THE EFFECT OF MANAGERS’ POWER ON EMPLOYEES’ ENTREPRENEURSHIP: AN EMPIRICAL STUDY IN THE PUBLIC OFFICES OF IRAN

Objectives: The aim of this paper is to examine the effect of managers’ power on employees’ entrepreneurship in the public offices by using five dimensions of power (Coercive, expert, legitimate, referent, and reward) to give constructive suggestions for situation improvement through identifying powers affecting employees’ entrepreneurship. Design/Methodology/Approach: Considering the research objective, it is an applied study, and regarding data collection and the data analysis method, it is a descriptive-correlational study which has analyzed data through structural equations modeling with partial least squares (PLS) approach. Data were collected using two questionnaires including managers’ power and employees’ entrepreneurship. A sample of 600 employees was selected from the public offices of Iran. Data were analyzed using structural equation modeling and Amos software. Results: The finding revealed that reward power, referent power, expert power and legitimate power had a positive and significant effect on employees’ entrepreneurship, but the coercive power had not significant effect on it. Limitations: Among limitations of the current research are the big size of the statistical population that made it difficult to collect data, as well as the existence of different cultures in the statistical population which could partly affect the results of the research. Practical implications: Research findings can be used to improve the entrepreneurship of employees in government organization. Originality/value: Authors confirm that the current research and its results are genuine and have been published nowhere so far. The proposed structural model in the current research can be used in government departments and improve the entrepreneurship status of employees in the organization.


INTRODUCTION
At the present, rapid changes in science and technology have created a lot of problems in social and economic systems. These changes have created threats and opportunities that require creative solutions when faced.
Economists, sociologists and management scientists have proposed different definitions and conceptual frameworks for the entrepreneurship process. The study of entrepreneurship can be defined as the study of opportunity resources. In fact, entrepreneurship process is defined as the process of discovery and assessment, and the use of job opportunities (De Amorim Braga et al., 2018). Entrepreneurship is one of survival forms of today's organization. entrepreneurship as the process of creating something different from value, by devoting the necessary time and effort, assuming the accompanying financial, psychological and social risks, and receiving the resulting rewards of monetary and personal satisfaction.
Due to the importance of entrepreneurship, the government has provided much Entrepreneurial assistance, such as funding, physical infrastructure and business advisory services through various entrepreneurial bodies, such as the Graduate Entrepreneur fund, the National Institute of Entrepreneurship, and the Permodalan Usahawen Dumiputera Nasional through the "skim Garduan", which is a pioneer graduate program (Muhammad et al., 2011;Sandhu et al., 2010). Therefore, adopting an attitude to ensure the organization's success is necessary (Daft, 2000). Entrepreneurship is an opportunity-based way of thinking and acting (Jena and Sahoo, 2014). In spite of this public belief that entrepreneurship only exists in the public sector, studies have shown that this category is also seen in nonprofit organizations (private sector). Currently, organizations in which the flow of innovation and dynamism exists are more successful than others. To achieve such a situation a manager and leader who can significantly affect the employees' performance is needed (Analoui et al., 2009). They can affect the employees' attitudes through the powers they have at their disposal and lead to the occurrence of different behaviors in the organization, such as strengthening entrepreneurial characteristics, reducing absence from work, and increasing efficiency (Watson and Papamarcos, 2002). Employees show mutual behaviors against managers (Coyle-Shapiro and Marrow, 2006); therefore, managers can attain competitive superiority through human resources with entrepreneurship characteristics and guarantee their success. On the other hand, in order to cope with the sudden changes and competitive space, it is essential to adopt an entrepreneurial attitude to ensure the organization's success. To this purpose, managers who are capable of properly using their power resources and tools are needed.
On the other hand, the power of managers is one of the concepts that affect the performance of the organization and management. Montana and Charnov (2008) state that power is composed of five components, such as coercive, expert, legitimate, referent, and reward. Generally, it should be said that the use of power requires a special skill and it can be considered as an art to increase and strengthen the given cases in government organizations. Power may be used vertically or horizontally in social systems. It is likely that power is interpreted among organizational units and their relationships, and/or gains meaning in the relationships between the upstream and downstream (Pfeffer, 1994).
Some questions may arise in our minds: what is power? Where can it be seen in organizations? Several centuries ago Machiavelli wrote his celebrated book, The Prince, in which he dealt with the principles of power as used by rules. Today power is an issue that has been the focus of attention for those studying politics over the years. in 1959 French and Raven defined the different types of power.
Power is a neutral force. It may be perceived as either positive or negative, depending on the frame of resource, life experiences, and objectivity of the people involved. Mintzberg (1983) sees power as the ability to achieve organizational outcomes; the manager in the role of leader has a significant impact on the performance of a group. Undoubtedly, power is one of the ways in which leaders and managers can influence the behavior of their followers. In fact, power is a key for manager to provide an effective basis in the organization. The managers are policy makers in the organizations and the performance of these organizations has been related to their power. In today's world, organizations' stability faces many threats. Perhaps some of these threats can be covered by innovation and entrepreneurship (Minatogawa et al., 2018). Considering the materials reviewed and given the conditions and situation of government departments, as well as regarding the existence of massive capitals and God-given resources, unfortunately, there are a lot of problems in Iran's government departments whose main reason is the improper use of human recourses, and it seems that by using the proper power bases by managers and modifying managers' selection in these organizations, some of the problems can be resolved. Therefore, the problems mentioned above caused the researcher to investigate the effect of managers' power resources (legitimate, reward, coercive, referent, and expert) on the employees' entrepreneurship with the aim to improve the status of these organizations so that at least some constructive suggestions can be presented to improve the status.

THEORETICAL BACKGROUND AND RESEARCH HYPOTHESES
In the following paragraphs, the literature of managers' power and employees' entrepreneurship are studied.

Research on managers' power
The study of managers' power and its impact is very important to understand organizations' performance.
The way organizational sub-units and employees are controlled is related to the issue of power and influence. Originated in Latin, power, which means to be able, has been associated with the ability to produce desired outcomes (Guinote, 2010). It is often defined as the ability of a person to influence others (De Dreu and Van Kleef, 2004;French and Raven, 1959;Lewin, 1941;Vescio et al., 2003) or control outcomes by providing or withholding resources valuable to others (Galinsky et al., 2003;Keltner et al., 2003;Thibaut and Kelley, 1959;Weber, 1947). The study of power and its effects and the way it is used in the government organizations is a very important issue (Pinnow, 2011). The way in which the organization's sub-units and individuals are controlled is associated with the power issue and its effect. It should be said that power is a function of influence without threat (Erkutlu and Chafra, 2016). Influence is associated with respect in this regard (Drea et al., 1993). For example, people's needs are satisfied when they are respected, so it can be argued that the managerial power is positive in this regard (Cho, 2006). In this study, the theory of French and Raven (1959) has been selected to managers' power. French and Raven have presented the most important analysis on power resources. They have identified five powers, including legitimate, reward, coercive, referent, and expert powers in organizational environments that are considered as the basis of many research studies. Legitimate power is the power that originates in the organizational position and status of individuals. Reward power refers to the ability to provide things that others like. Coercive power is based on the employees' fear and horror and is dependent on fear and punishment. Referent power is based on the individual's personality, and it is said that whenever the employees attribute good characteristics, such as honesty and trusteeship to their boss, the boss has the referent power. Expert power is the ability to control the behavior of another individual through knowledge and the experience he/she lacks but needs to have.

Research on employees' entrepreneurship
Entrepreneurship has been recognized as the engine of countries' development. (Schumpeter, 1934), Referred to as the "Fundamental phenomenon of economic development", entrepreneurship is responsible for technological innovation; exploitation of opportunities; and coping with uncertainty and risk taking.
Researchers in Entrepreneurship studies, factors affecting entrepreneurial behavior divided into three categories: social, environmental, and personal. In the Social factors model, factors such as personal history, family history, class career, recent experience of life, growth environment, and etc. are important. The environmental factors model, in turn, focuses on factors such as money value, tax, and indirect benefits.
On the other hand, the individual factors model that is widely known as features model focuses on personality traits (Gurol and Atsan, 2006). The approach features try to identify characteristics of entrepreneurial abilities through assessment personal capability, characteristics and Motivations (Haase and Lautenschlager, 2010). In fact, entrepreneurship is the key element to increase the innovation capacity in organizations and realize the competitive advantage in them (Castrogiovanni et al., 2011). Organizations can be innovative, risk takers and active through the activity of their members. The background of entrepreneurship studies shows that entrepreneurs have characteristics and attitudes that provide them with motive force and lead to their superiority over others (Matviuk, 2010). During the last few decades, researchers have been trying to discover the personality traits of entrepreneurs. The main axis of most of these research studies is motivation for advancement, creativity, self-confidence, risk taking, control focus, and accountability.
In the present study, after a comprehensive review of Entrepreneurship, studies following entrepreneurship features were selected as components of employees' entrepreneurial. Creativity is the creation of new thoughts and ideas (Furnham and Bachtiar, 2010). Need for achievement means the tendency to perform works in the best way and trying to reach the peak of success (Olakitan and Ayobami, 2011). Internal locus of control means the belief that incidences are the result of human performance, not environmental factors (Arslan et al., 2009). Risk taking means acceptance of calculated risks. Ambiguity tolerance is the tolerance of ambiguous situati ons and making the right decisions in these circumstances (Gurol and Atsan, 2006).

background research
Lots of studies have been done on power and entrepreneurship and some of them are menti oned below. In marking channels, Brown et al. (1995) investi gated the impact of the supplier's use of power on the retailer's commitment to the channel relati onship. They adopted the mediated (i.e. reward, coercion, legal, and legiti mate) and non-mediated (i.e. experti se, referent, informati on, traditi onal, and legitimate) typology of John Sonn et al. (1993). Newstrom and Davis (1997) also argued that the coercive power is likely to have an overall negati ve impact on the receiver. Jalilvand and Nasrolahi Vosta (2015) investi gated managers' power and aff ecti ve organizati onal commitment. They found two major fi ndings in this research. First, the relati onships among expert power, legiti mate power, reward power, referent power, and aff ecti ve commitment are positi ve and signifi cant. Second, the construct of the coercive power was not associated with the employees' aff ecti ve commitment. The fi ndings suggest that managers' power relates with a social exchange relati onship where employees exchange positi ve outcomes, including strong aff ecti ve commitment. When people perceive manager power, they feel more aff ecti vely att ached to their organizati ons. In the view of Smith et al. (2010), environment has a profound impact on the human atti tude toward himself/herself and others. Therefore, how the organizati on manager treats with individuals and the type of power used by manager contribute to the emergence of entrepreneurial behaviors as a fundamental facilitator. In their study, Jong and Hartog (2007) showed that the innovati ve behavior of entrepreneurs and the creati ve thinking of an organizati on's members (employees) are aff ected by the way in which the manager behaves toward the organizati on members. And the type of established communicati on, management style, and the power they use to deal with the organizati on employees play a pivotal role. Finally, Bell conducted a research in 2001 and concluded that managers and supervisors who have used appropriate management and supervisory procedures (appropriate power) have received far more performance and effi ciency from their employees.
Emanati ng from the previous discussion, the following hypotheses were developed (see fi gure1): • H2: legiti mate power has eff ect on the employees' entrepreneurship.
• H3: Referent power has eff ect on the employees' entrepreneurship.
• H4: Reward power has eff ect on the employees' entrepreneurship.
• H5: Coercive power has eff ect on the employees' entrepreneurship.

METHOD
Considering the research objecti ve, the current research is an applied study and, regarding data collecti on and the data analysis method, it is a descripti ve-correlati onal study which has analyzed data by structural equati ons modeling with the parti al least squares (PLS) approach. It should be noted that in the structural equati ons, the researcher analyzes data in three steps. In the fi rst step, the Confi rmatory Factor Analysis (CFA) is investi gated and if the measurement model has appropriate validity and reliability, the structural model of the research is studied. If it does not fi t, the measurement model is modifi ed and this process is performed again. As aforementi oned in the current research, the measurement model (CFA) was fi rst investi gated, and aft er confi rming the tool's validity and reliability, the structural model (path analysis) was used in order to test hypotheses.
This research focused on the public offi ces of Iran (in western Azerbaijan province). Data were collected from the public offi ces of sixth towns. The demographic variables of the respondents' sample were extracted by asking on gender, age, fi eld of study and the number of years they have worked in organizati on. Table 1 shows the sample popula-1. Load factor must be significant 2. Load factor must be above 0.5 3. AVE > 0.5 4. CR must be above 0.5 and CR>AVE 5. AVE > MSV 6. AVE >ASV

Number of years worked with current manager
Less than 3 years 13 3-6 years 20 7-10 years 25 More than 10 years 42 Confirmatory factor analysis (CFA) of the managerial items (i.e. 20 items) was conducted in order to check for construct independence. Based on the results of a CFA, the data supported the independence of five factors, namely; reward power (four items, ); referent power (four items, ); coercive power (four items, ); expert power (four items, ); and legitimate power (four items, ). One item from the coercive power was dropped due to low loading of the order of or less than 0.25.
Employees' entrepreneurship -Employees' entrepreneurship has six components and each of them was measured with four questions. These questions were made by the researcher and by all the tests that were approved. The Cronbach's for components was above 0.7. CFA results revealed that all of the items had significant positive loading on the CFA at the p<0.05 level. The participants were asked to assess their entrepreneurial characteristics on a sev-tion for this research. All the individuals in the sample were full-time employees of the participating organizations and had good cooperation in the present study. Respondents were composed of two parts, a part of site managers and deputies who responded to the power questionnaire and the other part were the employees that responded to the Entrepreneurship questionnaire.
the questionnaires were written in Persian, and were distributed to 600 people, of whom 200 people of them were managers and 400 were employees. It should be noted that as the sampling has been done in two groups of people in the statistical population, the data packaging method was used to prepare data to be entered into the software. Questionnaires were coded before distribution, and the mean of the employees' data related to each manager was entered into the software to compare them to manager data.
All of the 200 managers returned the questionnaires filled and 390 of the 400 employees returned the questionnaires filled, thus yielding 98 percent of the responses. Average age of managers was 42 (standard deviation=6.52) and tenure was 6.45 (standard deviation=5.12) years; and for employees they were 34.82 (standard deviation=6.31) and 6.63 (standard deviation=50.41) years, respectively.

Measures
Managerial power -In this paper, managers' power, described by using French and Raven's (1959) power, include: coercive, expert, legitimate, referent, and reward power. French and Raven's (1959) managers' power was measured using a modified version of Hinkin and Schriesheim's (1989) power scales.
For example, the following issues were applied: My Manager can "increase my pay level" (reward), "make things unpleasant here" (coercive), "provide me with needed technical knowledge" (expert), "make me feel that I have commitments to meet" (legitimate), and make me feel personally accepted" (referent).
The managers' power was evaluated on a seven-point response scale ranging from "agree" to "disagree". The validity and reliability of the researcher-approved questionnaires have been examined. To confirm the validity of the questionnaire, three cases were checked: the apparent validity, the content validity, and the construct validity. The apparent validity was confirmed by professors' management, the content validity of the questionnaire was confirmed by CVI and CVR indicators. The construct validity of the questionnaire was approved and it is described in the next section. Hair (2011) states that in order to verify construct validity the following conditions would have existed. en-point response scale (ranging from "strongly agree" to "strongly disagree").

Data collection procedures
Before the implementation, the questionnaires were pre-tested, using a small number of respondents. Their reliability and validity were assessed. To collect the data, questionnaires were distributed and then collected by the researcher. The participants were all given the opportunity to ask questions and were encouraged to answer the survey honestly; anonymity was guaranteed and no names or other identifying information was asked.

Validity and reliability of measures
The Confirmatory factor analysis (CFA) was used to assess the research questions in order to determine whether the questions of this research are capable of assessing variables. As shown in the Table 2, factor loadings are featured for all the questions above 0.5 in one of the questions of the coercive power. Also other conditions of validity and reliability are satisfying for this research, as shown in the Table 2.

Analytical procedure
Data analysis was conducted by descripti ve stati sti cs by using structural equati on modeling and AMOS (version 24) structural equati on program. Amos is designed to esti mate and test structural equati on models (SEMs). SEMs are stati sti cal models of linear relati onships among latent variables and observed variables. Its purpose is esti mati ng the coefficients in a set of structural equati ons. The path coeffi cients are tested for signifi cance and goodness of fi t. The esti mati ons of the parameters and the overall fi t index of the measurement model are based on the maximum likelihood (ML) method. The basic conditi ons adopted for the use of ML estimati on (Byrne, 2001) are met or approximated in the study. Furthermore, the sample is suffi ciently large (n=195 for manager and n=285 for employees). Over the recommended size of 200 cases (Medsker et al., 1994), the observed variable scale is conti nuous and no violati ons of multi variate normality are found in the survey responses. The factorial validity of the measurement model was assessed using CFA. Given the adequate validity coeffi cients of those measures, the number of indicators in the model was reduced by creati ng a composite scale for each latent variable (Hair et al., 1995).
The measurement of the suitability of the model was used to evaluate the fi t of the structural model. To esti mate the suitability of the model (goodness of fi t) indices (GFI) for measurement and structural model, X 2 -test was used.
In additi on, the root mean square error of approximati on (RMSEA) was used as an absolute fi t index. The incremental fi t index (IFI), the Tucker-Lewis index (TLI), and the comparati ve fi t index (CFI) were used as an absolute fi t index. The incremental fi t index (IFI), the Tucker-Lewis index (TLI), and the comparati ve fi t index (CFI) were used as incremental fi t indices. For GFI, AGFI, CFI, and TLI, coeffi cients closer to unity indicate a good fi t, with acceptable levels of fi t being a bare 0.90 (Hair, 2011). Root mean square residual (RMR) and RM-SEA evidence of good fi t are considered to be values lower than 0.05; values from 0.05 to 0.10 are indicati ve of moderate fi t and values greater than 0.10 are taken to be evidence of a poorly fi tti ng model (Browne and Cudeck, 1993). Standardized esti mates were used in reporti ng the causal relati onships between the exogenous and endogenous constricts.  As reported in Table 2, All hypothesis were accepted, except H5. Thus, all components of the managers' power predict employees' entrepreneurship, except the coercive power. Reward power had the greatest eff ect on employees' entrepreneurship and the legiti mate power had the least effect on employees' entrepreneurship.

DISCUSSION AND CONCLUSION
This paper has investi gated the eff ect of managers' power on employee's entrepreneurship in the public offi ces in Iran. Becoming aware of the infl uence of power on an organizati on is an important step in the achievement of the organizati on (Popovich and Warren, 2010). It is also necessary to be aware of the bases of employees' behavior and, this way, employees' entrepreneurship is strengthened.
It must be understood that Entrepreneurship is one of the soluti ons for the collapse of today's organizati ons. The results of the present study show that all components of the managers' power had a positi ve and signifi cant impact on employee's entrepreneurship except the coercive power.
This means that expert power, legitimate power, reward power, and referent power (personality power) have a positive and significant influence on employees' entrepreneurship in the public offices. The power of managers in some studies has been investigated but the researcher did not find any research that investigated this variable with entrepreneurship. Some researchers (Malhotra et al., 2007;Williamson et al., 2009;Newman & Sheik, 2012;Tornikoski, 2011;and Jalilvand and Nasrolahi Vosta, 2015) have found that managers' power has positive and significant influence on organizational commitment. Other researchers, such Rajan and Krishnan (2002) and Popovich and Warren (2010) have verified that power influence on other variables is similar to this research. Many components of power have had positive influence on dependent variables. In fact, the use of expert, legitimate, reward, and referent power will strengthen employees' entrepreneurship. By giving reward, law enforcement, charismatic, and expertise properties, managers can provide the motivation for employees' entrepreneurship. Thus, manager's use of reward power and legitimate power provides the extrinsic motivation for employees' entrepreneurship and, with the use of expert power or referent power, the intrinsic motivation for employees' entrepreneurship is provided.
As these intrinsic factors become more central, extrinsic factors such as rewards and legal action may become less important (Brown et al., 1995).
The findings are useful for managers to affect the employees in order to strengthen the entrepreneurial characteristics of employees. This study had showed that reward power has a positive and significant impact on employees' entrepreneurship. By giving a reward to employees, managers create the extrinsic motivation for employee's entrepreneurship in organizations. When reward power is used in an organization, it can be converted into a strong motivator (Newman and Sheikh, 2012). Reward power does not need to be monetary or tangible compensation to work, as managers can give various intangible benefits as rewards (Malhotra et al., 2007). The study shows that expert, referent, and legitimate power have a positive and significant effect on employees' entrepreneurial. By using expertise, referent, and legitimate power, managers increase positive attitudes toward the organization by increasing the positive relationship between them and the employees (Brown et al., 1995).
Finally, our study revealed that coercive power is not effect on employees' entrepreneurial. However, Managers should remember that coercive power might be a negative impact on employees' entrepreneurship.
Coercive power decreases employees' satisfaction, thus leading employees not to like showing entrepreneurial behavior.
The use of coercive power results in an atmosphere of insecurity or fear and a decline in productivity and creativity (Harris and Hartman, 2002). Therefore, if you want to have an organization with employees' entrepreneurship, you should not make use of such coercive power. Finally, with regard to the Results mentioned, managers should use power in a positive direction to be able to encourage entrepreneurship among employees and manage their organization in the face of difficult conditions. More specifically, given the path coefficients obtained from the research structural model, in order to have a greater influence, it is suggested that government organizations under study use referent power, and especially, reward power that has higher than average path coefficients to strengthen entrepreneurship in employees, so that they can increase entrepreneurial characteristics and lead to the occurrence of entrepreneurial behaviors in employees. For example, by identifying the rewards affecting employees, such as job promotion, increasing responsibility, and financial rewards the entrepreneurial process is accelerated, or the positive relationship with employees is strengthened by gaining trust, secrecy, trusteeship and honesty, and, by creating the positive feeling in employees, the way is paved to create entrepreneurial characteristics in workers and increase an entrepreneurial behavior in them. Additionally, it is emphasized that coercive power should not be used at all and whenever the use of coercive power is inevitable, it should be justified to employees, so that they can understand that it has been used as the last resort to improve and restore order in the system.

LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH
One of the limitations of this research was gathering data, firstly because statistical society was vast and all offices must be considered, and secondly, because there were a lot of problems in data distribution and collection because of exciting special rules in each office that has to coordinate separately with each of them. The second limitation is about some variables such as Iran's economic conditions and culture of results. Therefore, it is recommended for future researchers to carry out this research in one of the provinces that is united in terms of ethnicity. And some variables should also be considered, such as culture.

ACKNOWLEDGEMENT
This work was supported by the West Azerbaijan Governor of Iran. Therefore, we appreciate authorities and managers for their support.