Decentralization and Managerial Performance: A Mediated Budget Participation on the Village Government in West Halmahera Regency

- The research aimed to examine the budgetary participation variable as a mediator on the relationship between decentralization and managerial performance. The sample was determined using a purposive sampling method with the consideration of being involved in budgeting. They were village heads in 159 villages in West Halmahera Regency. The data were collected by a survey with a questionnaire. The used analytical tools are simple regression and hierarchical regression as a statistical analysis and IBM SPSS statistics version 24. The results showed that budgetary participation fully mediates the relationship between decentralization and managerial performance. The conclusion provides three main contributions as well as their implications, namely providing new findings related to budget participation as a mediating variable, terms of age at work, and the level of education.


I. INTRODUCTION
Budgetary participation has received attention in finance and accounting in recent decades. The results reveal that budgetary participation can be influenced by decentralization (Riyadi, 2007) and affect managerial performance (Indriani et al., 2020), and managerial performance can also be influenced by decentralization (Sun et al., 2021). The involvement of managers or subordinates in decentralized organizations in budgeting will have an impact on their high success, satisfaction and performance (Amir et al. 2020;Mulyanah & Puspanita, 2021).
It shows that budgetary participation can be a mediating variable for the effect of decentralization on managerial performance. Although previous research results are partially still very limited (decentralization and budgetary participation), inconsistent (budgetary participation and managerial performance), and consistently positive (decentralization and managerial performance), the literature review results show that research on decentralization and budgetary participation has only been studied by Riyadi (2007). Furthermore, the results of research between budgetary participation and managerial performance have significant positive results (Purwanti et al., 2017;Soleiman & Thalib, 2020;Pradana, 2021;Hussein et al., 2021), negative significant (Bryan & Locke, 1967) and insignificant (Candrakusuma & Jatmiko, 2017;Kahar et al., 2019;Firana, 2020). Consistently, there is a significant positive relationship between decentralization and managerial performance (Purwanti et al., 2017;Amran, 2020). Moreover, research on decentralization, managerial performance, and budget participation in the context of village government is considered very limited.
Decentralization at the village level has been described in Law No. 23 concerning Regional Government (2014) and Law No. 6 on Villages (2014). There are two types of scope of village authority, namely: 1) the authority that can regulate and administer based on origin rights and 2) the authority that only manages but does not regulate local ones within the village scope (Aritonang, 2018). With this decentralization, there will be additional budgets for villages, so participation is needed which will certainly improve performance both individually and in groups (Fritantus, 2019;Runtunuwu & Tussabaha, 2020;Kepramareni & Pradnyawati, 2021).
The synthesis of previous findings show that a number of researchers are considered negligent in assessing budgetary participation as a mediating effect of decentralization on managerial performance in the context of village government, so this becomes the research novelty.
Decentralization refers to the transfer of state or national responsibilities or functions from government domiciled at the center to the level of government domiciled in the regions, or from head office to branch offices, or to private affairs (Guler, 2017;Al-Sharafi et al., 2019). Decentralization is the practice of delegating authority to a lower level (Mowen et al., 2018). Decentralization is believed to reduce the burden on top management, so that managers will diagnose the organizational situation and choose the most appropriate level of decision making for the organization (Kesumawati et al., 2019).
According to Mahoney (1963) managerial performance is the accumulated performance of organizational members in managerial activities. This can be seen from the final achievement of an activity or program, policies in realizing the strategic planning of an organization (Ainanur & Tirtayasa, 2018) and is responsible for product quality, personnel development, budget achievement, increased revenue, and public affairs (Bedford & Speklé, 2018).
Budget participation is the frequency and scope of discussions from managers to subordinates regarding budgeting (Nguyen, Evangelista & Kieu, 2019). Participation in the budgeting process is an effective approach to increase manager motivation (Giusti et al., 2018). Participation in budgeting is related to how far the involvement of the leadership or head of the organization in determining or preparing the budget for the part or organizational unit (Aira, 2021). High participation tends to encourage managers to be more active in understanding the budget (Hartmann et al., 2020).
Decentralization can be regarded as a form of widespread distribution or delegation of decisionmaking power to levels (Sistiyan et al., 2019). Organizational structure has a role in influencing performance at the organizational level and sub-unit level which is decentralized, since managers have better understanding on the condition of the unit they lead (Hartmann et al., 2020). Decentralized power sharing allows managers to have the opportunity to actively participate in the budgeting process which in turn will provide motivational support to subordinates to improve their performance (Firmansyah & Mahardhika, 2018). The results also show that decentralization has a significant positive effect on managerial performance (Purwanti et al., 2017;Soleiman & Thalib, 2020;Amran, 2020). For this reason, the proposed hypothesis is formulated: H 1 : Decentralization has a positive effect on managerial performance Dakhli (2021) shows that managers or subordinates in decentralized organizations feel themselves to be more influential people, participate more in budget planning, and feel satisfied with budgetrelated activities. This is also supported by the results of research by Riyadi (2007) that the relationship between decentralization and budget participation is positive, so the hypothesis is: H 2 : Decentralization is positively related to budget participation Budget participation is defined as, "The involvement of managers in preparing the budget at the responsibility center of the manager concerned" (Kahar et al., 2019). Managers who are involved in the budgeting process will form positive attitudes and traits as they will feel satisfied and appreciated (Candrakusuma & Jatmiko, 2017). This certainly has an impact on increasing performance and being more responsible for their work. The results show that there was a positive and significant relationship between budgetary participation and managerial performance (Purwanti et al., 2017;Soleiman & Thalib, 2020;Pradana, 2021;Hussein et al., 2021). The hypothesis is proposed as: H 3 : Budget participation has a positive effect on managerial performance The granting of authority from the highest manager to the lowest manager to be involved in budgeting will have an impact on high performance both directly and indirectly (Dakhli, 2021;Kahar et al., 2019;Nguyen, Evangelista & Kieu, 2019;Murhaban & Adnan, 2020). According to Baron & Kenny (1986), a variable can be a mediating variable if there is a significant influence either directly or indirectly. Hypothesis 1, 2, and 3 indicate that budgetary participation can be a mediating variable on the relationship between decentralization and managerial performance, so hypothesis 4 is proposed: H 4 : Budgetary participation mediates the relationship between decentralization and managerial performance

II. METHODS
The research aims to re-test the hypothesis of previous studies by adjusting it in the context of village government, so the research applies a confirmatory approach. Data collection is carried out using a cross sectional method, namely collecting or retrieving data at a certain time (Bougie & Sekaran, 2019). The data collection technique used is survey, namely asking research respondents using a written questionnaire distributed directly to respondents (Neuman, 2020).
The research is conducted in West Halmahera Regency from March to May 2021. The population are 175 village heads in West Halmahera Regency. The sample is determined by purposive sampling method which has involvement in village budgeting totaling 159 respondents. According to Roscoe et al. (1975) the ideal number of samples in a study range from 300-500 respondents, so the total number of samples is considered appropriate. The research uses a validity test using factor analysis with a loading factor value of 0,5 and a reliability test using Cronbach's alpha 0,7 (Hair et al., 2018). Furthermore, testing hypothesis 1, 2, and 3 uses simple regression analysis, while for hypothesis 4 using hierarchical regression analysis (Baron & Kenny, 1986;Hair et al., 2018) by using IBM SPSS statistics version 24. The simple linear regression formula is as follows: Information: X = Decentralization and Budget participation Y = Managerial Performance and Budget Participation a = Constant b = Regression Coefficient Furthermore, the mediation hierarchy regression consists of three equations as: Variable M is called a mediator if it meets the criteria, namely (2) X significantly affects Y (c ≠ 0), (3), X significantly affects M (a ≠ 0) and (4), M significantly affects Y (b ≠ 0).
The questionnaire is adopted from previous research. For decentralization variables using a questionnaire adopted by Gordon & Narayanan (1984), then for managerial performance variables using a questionnaire developed by Mahoney (1963) and for budgetary participation variables using a questionnaire adopted by Milani (1975). The three variables use a Likert scale of 5, from strongly disagree to strongly agree. Specifically for the decentralization questionnaire, the researcher adjusted the questionnaire to the context of the village government.

III. RESULTS AND DISCUSSIONS
The results of the questionnaire distribution shows that from a total of 175 questionnaires distributed, only 163 (93,14%) questionnaires are returned and 159 (90,86%) questionnaires are declared eligible for the data testing stage. Based on the results of the distribution of this questionnaire, the response rate in the research is 93,14%.
The overall profile of the respondents in the research was male (159 respondents or 100%), with most of them ranging in age from 31-40 years (84 respondents or 53%) with the majority of the education level being high school graduates (152 respondents or 96%). The profile of the respondents is shown in Table  1. It shows that the sample in the research reflects the ideal age of the workforce (BPS, 2021; Arifin, et al., 2020). In addition, the respondents are also dominated by high school graduates who generally had a minimum education standard to be able to work as village officials (Wijayanti & Hanafi, 2018). Table 2 shows the results of testing the validity and reliability for the variables of decentralization, managerial performance, and budgetary participation. The validity test shows that the Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy value and the loading factor 0,5 are good for the variables of decentralization, managerial performance, and budget Source: Processed data participation. It shows that the three variables are considered to be valid. Furthermore, for the results of reliability testing in the research using the Cronbach's alpha value 0,7. Based on the results of the reliability test in Table 1, it shows that the three variables studied have a value greater than 0,7, so the three variables are declared reliable (Hair et al., 2018). Table 3 shows that respondents' perceptions of decentralization, managerial performance, and budgetary participation. The results show that respondents agree that there is a delegation of authority at the village level. In addition, respondents also agree that the village head has a good performance. similar to decentralization and managerial performance, village heads also agreed to always be involved in budgeting. Table 4 shows the results of testing hypothesis 1, 2, and 3. Hypothesis 1 shows that decentralization has a positive and significant effect on managerial performance (β = 0,568, t = 9,687, P < 0,05). Hypothesis 2 also shows that decentralization has a positive and significant effect on budget participation (β = 0,580, t = 8,112, P < 0,05). Similar to hypothesis 1 and 2, hypothesis 3 also shows that budget participation has a positive and significant effect on managerial performance (β = 0,705, t = 17,344, P < 0,05).
Finally, the results of testing hypothesis 5 show that budgetary participation mediates the relationship between decentralization and managerial performance. It can be seen from the direct influence between decentralization and managerial performance of 0,568. When the mediating variable of budget participation is added, the value increases to 0,409, so that budget participation in the research fully mediates the relationship between decentralization and managerial performance.
The results indicate that decentralization is positively related to managerial performance. Higher the level of decentralization given by the leader to subordinates or managers, the higher the managerial performance obtained by the manager. This is in Table 2 Validity and Reliability Tests  Bruns and Waterhouse (1975) that a manager or subordinate in a decentralized organization feels himself a more influential person, participates more in budget planning, and feels satisfied with activities related to the budget so that it will increase self-confidence and performance. The results are also similar to those of Purwanti et al. (2017), Soleiman and Thalib (2020), and Amran (2020) that decentralized power sharing encourages managers or subordinates to actively participate in the preparation of budget. The participation given by the manager in the preparation of the budget will result in an increase in managerial performance to be achieved by the manager or subordinate.
The research also provides results in accordance with the hypothesis that decentralization has a positive effect on budget participation. Luthans (2002) argues that the existence of decentralization provides relevance at lower levels of participating in decision making. With decentralization there will be employee empowerment since employees are more involved in various activities, especially in decision making. Apart from this, decentralization will also motivate subordinates to play an active role in every operational and managerial activity of the company, which in the end is an extremely important part of productivity. These results also confirm previous research that decentralization has a positive influence on budget participation, since higher level of decentralization in decision-making given to managers increases their participation in budgeting involvement (Riyadi, 2007).
The results show that budgetary participation is significantly positively related to managerial performance. It shows that the role of participation in the preparation of the budget can be related to the process of involvement of managers in preparing the budget and its implementation to achieve certain targets. Direct involvement of managers in budgeting will have an impact on budgeting. This is in accordance with the theory put forward by Nguyen, Evangelista & Kieu (2019) that budgetary participation will have an impact on performance when managers are able to be further involved in budgeting. This refers to the frequency of effective consultation between managers and subordinates on budgets. The results are also in accordance with the results of research by Purwanti et al. (2017), Soleiman andThalib (2020), Pradana, (2021) and Hussein et al., (2021) that involving managers in budgeting will ultimately improve their performance. Table 5 shows that the budget participation variable can be a full mediating variable both directly and indirectly on managerial performance, especially in the context of the village head, especially the village government of West Halmahera Regency. The result also strengthens the theory of the relationship mechanism described previously in Hypothesis 4 (Dakhli,

IV. CONCLUSIONS
The research yields three main contributions as well as implications. First, the research provides new findings, namely budgetary participation as a mediating variable on the relationship between decentralization and managerial performance in the context of village government. Results also contribute to the management and accounting literature, especially on decentralization, budgetary participation, and managerial performance. Second, in terms of age, research provides evidence that the productive age in developing villages ranges from 30-40 years. Third, the preliminary descriptive analysis has shown a tendency in the educational aspect to show that the level of education does not guarantee that higher education is capable of managing village government better than those with lower levels of education.
It can be concluded that: 1) decentralization has a positive effect on managerial performance, 2) decentralization has a positive effect on budget participation, 3) budget participation has a positive effect on managerial performance, and 4) budgetary participation fully mediates the relationship between decentralization and managerial performance. In addition, the research also has limitations and recommendations for future research. The research uses measurements for managerial performance variables that are personal assessment, so it reveals that managers or village heads in assessing their own performance. Therefore, it is suggested that future research express performance using a 360-degree assessment. In addition, the research can examine several other variables to see the mediation between the relationship between decentralization and managerial performance such as self-efficacy, organizational culture, management accounting system, environmental uncertainty, and internal control which have been neglected to be studied in the context of village government.