitable for public service organizations context, with the strengthening of self-management mechanism and restructuring to adapt to environmental instability (Camilleri, 2021;. NPM cultural orientations are types of organizational culture that reflect organizational change trends as a result of NPM movements and related activities, emphasizing on flexibility, innovation and towards improving organizational performance (Nitzl et al., 2019). NPM cultural orientations are often referred to in public administration studies as results-oriented culture (Verbeeten & Speklé, 2015) and innovation-oriented culture (Wynen et al., 2014). In organizations with a results-oriented culture, the organization's activities stress on action and results, and high expectations for performance (Sheridan, 1992). Specifically, a results-oriented culture focuses on a manager's level of strong commitment to achieving goals, willingness to take responsibility, and recognition of employees' contributions (Nitzl et al., 2019;Verbeeten & Speklé, 2015). Therefore, results-oriented culture is a culture that is encouraged to be promoted in the public sector according to NPM theory in order to improve organizational performance (Yen et al., 2020). In addition, innovation-oriented culture is considered an important element to help public organizations adapt to environment with the rapid development of science and technology, economic pressures, and political changes (Wynen et al., 2016). Organizations have an innovation-oriented culture, which means change and creativity are encouraging, including taking risks into new areas or areas where members have little or no prior experience (Chia & Koh, 2007). A high organizational innovation environment will include a high degree of involvement of members in developing and promoting innovation (Bendak et al., 2020).

Organizational performance
Due to the complexity of the operating mechanism of the public sector, most of the literature suggests that organizational performance in the public sector is a multidimensional concept (Andersen et al., 2016;Andrews et al., 2011;Pollanen et al., 2017). The basic feature of public organizations is that there are many stakeholders with diversity goals to be achieved but noted that goals often are ambiguous (Andersen et al., 2016). These characteristics lead to variety of expectations for organizational performance (Andersen et al., 2016); hence, organizational performance is perceived to be quite diverse. Many authors initially consider organizational performance only quantitatively, but this view is often inadequate for the public sector, therefore the performance of public sector should be considered both quantitatively and qualitatively aspect (Van de Ven & Ferry, 1980;Verbeeten, 2008). Quantitative results are budget efficiency, quantity of products and services produced, revenue, profit, while qualitative results are service quality, innovation, customer satisfaction, long-term effectiveness (Song & Meier, 2018). Recently, George et al. (2019) argued that organizational performance includes aspects of performance, financial results, social outcomes, and the ability to meet public needs. Although the authors have different interpretations, the approach to the concept of public organizational performance has one thing in common, it is necessary to comprehensively consider both quantitative and qualitative aspects.

Transformational leadership and organizational performance
Leaders have always been seen as a strong factor in promoting organizational performance, so leadership research has always dominated the public administration literature (Crosby & Bryson, 2018;Bradley E. Wright & Pandey, 2009). First of all, leaders directly give strategies and guide all activities taking place in the organization (Ahmed et al., 2018). Next, leadership also indirectly affects activities to achieve organizational goals by encouraging and supporting employees to come up with ideas to improve work (N. T. Nguyen et al., 2021). This means that all the personal characteristics, behaviors, and leadership styles of a leader will have effect on organizational performance. The effectiveness leadership is the main factor determining the success or failure of an organization. Indeed, to adapt internal volatility and growing external uncertainty, organization should focus on training leadership skills for managers so that they are skilled enough to deal with problems arise (Hennessey, 1998). The general principle of the influence of leadership on performance depends on the leader's ability to diagnose and understand the situations in the organization and more importantly, the application of the right style in any situations (Ogbonna & Harris, 2000). Many empirical studies have shown that leadership style is a key factor for organizational performance (Trottier et al., 2008;Van Wart, 2013), in which the influence of transformational leadership style on performance is the most discussed (Elmasry & Bakri, 2019;Katou, 2015) because of its prominent influence on mainstream organizational theory (Moynihan et al., 2013). Transformational leadership theory facilitates change in mission, vision, values and culture (Hsieh & Liou, 2018), with its inherent characteristics transformational leaders is seen as an importance predictor of fundamental issues in organizations such as: motivation, innovation and performance enhancement (Alrowwad et al., 2020;Andersen et al., 2018;Campbell, 2018;Tran et al., 2021).
For public service organizations, competitive pressure is becoming stronger following the trend of socializing public services (Võ & Löfgren, 2019). Transformational leaders can control the impact of environmental uncertainty by communicating a continuous improvement mindset to subordinates and help reduce conflicts within the organization (D. I. Jung et al., 2003), thereby improving the ability of the organization to achieve goals (Elmasry & Bakri, 2019). Transformational leaders also improve efficiency and quality of public services by supporting employees to implement new solutions to complex problems (Sarros et al., 2008). Paarlberg and Lavigna (2010) further emphasized that transformational leadership is a value-based leadership strategy with particular relevance to the public sector, which improves organizational performance. In addition, the characteristic of a public service organization is to serve the community, so the condition for good results is that employees need to prioritize the common good instead of focusing on personal interests. And transformational leadership drives this mechanism (Tran et al., 2021), transformational leaders communicate to employees a clear understanding of what need to contribute to realize the organization's mission (Andersen et al., 2018); repeatedly emphasizes why employees' work contributes to organizational success . This is the basic reasoning why transformational style strengthen employees' awareness of the importance of the task as well as motivating employees to work beyond self-interest, towards public-interest. According to these arguments, we offer the following hypothesis: Hypothesis 1 (H 1 ): Transformational leadership has a positive impact on organizational performance.

The mediating role of result-oriented culture
Transformational leadership theory indicates the emotions, values, and importance of leadership to encourage employees to put more effort into achieving the overall goals of the organization (Bass & Avolio, 1994). To inspire employees, transformational leaders also strive to be a role model for employees in organization (Paarlberg & Lavigna, 2010). First, transformational leaders make a strong commitment to achieving the set goals. Next, transformational leaders are willing to take responsibility for their results and face with their own failures. On the other hand, transformational leaders are also concerned with the needs of subordinates in terms of recognizing employee achievements and facilitating employee advancement. Indeed, the characteristics of transformational leadership are well suited to forming a results-oriented culture (Bass & Avolio, 1994). Many previous studies have paid attention to the relationship between leadership and organizational culture (Li et al., 2018;Tipu et al., 2012). According to Bass and Avolio (1994), transformational leadership has a strong influence on the formation of organizational culture. More specifically, transformational leadership promotes organizational change by forming a strategic vision for the organization and motivating all employees to work towards a common goal (Sun & Henderson, 2017). While a results-oriented culture also emphasizes the importance of employees striving to accomplish organizational goals (Verbeeten & Speklé, 2015). From these arguments, we propose the following hypothesis:

Hypothesis 2 (H 2a ): Transformational leadership has a positive impact on results-oriented culture
NPM public sector reform requires public institutions to strengthen performance-based governance (Wynen & Verhoest, 2013). To do this, public organizations need to first replace the traditional bureaucratic culture with a results-oriented culture, reducing dependence on procedures and regulations (Tallaki & Bracci, 2019). When an organization operates with a results-oriented culture, all employees will have a strong commitment to achieving goals and managers are willing to take responsibility for the results achieved (Verbeeten & Speklé, 2015). Therefore, a results-oriented culture helps an organization to focus more on outputs with the expectation of improving performance (Nitzl et al., 2019). Moreover, the positive impact of a results-oriented culture on the performance of public organizations has been confirmed in a number of previous studies, for example, Nitzl et al. (2019); Verbeeten and Speklé (2015); Yen et al. (2020). Thus, both theory and empirical research support the view that the clearer the results-oriented culture, the better the public organization's performance. Given related frameworks, we suggest the following hypothesis: Hypothesis 2 (H 2b ): Results-oriented culture has a positive impact on organizational performance.
Transformational leadership theory has emphasized the importance of transformational leadership for improving performance directly and indirectly through other mechanisms (Katou, 2015;Moynihan et al., 2013;Paarlberg & Lavigna, 2010). In which, organizational culture can be seen as an important mediator for the impact of transformational leadership on performance (Bass & Avolio, 1994). Transformational leaders are good inspirers and always encourage their followers to think towards a common goal, which is to build a results-oriented culture (Çelik, 2018;Den Hartog et al., 1996). Transformational leaders are intellectually motivated, sophisticated, inspire their follower to pursue a better future (Alrowwad et al., 2020). Public organizations have a working environment in which employees accept change, strive for common goals and are willing to take responsibility for results, leading to high performance (Verbeeten & Speklé, 2015;Yen et al., 2020). Therefore, given the arguments for the direct impact of transformational leadership on resultsoriented culture (H 2a ) and the direct impact of results-oriented culture on organizational performance (H 2b ), we predict the following hypothesis: Hypothesis 2 (H 2 ): Results-oriented culture mediates the relationship between transformational leadership and organizational performance.

The mediating role of innovation-oriented culture
Currently, public service organizations around the world have been carrying out innovative activities in order to adapt to competitive pressures and the development of technology (Gieske et al., 2020;Moussa et al., 2018). To better support innovation, leaders in public organizations need to adopt a transformational leadership style (Gieske et al., 2020). It is understood that leaders need to inspire, encourage and drive innovation by creating the right organizational culture that enables employees to work effectively (Nusair et al., 2012). Transformational leadership ensures the longterm survival of the organization by supporting and driving innovation (García-Morales et al., 2012).
Innovation is seen as one of the key factors influencing the sustained success of any organization and it is influenced by organizational culture (Bendak et al., 2020). This matter only happen if leaders create a work environment that favors innovation (Borins, 2002). Therefore, instead of addressing innovation behavior directly, our study is interested in the basic premise that leads to this behavior, which is innovation-oriented culture (Wynen et al., 2014). The role of leadership in shaping an organization's innovation-oriented culture has been explored in previous studies (Nusair et al., 2012;Szczepańska-Woszczyna, 2015;De Vries et al., 2018). In particular, leaders who adopt a transformational style are able to orient their organizations to focus on innovation by motivating their employees to be creative in problem solving to achieve goal (E.-J. E.-J. Kim & Park, 2020;Bradley E Wright et al., 2012). Furthermore, transformational leaders clearly communicate their vision and motivation, which creates a workplace that fosters innovation by broadening feedback channels and enabling employees to propose new ideas (Al-Husseini & Elbeltagi, 2016;Jaskyte, 2004). In addition, by surveying employees working in local government organizations in Korea, S. S. Kim and Yoon (2015) provided empirical evidence on the positive influence of transformational leadership styles on innovation-oriented culture. Extending the scope of the study, by collecting data in both public and private institutions, Xenikou (2017) confirmed that this relationship exists. In sum, the following hypothesis is proposed and tested in our study:

Hypothesis 3 (H 3a ): Transformational leadership has a positive impact on innovation-oriented culture.
Responding quickly and creatively to environmental uncertainty is an important capability that any organization must possess to avoid the risk of exclusion (Bendak et al., 2020). The operating environment of the public sector is increasingly changing with the development of society, political changes and economic pressure (Wynen et al., 2016). To use resources more efficiently, public institutions need to continuously innovate to adapt to circumstances and better serve citizens (Clausen et al., 2020), which implies that innovation is an important precursor to the performance of public organizations (Park et al., 2016). When an organization encourages employee innovation, allows employees to experiment with new ways of doing things, takes risks, and supports personal growth, innovation will be maintained, and this tends to more effective and productive (Park et al., 2016). Many authors also agree that employees are willing to take risks, be more innovative and proactive in problem solving, which will contribute to improved organizational performance (Kim, 2010b;Langer & LeRoux, 2017). Based on strong relationship between innovation-oriented culture and performance from theoretical and empirical evidence, the hypothesis is suggested as follow: Hypothesis 3 (H 3b ): Innovation-oriented culture has a positive impact on organizational performance.
When employees work under transformational leaders, they are encouraged to learn, improve their knowledge and embrace innovation (García-Morales et al., 2012). In addition, employees will have goals that are consistent with the leader's vision, have a positive attitude from the leader's charisma, and feel energized to better complete their tasks (Xenikou, 2017). Transformational leaders, through their influence on the organization's environment, systems, and strategies, impact organizational performance (D. I. Jung et al., 2003). For the public sector, leaders should aim to develop an innovative organizational culture, which is a necessary condition for effectively implementing government reform initiatives (Hartmann & Khademian, 2010), thereby, improving the performance of the organization (Tabassi et al., 2017). As argued by Moynihan et al. (2012), transformational leaders foster a culture that helps an organization adapt to its environment in a timely manner, acquiring the resources essential to the organization's growth. An operating environment that encourages innovation is an important foundation for innovative activities (Borins, 2002;Moussa et al., 2018), and innovative activities will enhance performance (Naranjo-Gil, 2009;Van Der Wal & Demircioglu, 2020). Therefore, we propose the following hypothesis:

Hypothesis 3 (H 3 ): Innovation-oriented culture mediates the relationship between transformational leadership and organizational performance.
The proposed model and corresponding hypotheses are shown in Figure 1.

Data and sample
Public service organizations in Vietnam are organizations which established in accordance with the law to provide various public services, such as education, health care, culture, sports, etc. (Government, 2021), these are in the process of renewal to improve quality and operational efficiency. For example, these organizations have rearranged their apparatus to be more streamlined, improve governance, enhance autonomy, and maintain their leading role in the public service market (Central Executive Committee, 2017). We use survey questionnaires to collect data. Respondents are senior managers, middle managers and chief accountants working in public service organizations. In each organization, we asked a representative to answer the questionnaire. In order to increase the reliability of the results, we selected respondents with at least five years of work experience. In this study, convenience sample is chosen, which is considered appropriate in the context of the public sector in Vietnam (Yen et al., 2021).
To ensure that the questionnaire content was understandable and relevant to the research context, we pre-tested by eight managers of public organizations in Vietnam. The results show that the content of the questionnaire is clear and easy to understand. After that, the questionnaire was completed and sent directly, or via email and social networks, to the survey respondents. The survey period was conducted from August 2020 to December 2020. Out of 483 questionnaires sent out, we received 224 responses. In which, 19 responses are invalid because of lack of information or tend to respond to the same degree for observed variables. Therefore, 205 valid responses were used for data analysis.
Research model in this study is the structural equation modeling (SEM). Sarstedt et al. (2020) suggest that the partial least square method (PLS) is considered suitable for analyzing complex relationships in SEM in social science research. Furthermore, the PLS-SEM method does not require normally distributed data and is more suitable for small sample sizes (Hair et al., 2019). PLS-SEM is a non-parametric method based on conventional least squares regression, and it is designed to maximize the explained variance (Reinartz et al., 2009). PLS-SEM is also a widely accepted statistical technique and applied in many researches in different fields of accounting and public administration (Ghufran Ali Khan et al., 2022;Verbeeten & Speklé, 2015). Therefore, our study using PLS-SEM method with sample size 205 is considered suitable. We analyze descriptive statistics of the survey sample in Table 1.
In Table 1, among the respondents, 84.8% are senior managers (the sum result of Head of organization and Vice-director), the rest are chief accountants. All of respondents have working experience over 5 years. Respondents play an important role in the organization, have sufficient knowledge and experience on issues related to leadership, organizational culture and performance. These characteristics ensure the quality and reliability of the information collected. In term of field, the majority are public organizations providing education and health services with 73.2%, the rest are other fields. According to the degree of financial autonomy, public organizations with the state budget guaranteeing all operating expenses have the highest percentage (50.7%), followed by public organizations with self-guaranteed part of recurrent expenses (32.7%), followed by public organizations with self-guaranteed recurrent expenses (12.2%), and public organizations with self-guaranteed recurrent and investment expenses (4.4%). These characteristics of the survey sample reflect quite similar to the structural characteristics of the overall public service organizations in Vietnam (General Statistics Office, 2018). Therefore, the survey sample can be representative of the population, supporting the generalizability of the research results.

Variable measurement
Constructs are measured by many observed variables, which have been tested and confirmed in previous studies. Transformational leadership (TL) is measured by 5 observed variables, inherited from the scale of House (1998). This is a commonly used scale in public administration studies (e.g., Campbell, 2018;Kroll & Vogel, 2014;Wright et al., 2012). Accordingly, respondents were asked to rate the transformational leadership skills of senior managers in their organizations. The level of agreement of the respondents was ranked from 1 (strongly disagree) to 5 (strongly agree).
Our study considers two important types of organizational culture as recommended by NPM theory (Nitzl et al., 2019;Wynen & Verhoest, 2013). First, results-oriented culture (CUL) is a concept developed under the OECD-NPM doctrine (Verbeeten & Speklé, 2015). This scale addresses aspects of accountability, performance, management's commitment to achieving goals, and recognition of individuals' contributions to the success of the organization. This scale includes four items and to be measuring by 5-point Likert scale. Next, innovation-oriented culture (INNO) is adapted from Wynen et al. (2016), which comprises four items: innovation, risk-taking, willingness to experiment and creativity. Respondents rate the existence of innovation-oriented cultural characteristics in their organizations on a 5-point Likert scale.
In term of organizational performance (PER), we use perceptual measurement instead of archival data. Many previous studies have shown that measuring organizational performance in the public sector using self-assessment reports based on personal is more appropriate than archival data (Giauque et al., 2013;Pollanen et al., 2017). This assessment was developed by Van de Ven and Ferry (1980) and has been widely used in public sector performance management studies (Verbeeten, 2008;Verbeeten & Speklé, 2015;Yen et al., 2021). Therefore, our study also uses this scale to measure organizational performance, including seven observed variables. This scale comprehensively reflects both quantitative and qualitative aspects of performance in public organizations. Specifically, the respondents will rate their personal feelings about For control variables, the analysis includes control variables for organizational size and organizational age (Gomes et al., 2017;Pollanen et al., 2017;Song & Meier, 2018). Based on research by Pollanen et al. (2017), we measured the organizational size (SIZE) by the total number of long-term employees at the organization. Organizational age (AGE) was measured according to Glisson and Martin (1980). This scale is calculated from the time of establishment to the time of the survey and is rounded to the number of years. These control variables are continuous variables. Surveyed public service organizations have the average of 92 full-time employees and the average of 23 years old.

Common Method Bias
The problem of common method variance (CMV) is considered to be common in studies using survey data, it can cause results to be biased (Podsakoff et al., 2012). As recommended by Tehseen et al. (2017), we use both procedural and statistical remedies to assess and control CMV. For procedural remedies, protecting the anonymity of the answer is applied by us to minimize the evaluation apprehension. Next, the improving of scale items is also applied by us to ensure the clarity and understanding of the question (Podsakoff et al., 2003). Example, we try to exclude items ambiguity and provide their examples so that respondents understand the question correctly. In addition, we also recommend that the respondent estimate organizational performance measures according to organizational meetings and documentation (Liang et al., 2007).
After data collection, we apply two statistical remedies. The first test technique is Harmon onefactor test, was conducted by SPSS software (Chang et al., 2010). This test is performed by analyzing the four conceptually important variables in our model including transformational leadership, results-oriented culture, innovation-oriented culture and organizational performance. The analysis results show that there are four distinct factors, accounting for 61.13 percent of the total variance. Especially, the first unrotated factor accounts for 33.78 percent, indicating that CMV is not an issue in this study (Tehseen et al., 2017). The second test technique is using a common method factor in the PLS model (Liang et al., 2007;Podsakoff et al., 2003). The common method factor includes all the main constructs' indicators. The research results show that the average substantively based variance (0.59) is larger than the average method based variance (0.33). And some of the method factor loadings are insignificant. This result indicates that CMV is unlikely to be a serious concern in this study (Williams et al., 2003).

Measurement model testing
Before testing the hypotheses, the measurement model was analyzed to estimate the relationship between the observed variables and the latent construct (Hair et al., 2019). The measurement model in this study has ensured reliability, convergent validity and discriminant validity. The results are presented in Table 2 and Table 3.
In Table 2, the composite reliability of four constructs has values from 0.81 to 0.93, which is greater than the threshold value of 0.708 (Hair et al., 2017). Furthermore, the factor loadings of all items are greater than 0.50, being significant at the 0.01 level, in an acceptable range (Hulland, 1999). These results indicate that the measurements are guaranteed to be reliable. Average variance extracted (AVE) is the criterion to evaluate the convergence value of the scale, and the acceptable threshold is 0.5 (Hair et al., 2019). The analysis results in Table 2 show that the scales have good convergence value with the lowest AVE value of 0.52.
Next, the results in Table 3 show that the discriminant validity of the structure is achieved according to the test of Fornell and Larcker (1981). Specifically, the square root of the AVE of the  key scales (ranging between 0.72 and 0.82) is greater than the corresponding bootstrapped correlations between these scales (ranging between 0.18 and 0.51). Furthermore, Henseler et al. (2015) suggest that the degree of difference between structures should be evaluated by the Heterotrait-Montrait (HTMT) test, which is superior to the Fornell Larcker criterion. The highest HTMT value in Table 3 is 0.64, which was significant below 0.85 (Henseler et al., 2015). Thus, latent constructs achieve good discriminant validity.

Structural model testing
After the measurement model is satisfactory, we test the structural model to estimate the explanatory level of the path coefficients (Hair et al., 2019). First, we test multicollinearity by using the variance inflation factor (VIF) in our model (O'brien, 2007). The results indicate that the internal VIF values was lower than the threshold of 5 (ranging from 1.00 to 1.48); therefore, multicollinearity is not a serious problem in our study. Next, the estimated results from PLS analysis are shown in Table 4. The R 2 value of organizational performance is 0.30, indicating that the model has a good fit for the survey dataset.
Hypothesis H 1 suggested a significant positive association between transformational leadership and organizational performance. Results in Table 4 shows that a significant path coefficient (β = 0.22, t-value = 3.21) for this direct relationship. It means that leaders in public service organizations adopt a transformational style that will enhance organizational performance. This result provides further evidence for Paarlberg and Lavigna's (2010) statement that transformational leadership has a positive effect on both individual and organizational performance (Katou, 2015;Paarlberg & Lavigna, 2010).
Next, consistent with the prediction in hypothesis H 2a , transformational leadership has a significant positive effect on results-oriented culture. The data analysis in Table 4 presents that this path coefficient is significant (β = 0.51, t-value = 9.41), thus supporting H2a. This implies that leaders with a transformational style pose challenges and inspire subordinates to accomplish their goals of improving performance (Bass & Avolio, 1994;Den Hartog et al., 1996). Similarly, the positive influence of results-oriented culture on organizational performance (H 2b ) is also accepted. The path coefficient of this relationship is 0.41 with one percent significance level (t-value=4.82). The present study reinforces the assumptions of NPM theory (Hood, 1995), the key to improving public sector performance is the formation of a results-oriented culture. Employees who are committed to their duties and want to contribute to the success of the organization will improve organizational outcomes (C. S. Jung & Lee, 2013). Therefore, the emergence of a results-oriented culture in public institutions to lead to higher results is inevitable. Hence, both H 2a and H 2b are proved.
Furthermore, hypothesis H 3a proposed that transformational leadership has a positive effect on innovation-oriented culture of public service organizations in Vietnam. This path coefficient is significant (β = 0.34, t-value = 5.35), supporting the expectation that transformational leaders can create work environments that foster innovation. This result is similar to previous studies such as S. S. Kim and Yoon (2015); Xenikou (2017). Contrary to a predicted positive association between innovation-oriented culture and organizational performance in H 3b, an insignificant positive path coefficient (β = 0.05, t-value = 0.75) was found. This means that the innovation-oriented culture in public service organizations in Vietnam is still weak, so it has not contributed to improving organizational performance. This result is not the same as the work by Y. , conducted at U.S state agencies. However, this result also supports the assertion of Hood and Dixon (2015) that the effects of NPM-style reforms of public services are very complicated. Thus, H 3a is corroborated but H 3b is not.
Following these tests of direct effects, we test the indirect effects of transformational leadership on organizational performance through results-oriented culture (H 2 ) and innovation-oriented culture (H 3 ). To examine the mediating effects of results-oriented culture and innovation-oriented culture, following Hair et al. (2017), we use bootstrap analysis with with 3,000 subsamples. In addition, we calculated confidence intervals for each specific indirect effect in our research (Fritz & MacKinnon, 2007). The results demonstrated that results-oriented culture partially mediated the relationship between transformational leadership and organizational performance (β = 0.21, t-value = 4.14, CI = [0.12; 0.31]), supporting H 2 . This confirms that the transformational leader promotes the creation of a results-oriented culture, which in turn enhances organizational performance. This evidence also further supports transformational leadership theory. Specifically, transformational leaders inspire employees to focus on organizational performance goals, thereby boosting employees' intrinsic motivation, ignoring self-interest, toward organizational performance. Furthermore, this result shows that transformational leadership contributes to promoting results-oriented culture, which is a type of culture that should be promoted in the public sector according to NPM theory (Hood, 1991).
Contrary to a predicted mediating role of innovation-oriented culture in H 3 , the indirect effect of transformational leadership and organizational performance through innovation-oriented culture is not statistically significant (β = 0.02, t-value = 0.71, CI = [-0.03; 0.07]). The reason for this is leaders of public service organizations in Vietnam have adopted a transformational leadership style that has created an innovation-oriented culture, the extent of its influence on organizational performance is unclear. This result is not line with the study by Parry and Proctor-Thomson (2002), transformational leadership impacts on organizational performance through its influence on climate for innovation. However, consistent with Hood and Dixon (2015), this result shows that the effectiveness of NPM public finance reforms should be observed over a longer period of time.
In addition, following the instructions of J. J. F. Hair et al. (2021), we use Cohen's effect size (f 2 ) and PLS-Predict procedure to assess the predictive power of the research model (in Table 5). The f 2 values of direct effect from transformational leadership to results-oriented culture is 0.35, indicating that the effect sizes of this direct effect are very large. The f 2 values of other direct effects (except innovation-oriented culture -> organizational performance) ranged from 0.05 to 0.17, indicating that the effect sizes of these direct effects were weak and medium.
Finally, we evaluate predictive power through Q 2 _predict values and comparing the root mean square error (RMSE) values resulting from PLS-Predict procedure (J. J. F. Hair et al., 2021). The  Table 5 show that all Q 2 _predict values are greater than 0. Furthermore, the PLS-SEM analysis (PLS-RMSE) produces smaller prediction errors than the linear regression model -LM (LM-RMSE) for all items measuring results-oriented culture, innovation-oriented culture and performance. Therefore, the research model in our study had high predictive power.

Theoretical implications
The present study enriches the literature on the relationship of leadership, culture and performance in the public sector. Many previous studies have been interested in organizational culture as a mediating mechanism explaining the indirect influence of leadership on organizational performance (Parry & Proctor-Thomson, 2002;Sun & Henderson, 2017), however, NPM cultural orientations (eg. results-oriented culture and innovation-oriented culture) seem to have received little attention. For example, Parry and Proctor-Thomson (2002) stated that the importance of a culture that encourages innovative thinking; while Sun and Henderson (2017) focused on a collaborative culture, it encourages employees to work together to solve problems that arise at work for the goal of the organization. Therefore, the important contribution of this study is to provide more evidence confirming the partial mediating role of results-oriented culture in the relationship between transformational leadership and performance of public service organizations. Another finding, this research also shows that the indirect impact of transformational leadership and organizational performance through innovation-oriented culture is non-existent in Vietnam public service organizations. In sum, the findings of this study provide empirical evidence on the indirect impact of transformational leadership on organizational performance through different types of organizational culture is different. This also means that the expectation of a positive impact of the implementation of NPM public finance reforms (Hood, 1995) on organizational performance should be observed over a longer period of time and in different public organizational contexts (Hood & Dixon, 2015;Tallaki & Bracci, 2019).
This study also contributes to public administration research by providing empirical evidence on the direct positive impact of transformational leadership on both results-oriented culture and innovation-oriented culture in public service organizations. These results confirm the view of Bass and Avolio (1994) that the development of organizational culture is largely influenced by leaders, specifically transformational leadership style. Transformational leadership not only foster to results-oriented culture but also promote innovation-oriented culture (S. S. Kim & Yoon, 2015;Li et al., 2018;Owusu-Agyeman, 2021). Transformational behaviors of leader play core role to helping public service organizations increase innovation through creating an innovative working environment (Al-Husseini et al., 2021;Li et al., 2018), sharing knowledge and promoting employee creativity (Lei et al., 2020;Tran et al., 2021).
Next, our study investigated the impact of organizational culture types on organizational performance in public sector. For results-oriented culture, the findings are consistent with study of Verbeeten and Speklé (2015); Yen et al. (2020), which means results-oriented culture has a positive effect on organizational performance. This result gives a donation to the argument of NPM theory that public organizations should promote increased accountability and a strong commitment to organizational goals in order to improve organizational performance (Verbeeten & Speklé, 2015). Our research is even more meaningful when carried out in a developing country because P. S. P. S. Kim (2009) found that one of the major challenges in public administration in developing countries is the lack of results orientation. Public institutions in developing countries seem to be concerned only with short-term achievements, with little attention to the long-term goals of the program (Tallaki & Bracci, 2019). This lead to short-term goals being achieved but can negatively affect long-term goals (P. S. P. S. Kim, 2009). Therefore, strengthening a results-oriented culture is a necessary condition for the long-term performance goals of the public sector.
Interestingly, there is no evidence of the positively influence of innovation-oriented on organizational performance in our study while many previous studies concluded that the characteristics of the innovation-oriented culture in order to adapt to the change of the operating environment; thus, it has positive impact on the performance (Jacobs et al., 2013;Park et al., 2016). In addition, this study did not find the influence of innovation-oriented culture on the performance of public service organizations, which can be explained as follows: Firstly, Vietnamese's public sector still faces the same problems as other transition economies, including poor governance, high corruption, and cumbersome bureaucracy (Pham, 2018). The performance of public service organizations in Vietnam is still weak due to the slow innovation process, the lack of transparency, low autonomy and inadequate inspection and supervision (Central Executive Committee, 2017). Furthermore, the effectiveness of an innovation-oriented culture requires a certain time lag and must be evaluated over a long period of time (Xenikou, 2017). Our study collected data at one point in time, so it may not be clear how an innovation-oriented culture affects the performance of public service organizations. In summary, the results of this study help to emphasize the point of Hood and Dixon (2015) that the process of public governance reform under the NPM theory always has different results in different public sector contexts. Specifically, in a developing country like Vietnam, public governance reforms under NPM theory are having a positive effect on performance. In some cases, in less developed countries or other developing countries, NPM principles introduce unwanted organizational behavior or inefficiencies (Andrews, 2012;Rahaman & Lawrence, 2001), with very little capacity for organizational innovation (Tallaki & Bracci, 2019).
Finally, our research supports the statement of Moynihan et al. (2013) that the goal of adopting a transformational leadership style is to increase organizational performance. Therefore, our study adds the empirical evidence that transformational leadership increase public organizational performance which has been explored by previous studies (Alrowwad et al., 2020;Bellé, 2014;Paarlberg & Lavigna, 2010). In summary, the findings of this study are in accordance with the argument of transformational leadership theory that transformational leaders inspire employees to achieve higher-than-expected performance (Al-Husseini et al., 2021;Bass, 1999;Owusu-Agyeman, 2021).

Managerial implications
Besides theoretical contributions, our research provides some governance implications for public service organizations. With the direct and indirect impact of transformational leadership on organizational performance, this study has confirmed the important role of applying transformational leadership style, especially in public service organizations in transition economies (Walumbwa & Lawler, 2003). The application of transformational leadership style is still limited in public organizations in developing countries (Dartey-Baah et al., 2021;T. T. Nguyen, 2020). Moreover, the provision of public services is increasingly socialized, so the competition in service quality is increasing for public service organizations (Gieske et al., 2020); hence, the findings also imply that leaders in public service organizations need to apply the principles of transformational leadership to model behaviors in the organization, towards forming a new culture for the organization. This will motivate employees to embrace innovation and drive results for the overall development of the public organization. First of all, Government needs to promote human resources management strategies to enhance transformational leadership of leaders in public organizations. Next, public organizations should provide relevant need to strengthen transformational leadership training program for managers and prioritize recruiting leaders who have a transformational characteristic.
Based on the mediating effect of results-oriented culture, we suggest that leaders should be concerned with creating and spreading a results-oriented culture in their organization. Accordingly, leaders need to demonstrate a strong commitment to achieving their goals, taking responsibility for the results achieved; and recognize employee contributions. Another practical recommendation, leaders should take responsibility when they are not achieved their goals because this behavior will motivate for their employees to achieve higher performance (Jin et al., 2016;Lei et al., 2020). In addition, strengthening the results-oriented culture also benefits for public organizations such as increased transparency, financial accountability, and improved organizational performance (Yen et al., 2020). Therefore, in training and development projects for public managers, we recommend that policy-making agencies take courses on transformational leadership theory and building a results-oriented culture.
In addition, given the positive effect of transformational leadership on innovation-oriented culture, another practical recommendation is that leaders should facilitate employees to trial new ideas and recognize their creativity. Leaders need to support resources that enable the climate for innovation. Especially, leaders need to communicate effectively and evaluate employee performance fairly because these practices foster an innovation-oriented culture (S. Al-Husseini et al., 2021;S. Kim & Yoon, 2015). Finally, given the evidence that innovation-oriented culture has no impact on performance, leaders in public service organizations in transition economies as Vietnam need to demonstrate a stronger commitment to innovation. To improve public sector performance, leaders need to demonstrate strategic vision and create favorable conditions for employees to carry out innovative activities. To sum up, transformational leadership style is essential role for public service organizations to adopt to promote employee creativity, foster organizational innovation, and improve performance (Gieske et al., 2020;Owusu-Agyeman, 2021).

Limitations and further research
Besides the theoretical and practical implications as presented, the study has some limitations. First of all, this study used cross-sectional data with a one-time survey, we did not control for time bias when measuring variables. More specifically, NPM cultural orientations (e.g. resultsoriented culture and innovation-oriented culture) are accumulated through a certain period, and its influence on organizational performance should be observed in the long time. This is not expected to help explain the cause-and-effect relationships between variables (Wiley, 2011). Therefore, future researchers should collect data of variables at different points in time, for example, about six months apart. Furthermore, our study used PLS-SEM technique, it seems to be appropriate for the topic. However, this is a frequentist method which faces several drawbacks. For example, significant and non-significant p-values can be interpreted meaningfully in the frequentist framework. Thus, future researchers should use a Bayesian or other non-frequentist method instead.
Next, organizational culture is a broad concept (Bendak et al., 2020) but we only consider NPM cultural orientations (Nitzl et al., 2019), specifically, results-oriented culture (Verbeeten & Speklé, 2015) and innovation-oriented culture (Wynen et al., 2014). According to previous studies, the indirect impact of transformational leadership on organizational achievements is also influenced by other cultures such as goal culture (Xenikou, 2017), cooperation culture (Sun & Henderson, 2017), development culture (Langer & LeRoux, 2017), etc. Therefore, future studies can further explore other aspects of organizational culture to provide a more comprehensive picture of the mediating role of organizational culture in the relationship between transformational leadership and organizational performance in public sector. The impact of transformational leadership on organizational performance may be mediated by variety of factors such as: public service motivation (Paarlberg & Lavigna, 2010;Schwarz et al., 2020), performance information use (Kroll & Vogel, 2014;Moynihan et al., 2012), or organizational learning (Al-Husseini et al., 2021;García-Morales et al., 2012). These are possible directions for future research.
Finally, the transformational leadership conceptual scale in this study is considered as a unidimensional scale. As originally proposed by Bass and Avolio (1994), transformational leadership is a multidimensional scale consisting of four behavioral dimensions (e.g. inspirational motivation, intellectual stimulation, idealized influence, and individualized consideration). Therefore, we suggest exploring examine more detail the components of transformational leadership because each component can exert different effects on organizational performance (Deinert et al., 2015). For example, future researchers can measure transformational leadership by a multidirectional scale of Antonakis et al. (2003). These work will help researchers have more evidence to better understanding the influence mechanism of transformational leadership on organizational performance, propose more detailed governance implications to improve performance (Katou, 2015;Paarlberg & Lavigna, 2010).