Employing Structural Equation Modeling to Examine the Determinants of Work Motivation and Performance Management in BUMDES: In Search of Key Driver Factors in Promoting Sustainable Rural Development Strategies

Abstract

This study aimed to analyze the influence of local politics, village facilitators, recruitment of administrators, training and education, and organizational culture on work motivation and management performance. The study was conducted in Wajo Regency, South Sulawesi Province, Indonesia, utilizing primary data collected from 250 participants, including administrators of village-owned enterprises (BUMDES), community leaders, and representatives from the private sector. The data were analyzed using structural equation modeling (SEM) with the LISREL program. The results indicated that the latent variables of local politics, village facilitator, recruitment of administrators, training and education, and organizational culture had a positive and significant impact on work motivation and management performance. These findings are valuable key indicators and provide essential insights for promoting and driving the BUMDES as a pillar of rural development strategies. Based on these findings, it is recommended that the local government revitalize the local political system, reorient the organizational culture of the BUMDES toward a modern business-oriented culture suited to rural conditions, and enhance the training and education of village facilitators to improve their motivation and performance. This recommendation will empower the BUMDES to promote rural economic improvement and sustainable rural development by enhancing work motivation and management performance.

1. Introduction

Rural development encompasses initiatives and interactions among physical, technological, economic, socio-cultural, and institutional elements [1]. It aims to promote growth, retention, poverty alleviation, and expansion in rural areas, ultimately enhancing the quality of life for rural residents. Rural development policies should capitalize on the government’s strategy to promote sustainable development in rural regions and alleviate poverty nationwide [2]. The Sustainable Development Goals (SDGs) aim to eradicate poverty in all its forms. One approach to achieving this objective is to strengthen local communities, thereby mitigating the effects of poverty in rural regions [3]. To revitalize the country’s economy and reduce poverty, the Indonesian government has designated village-owned enterprises (BUMDES) as a front-runner in local economic development [4,5]. A spirit of a family-oriented system governs the BUMDES, and a locally based cooperation institution is established through village meetings [6]. The BUMDES performs two main functions for the hamlet: social and economic supporting functions. The BUMDES primarily focuses on social service provision, rather than the profit-driven nature of the commercial function [7]. Through the implementation of a cooperative, participative, emancipatory, transparent, accountable, and sustainable form of productive management, BUMDES can facilitate local economic independence for rural communities [8]. The village is led by a village head who is democratically elected by the villagers [9].

Village authorities are responsible for coordinating, planning, executing, utilizing, developing, and preserving development within the village; managing the local government’s administration; and fostering the village economy [10]. The BUMDES, an economic organization serving the village, is granted the authority to be administered by the villagers, as per Government Regulation No. 11 of 2021 on village-owned enterprises. Empirically, the BUMDES also increased PAD (village original income) in Bandung Regency’s Pameingpeuk District [11]. According to research conducted by Pariyanti [12] in the BUMDES Tirta Kencana, villagers engaging in marine gardening and cage gardening could contribute significantly to the community’s income and meet the community’s needs. Research in Ponggok Village, Tlogo, Ceper, and Manjungan Klaten Regency found that the BUMDES effectively improves people’s livelihoods [13]. The BUMDES was set up to meet the community’s needs and increase PAD (village revenue) [14].

Despite being supported by a solid legal framework, the implementation of BUMDES in the field still faces various obstacles. One of the main obstacles is political interference at the local level. Abdullah [15] argues that the involvement of village elites in the decision-making process often leads to practices that are not transparent and accountable and which disregard the principle of participation because they prioritize the interests of particular groups [16]. Hapsary and Fyniel [17] demonstrate that competent facilitators can drive improvements in financial reporting quality and governance transparency. However, variations in the number and quality of facilitators across regions pose challenges for standardizing BUMDES management at the national level. Another issue relates to the recruitment mechanism for BUMDES administrators, which often fails to consider their background experience and technical competence. Fajar et al. [18] state that the selection process is still frequently conducted informally or based on personal connections (nepotism) rather than expertise. This impacts the weakness of management systems and the sustainability of village enterprises. However, training and education programs have shown positive results. Ansori et al. [19] noted that regular and structured training can improve the managerial and financial capabilities of BUMDES managers, which in turn has an impact on increasing village revenue. Organizational culture is also a crucial factor in BUMDES’ governance. Anjeli et al. [20] found that a work culture that upholds the values of participation and cooperation can increase individual motivation while strengthening the collective performance of the organization. Conversely, a weak organizational culture tends to cause internal conflicts and a decline in performance. In line with this, Muizu et al. [21] emphasize that strengthening BUMDES institutions cannot be achieved through a structural approach alone and that such approaches must also be accompanied by strengthening local cultural values that are in line with the principles of sustainable development and environmental sustainability [22]. Aeni [23] highlights that low accountability in financial management remains a major challenge, which presents various administrative obstacles.

In Wajo Regency, South Sulawesi Province of Indonesia, poor performance was frequently encountered in many aspects when implementing the BUMDES. There were still many mistakes in the BUMDES management, according to the 2021 Inspectorate Report on the subject of BUMDES. These include accountability reports that did not comply with the provisions, business units that did not have good human resource analyses, no profit and loss statements, and management capacity that did not meet BUMDES’s requirements [24]. For the BUMDES to be effectively managed, all relevant parties must pay close attention to the project and provide direction in line with their roles and duties [25]. The BUMDES frequently encounters issues such as insufficient communication with village elites, subpar human resource quality stemming from challenges in identifying individuals with effective managerial capabilities, an emphasis on physical and administrative development over community empowerment and participation, and a neglect of the village’s inherent strengths and weaknesses [26].

Another problem with BUMDES is that local elites, such as village chiefs and community leaders, often intervene in managing the BUMDES by leveraging their social status. According to Rahmadani et al. [7], their findings show that the management of village-owned enterprises often experiences political intervention from village officials. This is contrary to the core principles of BUMDES, which are cooperative, participatory, emancipatory, transparent, accountable, and sustainable [27,28]. Using their influence and connections, players execute a variety of interventions. Consequently, BUMDES will fail to fulfill their commitment as trailblazers in the economic rejuvenation of rural regions if this issue remains unaddressed [29]. The state will suffer a considerable loss due to the mismanagement of large amounts of local funds channeled through the BUMDES [30]. Therefore, the purpose of this study is to examine the influence of local politics, village facilitators, administrator recruitment, training and education, and organizational culture on work motivation and management performance. This analysis is based on empirical facts and issues encountered during the implementation of BUMDES, which support the findings of this study. The findings of this study are expected to provide significant indicators and profound insights that will encourage and facilitate rural development. This study is not only theoretical but also based on empirical data from the field. It is hoped that the results of this study will make a significant contribution to stakeholders in building more resilient villages and strengthening the strategic role of BUMDES as drivers of the local economy. The results of this study are also expected to have a meaningful impact on improving the work motivation and performance of BUMDES by identifying key factors that influence these aspects.

2. Literature Review

2.1. Local Politics

Reform severed ties between the central and local political elites, which benefited the local elites. As a result, achieving democracy at the regional level proved to be challenging [31]. To be independent, communities need to build social and economic strength, enabling them to innovate and determine their direction in life [32,33]. Effective management of village-owned enterprises (BUMDES) is crucial for enhancing the village economy [34]. Empowered communities can shift power, become agents of change, and drive sustainable transformation [35,36]. The contemporary literature has increasingly focused on the impact of local politics on organizational performance, notably via the lens of political embeddedness. In developing nations, the interplay between organizations and local political actors is crucial for enhancing legitimacy, broadening access to public resources, and ensuring regulatory protection [37,38]. However, this relationship is ambivalent. When political ties foster stability and symbolic support, they can enhance work motivation by providing a perception of external support, job security, and organizational clarity [39,40]. However, if these relationships are perceived as forms of political pressure or intervention, work motivation may decline due to role uncertainty and value conflicts [41]. In this context, work motivation acts as a psychological mediator explaining how complex political embeddedness can have dual effects on organizational performance.

Empirical evidence suggests that the impact of motivation on political embeddedness is highly contextual and contingent upon the nature of the relationship established. When the relationship is founded on trust and reciprocity, it has a positive impact on performance by increasing affective commitment and orientation toward shared objectives [42,43]. Conversely, political embeddedness that is too deep without being balanced by professional work ethics can hinder the structural effectiveness of the organization [44]. Therefore, this synthetic approach emphasizes the importance of psychology-based mediation analysis, particularly work motivation, to comprehensively understand the dynamics of local political influence on performance.

The power relations between local actors, including village presidents, village officials, community leaders, and other village institutions, which are involved in the decision-making and policy implementation process, are represented by local politics at the village level [45]. Previous research indicates that positive perceptions of village governance and community participation are highly dependent on how village elites conduct political practices and manage conflicts of interest [46]. In line with Foucault’s [47] thinking, power at the local level is not only exercised by formal actors but also distributed through informal social networks, which in practice is reflected in the dominance of village heads in managing village funds and the appointment of BUMDES administrators based on political proximity rather than professional competence [46]. While prior research has thoroughly examined the correlation between power and the quality of village governance, limited studies have emphasized the indirect impact of local politics on the performance of BUMDES administrators via work motivation. Work motivation is a crucial psychological factor that can influence the effects of power hierarchies on organizational outcomes [48]; thus, examining the interaction between political dynamics and motivation becomes vital in designing performance-based institutional governance models in rural settings.

Various recent studies confirm that local politics plays a central role in shaping the quality of village institutional governance, particularly in the management of village-owned enterprises (BUMDES). Anhari et al. [49] found that the selection process for BUMDES managers in several regions is often influenced more by political proximity than by managerial capacity, which ultimately disregards the principle of meritocracy and results in low efficiency, weak accountability, and an increased potential for the misuse of public funds. In a similar context, Anhari [49] noted that without clear political accountability, BUMDES are vulnerable to being misused as political transactional tools and are not directed toward achieving sustainable development goals. Local politics is not only shaped by formal institutions, but also through dynamic interactions between village bureaucracies, traditional elites, and informal groups, which significantly influence resource management and local policy. However, research integrating the influence of local politics with psychological aspects such as managerial work motivation remains limited [50]. Within the framework of Self-Determination Theory, motivation is understood as a function of individuals’ intrinsic motivation, which in turn contributes to stronger commitment and enhanced organizational performance [48]. Conversely, local politics that is run in a transparent, accountable, and participatory manner can strengthen the BUMDES governance system and encourage the improvement of its economic performance [51]. Therefore, this study occupies a strategic position in filling the literature gap by empirically investigating the role of motivation as a connecting pathway between local political dynamics and BUMDES management performance.

2.2. Village Facilitators

A village facilitator is not limited to implementing projects that enter the village or simply overseeing the use of village funds. The role encompasses comprehensive support for all aspects of village life and development [52]. Carlsson-Kanyama et al. [53] state that the role of facilitators in supporting sustainability at the local level can be maximized through a participatory backcasting approach. Participatory backcasting is a methodology in strategic planning that combines the active participation of stakeholders with a future-oriented approach [54,55]. This entails the involvement of stakeholders in developing a long-term vision. Effective village facilitators are responsible for visionary planning that focuses on developing local potential.

Zhu et al. [56] also highlighted the strategic role of local non-state actors, including village facilitators, in overcoming institutional exclusion and encouraging constructive informal engagement in the planning process. Village facilitators function as intermediaries capable of bridging unequal power relations in the implementation of development at the village level. Facilitators have tasks, including ensuring community participation, including the ability to build village discussion and decision-making forums, and the preparation of work plan documents and village meetings [54]. Furthermore, capacity building of village human resources, as per Carlsson-Kanyama et al. [53], namely, training (including training in the use of adaptive technology tailored to the needs of BUMDES), technical guidance, and mentoring of village officials, as well as the mediation of relations between villages and the government, which involves liaising between the aspirations of village communities and local or central government, and advocacy for collective village rights and ensuring alignment between local development and central government priorities.

Marsden and Murdoch [57] suggest that the formation of new economic entities in rural areas is a consequence of changes in rural governance that emphasize collaboration between state institutions and local communities. In this context, the BUMDES acts as a tool to encourage community participation and strengthen economic decentralization at the village level. Sisto et al. [54] elaborated that the success of local stakeholders in formulating adaptive and solution-oriented long-term strategies dramatically influences the success of BUMDES.

The relationship between the role of village facilitators and BUMDES performance is based on participatory development theory and empowerment theory. The facilitator not only functions as an educational facilitator but also plays a role in strengthening the village’s economic structure [53,56]. The participative backcasting approach has proven effective in building a collaborative planning platform that aligns BUMDES strategies with local socio-economic conditions [54]. In this case, the village facilitator plays a crucial role in facilitating the process. Although the concept of mentoring is considered ideal in encouraging village economic development through BUMDES, there is a possibility of technocratic bias that tends to overlook the social dimension and conflicts of interest at the local level. Therefore, it is necessary to examine the extent to which the village facilitator indeed carries out its role as a community capacity booster, not just an extension of the central government’s control [57].

Based on the above explanation, to achieve sustainable and beneficial management of the BUMDES for the community’s economic well-being, it can be suggested that village facilitators can assist the village government and community in designing, managing, and supervising BUMDES, thereby motivating management to work effectively and positively impacting the performance of BUMDES. However, there are many conflicts of interest in village administration.

2.3. Recruitment of Administrators

Human resources are key to creating a competitive organization, and this should be the basis for BUMDES to recruit BUMDES administrators amid strong political dynamics in the village [58]. In general, recruitment is the first step in human resource management and has a significant impact on organizational effectiveness. In the context of the digital economy and industry under the 4.0 standard, it should be recalled that, as Popkova [59] said, the quality of human capital is one of the key factors of organizational competitiveness. Therefore, recruitment is a much more strategic mechanism than the purchasing process of a company’s administrative functions. Effective recruitment matches candidates with the positions to be filled and the desired characteristics, not only with technical skills, but also other relevant factors, such as soft skills, values, and innovation capacity. The analysis by Larios-Francia and Ferasso [51] shows that the success of innovation, which in turn refers to competent and creative human resources, is a determinant of performance [60].

An effective recruitment process can produce human resources (HR) with the ability to innovate and adapt in response to a changing business environment. This aligns with the Resource-Based View (RBV) approach, which posits that long-term competitive advantage stems from an organization’s internal resources that are rare, difficult to imitate, and possess strategic value, including high-quality human resources [60]. Resource-Based View (RBV) and Knowledge-Based View (KBV) theories, as explained by Anjaningrum et al. [61], are key bases for understanding the linkages between recruitment, organizational learning, innovation, and performance. Recruitment is viewed not just as an effort to acquire manpower, but rather as a long-term strategic investment in organizational knowledge and capabilities.

Rodriguez et al. [62] confirmed that successful implementation of digital systems, such as ERP systems, and business model innovation is determined mainly by the organization’s ability to adapt and manage internal resistance. Proper recruitment plays a role in reducing resistance to change while accelerating the process of adaptation to new systems and procedures. This happens because individuals who are ready to learn and accept change tend to adapt more quickly to complex systems, such as ERP systems, which significantly affects organizational performance. The above explanation indicates that a good recruitment process for administrators influences their work motivation because they feel valued, fit in with the organization, and have room for growth, so it is hoped that they will be able to improve the management performance of BUMDES. Local elites are expected to properly supervise the recruitment process for administrators and refrain from negative intervention.

2.4. Training and Education

Training is a component of education that helps participants quickly enhance their skills. Good performance and results can only be achieved when workers are equipped with the necessary information and skills to reach their full potential [63]. The results of the study emphasize the importance of training and education both in and outside the workplace in improving their competitiveness and skills [64]. Separating training into its component elements reveals a favorable and statistically significant correlation between performance and on-the-job and off-the-job training. The connected effect, however, is greater than the sum of the two forms of training when applied simultaneously [65]. In the context of BUMDES management within a rapidly digitalizing rural economy, training becomes even more critical as it equips administrators with the adaptive skills necessary to engage with stakeholders and innovate. Knowledge about customers, from customers, and for customers can help BUMDES owners and managers [66]. Customer feedback may be a great source of innovative ideas for small and medium-sized enterprises (SMEs) by providing fresh perspectives, as well as more realistic and original ways of thinking about how to improve existing products and services [66].

Several studies reveal that entrepreneurs who receive training and education contribute more significantly to the development of positive self-evaluation, increased self-confidence, adaptability, competence, decision-making skills, and risk-taking [67]. Furthermore, internal motivation, including the drive for success and the willingness to take risks, is greatly influenced by learning experiences and education [68]. Therefore, it is indicated that through proper education and training, work motivation can be directly enhanced by strengthening feelings of competence and self-confidence, thereby improving the commitment and management performance of BUMDES. However, external motivation is often more emphasized in rural communities, as social status, recognition, and acknowledgment from local elites are considered more important.

2.5. Organizational Culture

Organizational culture and the work environment impact are critical determinants of employee performance, particularly in dynamic and innovation-driven organizations [20,69]. The imaginative work behavior of an organization’s personnel is one factor that determines its competitive advantage in culture. The process of innovative behavior in the workplace involves developing new ideas, tools, and procedures, which are subsequently tested and implemented within the firm’s operations. In the current era of unpredictable market conditions, changing consumer expectations, and increasing global competition, companies must innovate to remain viable [70]. To address their difficulties, companies must engage in innovative problem-solving. Findings indicate that open innovation and organizational creativity are key.

An organizational culture is a set of values, norms, attitudes, and behaviors that develop within an organization and serve as guidelines for interacting and conducting work. Rozak et al. [71] stated that an organizational work culture is defined as the extent to which organizational work values are applied by BUMDES managers, encompassing aspects such as collaboration, innovation, adaptability to change, and technology utilization. This work culture concept refers to the development of Shanker et al. [72], which includes innovative values, namely, support for creativity, independence at work, and exploration of ideas; innovative work behavior, which consists of the creation, promotion, and implementation of ideas; and organizational openness, namely, the level of organizational support for open discussion and acceptance of change.

Research by Kadam et al. [73] indicates that cultural competence significantly contributes to the performance of MSMEs and is an essential skill that owners or managers must possess, particularly when operating in a multicultural environment. The performance pyramid theory, proposed by Inthavong [74], illustrates that organizational performance consists of four hierarchical levels, starting from the company’s vision, business goals, operational indicators, and process quality and efficiency levels. This framework is relevant to understanding how BUMDES integrates village strategies into adaptive operational practices, particularly in terms of its ability to respond effectively to changes in the work environment. According to Shanker et al. [72], an organizational climate that supports innovation can encourage the emergence of innovative work behavior that directly contributes to improved performance. Rumanti et al. [75] stated that organizational performance, especially in the context of SMEs and village-level enterprises, is not solely judged by financial aspects, but also includes the achievement of strategic goals, the level of flexibility, and customer satisfaction. This view is in line with the findings of Inthavong et al. [74], which emphasize that organizational learning and the ability to innovate are essential elements in realizing sustainable performance. Meanwhile, a study by Rozak et al. [71] revealed that the implementation of a digital culture through digitalization strategies and strengthening digital skills has a positive impact on organizational agility, which ultimately improves the performance of MSMEs and, in this context, is also relevant for BUMDES. Within the framework of BUMDES, a work culture that emphasizes collaboration and innovation is a key factor in delivering optimal public services, effective professional business management, and adaptability to the dynamic needs of village communities [22].

One of the foundations used is Organizational Support Theory (OST), which states that organizational support motivates employees, thereby fostering commitment and loyalty to the organization. Based on previous theories and research findings, it can be indicated that, within the framework of BUMDES, an organizational culture that emphasizes collaboration and innovation is a key factor in providing optimal public services, effective professional business management, and the ability to adapt to the dynamic needs of rural communities. This will motivate BUMDES administrators, who are expected to improve performance.

2.6. Work Motivation

A person’s natural drive to achieve their goals is known as motivation [76]. Employee performance is positively and significantly influenced by their level of work motivation [77]. Self-determination theory, developed by Deci and Ryan [78], is one of the main foundations. This theory emphasizes the importance of intrinsic motivation, which comes from within the individual, in achieving goals and increasing job satisfaction. In this context, basic human needs such as autonomy, a sense of competence, and social relationships are important factors influencing employee motivation. The management of BUMDES is influenced by various factors, both internal and external, thus requiring high levels of work motivation. Motivation is divided into two categories: intrinsic motivation, which stems from individual drive, and extrinsic motivation, which comes from external sources such as rewards or social status recognition [79]. Gagné and Deci [48] state that this perspective is expanded by showing that when employees feel they have autonomy in their work, they tend to have higher commitment and perform better. This suggests that fulfilling psychological needs can enhance intrinsic motivation, which in turn has a positive impact on productivity and job satisfaction. The presence of digital tools and data-driven management approaches can enhance a sense of ownership, self-confidence, and work motivation among managers [80,81]. Several studies emphasize that when the digitalization process is integrated with strategic management practices such as human resource planning, positive local political engagement, and strengthening the role of village facilitators, the motivation of village officials increases significantly [82,83,84]. In the management of village-owned enterprises (BUMDES), the use of digital technology and the implementation of managerial innovations are not just trends but strategic necessities. Digital innovations, such as app-based accounting systems, local e-commerce platforms, and real-time financial dashboards, enable BUMDES administrators to work more efficiently, transparently, and responsively in response to community needs. However, the impact does not stop there. When managers are involved in the innovation process and given space to learn and experiment digitally, this creates a work environment that fosters active participation and a spirit of collaboration [85,86].

In the context of this study, the authors indicates that local politics (external), village assistants (external), recruitment of managers (internal), educational training (internal), and organizational culture (internal) influence the work motivation of BUMDES managers, which in turn affects their performance.

2.7. Management Performance

Effective BUMDES managers are able to increase village revenue, support economic improvement of the community, provide services to the community, and ensure sustainability [28]. Performance refers to the efforts made by employees in carrying out their duties in accordance with the quality and quantity standards set by their superiors [87]. The established objectives and targets of BUMDES are a shared responsibility. Employee performance can be interpreted as the level of success in achieving targets and other tasks [88]. One of the factors that significantly influences employee performance is compensation, particularly in the form of a salary, which is one way to motivate employees, considering that the salaries of BUMDES managers are highly dependent on the profits of the BUMDES themselves [89]. Furthermore, the overall performance of an institution is greatly influenced by the performance of its individual employees, both positively and negatively [90]. Therefore, the recruitment, education, and training processes for BUMDES managers must be given attention and guidance. Training and utilization of digital transformation integrated with management strategies not only increase market access, productivity, and revenue but also act as a catalyst in shaping an adaptive and dynamic organizational culture [91,92].

Organizational performance encompasses not only the final results (output) but also the implementation process. BUMDES management involves numerous stakeholders, which creates a potential for conflicts of interest. Further research using longitudinal data shows a positive reciprocal relationship between the implementation of high-performance work systems (HPWSs) and increased productivity [93]. Concerning BUMDES, the use of technology, the internet, and social media is expected to increase BUMDES’ productivity. Based on the previous explanation, it is evident that factors such as local politics, village facilitators, recruitment of administrators, education and training, and organizational culture can influence the management performance of BUMDES.

2.8. Synthesis of Variable Justification and Research Hypothesis

The construct selection in this research is grounded in a theoretical rationale linking performance and latent variables and is reinforced by empirical findings from prior research integrated into the theoretical model. In the context of digital economic transformation and sustainable rural development, the selection of variables in this study is not only based on theory but also on the practical urgency of developing village institutional capacity and sustainability. Local politics influences the direction of policy and strategic decision-making in the utilization of technology and digital resources by the BUMDES. Village facilitators play a crucial role in the digitalization process, particularly in capacity development and innovation facilitation. The recruitment process for village officials, based on transparency and digital competence, is a key determinant of technological adaptation success within village organizations. Education and training are essential to equip officials with adequate managerial and digital skills. An organizational culture open to innovation accelerates the technology adaptation process and strengthens the sustainability of village institutions. Work motivation is a key element bridging the influence of structural factors on performance. Thus, all variables are structured to reflect how internal and external factors interact in driving the performance of BUMDES managers based on technology, participation, and sustainability. The elaboration of latent variables and their indicators can be seen in Section 3.2.2. Next, these variables are formulated into hypotheses.

Hypotheses are formed not only through theoretical reasoning but also as a means of connecting abstract concepts with empirical data that can be measured. Therefore, this research formulates hypotheses systematically by referring to previous studies, relevant theoretical foundations, and existing study gaps, to examine the cause-and-effect relationships contained in a structural model designed to identify key factors in the development of BUMDES as a strategy for sustainable development in rural areas. The following are the hypotheses in this study based on the theoretical framework model presented in Figure 1:

H1: 

The latent variable (LV) of local politics (PL) influences the LV of work motivation (MT).

H2: 

The LV of village facilitators (PD) influences the LV of work motivation (MT).

H3: 

The LV of recruitment of administrators (RP) influences the LV of work motivation (MT).

H4: 

The LV of training and education (DL) influences the LV of work motivation (MT).

H5: 

The LV of organizational culture (BU) influences the LV of work motivation (MT).

H6: 

The LV of local politics (PL) influences the LV of management performance (KN).

H7: 

The LV of village facilitators (PD) influences the LV of management performance (KN).

H8: 

The LV of recruitment of administrators (RP) influences the LV of management performance (KN).

H9: 

The LV of training and education (DL) influences the LV of management performance (KN).

H10: 

The LV of organizational culture (BU) influences the LV of management performance (KN).

H11: 

The LV of local politics (PL) influences the LV of management performance (KN) through the LV of work motivation (MT).

H12: 

The LV of village facilitators (PD) influences the LV of management performance (KN) through the LV of work motivation (MT).

H13: 

The LV of recruitment of administrators (RP) influences the LV of management performance (KN) through the LV of work motivation (MT).

H14: 

The LV of training and education (DL) influences the LV of management performance (KN) through the LV of work motivation (MT).

H15: 

The LV of organizational culture (BU) influences the LV of management performance (KN) through the LV of work motivation (MT).

3. Materials and Methods

3.1. Construction of Conceptual Framework

In behavioral studies, both independent and dependent research variables are defined to help systematically design and structure research methods [94]. Dependent variables are used as measurement tools to assess behavioral responses to changes in one or more independent variables. To clarify the understanding of the variables used in this study, we categorize them into two types: independent variables and dependent variables. The relationship between the independent variable and the behavior of the dependent variable is interrelated. We focused our analysis on the characteristics that reflect the independent variables to gain a deeper understanding of their relationship with the dependent variables, specifically work motivation and board performance. We hypothesized that the five independent variables depicted in Figure 1 are factors that influence the work motivation and management performance of BUMDES in Bola District. Figure 1 shows the conceptual framework for factors influencing work motivation and management performance.

The research methodology employed in this study consisted of four steps, as illustrated in Figure 2. These steps were (1) model development, research location determination, and primary data collection; (2) measurement model evaluation; (3) SEM evaluation/assessment of SEM; and (4) hypothesis model testing and interpretation. A description of the activities carried out in each stage is outlined below.

Figure 2 shows the steps of the research and design process. The research process consisted of four main steps, as shown in Figure 2. First, a structural equation model (SEM) was developed, and then the research site was visited. After that, primary data were gathered and analyzed. Then, the measurement model was evaluated. Next, we assessed the SEM model and tested the hypothesis. Finally, the results were interpreted, and conclusions or recommendations were drawn.

3.2. SEM Model Development, Research Site, Primary Data Collection, and Data Analysis

3.2.1. SEM Model Development

The initial stage of this research involved building data for the development of structural equation models. Figure 2, a flowchart for this research, illustrates the stages involved in conducting this study. In the process, we designed a model that was assumed to influence work motivation (MT) and management performance (KN). Based on the literature review, we identified five latent variables that are expected to affect work motivation and management performance, namely, local politics (PL), village facilitators (PD), training and education (DL), recruitment of administrators (RP), and organizational culture (BU). The five latent variables were then connected in the form of a path diagram, with the direction of influence referring to the results of the previous literature review (Figure 1).

SEM-Lisrel analysis was used to test the hypotheses. One statistical method that examines dependent and interdependent multivariate data is structural equation modeling (SEM), which focuses on confirmatory factor analysis and path analysis. Here we examined constructs, which are latent variables. Instead of generating models, SEM mainly aims to assess and verify theory-based models, particularly measurement and structural models [95]. Intelligent statistical software, LISREL8.80, can execute a wide range of analyses, including path analysis and structural equation modeling (SEM) [96,97].

3.2.2. Research Site and Primary Data Collection

This study was conducted in Bola District, Wajo Regency, South Sulawesi Province, Indonesia. Primary data were collected through structured interviews with 250 respondents. The respondents’ answers to the questionnaire were the primary source of data in this study. The sample consisted of village heads, Village Consultative Bodies (BPD), village officials, administrators of village-owned enterprises (BUMDES), and community and private sector leaders. The criterion for inclusion was that respondents had an understanding of BUMDES. The demographic profile of the respondents showed that 60.4% were male and 39.6% were female. In terms of education, 1.2% held a master’s degree (S2), 35.6% a bachelor’s degree (S1), 61.6% a senior high school (SMA) diploma, 0.4% a junior high school (SMP) diploma, and 1.2% an elementary school (SD) diploma. In the field, 1.2% of elementary school graduates and 0.4% of junior high school graduates were members of the BUMDES management. The Likert scale was used to measure various dimensions in this study, where 1 indicated strong disagreement and 5 indicated strong agreement [98]. An explanation of the data sample description, presented in the form of descriptive statistics, is outlined in

Table 1. Descriptive statistics of the research data.

Latent Variables (LV)	Indicator Variables	Number of Respondents	Likert Scale Frequencies					Means	Min. Value	Max. Value	Std. Deviation
			1	2	3	4	5
LV of Local Politics (PL)	Local Leader Intervention (PL1)	250	0	44	81	86	39	3.48	2	5	0.958
	Kinship Relationship (PL2)	250	0	21	117	71	41	3.53	2	5	0.865
	BUMDES Development Priority (PL3)	250	0	36	67	112	35	3.58	2	5	0.902
	Political Elite Intervention (PL4)	250	0	20	72	128	30	3.67	2	5	0.789
LV of Village Facilitator (PD)	Facilitator (PD1)	250	0	13	63	137	37	3.79	2	5	0.753
	Motivator (PD2)	250	0	11	73	125	41	3.78	2	5	0.767
	Supervisor (PD3)	250	0	18	66	124	42	3.76	2	5	0.816
	Communicator (PD4)	250	0	17	57	129	47	3.82	2	5	0.812
LV of Recruitment of Administrators (RP)	Human Resource Needs (RP1)	250	0	70	64	71	45	3.36	2	5	1.075
	Source of Human Resources (RP2)	250	0	71	77	59	43	3.30	2	5	1.061
	Open Recruitment (RP3)	250	0	67	81	56	46	3.32	2	5	1.062
	Transparency (RP4)	250	0	39	53	93	65	3.74	2	5	1.015
LV of Training and Education (DL)	Knowledge (DL1)	250	0	33	65	112	40	3.64	2	5	0.905
	Skills (DL2)	250	0	35	75	84	56	3.64	2	5	0.980
	Attitude (DL3)	250	0	27	77	90	56	3.70	2	5	0.937
	Cooperative Relationship (DL4)	250	0	39	56	107	48	3.66	2	5	0.962
LV of Organizational Culture (BU)	Risk Tolerance (BU1)	250	0	23	77	107	43	3.68	2	5	0.865
	Supervision (BU2)	250	0	24	87	87	52	3.67	2	5	0.913
	Result-Oriented (BU3)	250	0	20	83	87	60	3.75	2	5	0.912
	Reward (BU4)	250	0	20	79	92	59	3.76	2	5	0.904
LV of Work Motivation (MT)	Willingness to Work (MT1)	250	0	21	62	112	55	3.80	2	5	0.877
	Desire for Achievement (MT2)	250	0	12	72	103	63	3.87	2	5	0.847
	Perseverance at Work (MT3)	250	0	11	60	125	54	3.89	2	5	0.789
	Forward-Oriented (MT4)	250	0	12	56	132	50	3.88	2	5	0.777
LV of Management Performance (KN)	BUMDES Income (KN1)	250	0	12	64	118	56	3.87	2	5	0.811
	Asset Addition (KN2)	250	0	19	44	127	60	3.91	2	5	0.846
	Increase in Management Salary (KN3)	250	0	15	44	142	49	3.90	2	5	0.777
	Improving the Quality of Human Resources of the Management (KN4)	250	0	9	42	140	59	4.00	2	5	0.742

.

In general, Table 1 presents descriptive statistics indicating that the average perception of respondents toward the variables fell within the high category (mean: 3.30–4.00, on a scale of 1–5). The highest value was found for indicator KN4 with a mean of 4.00, while the lowest was for indicator RP2 with a value of 3.30. The standard deviation ranged from 0.742 to 1.075, indicating moderate variation in responses. This reflected that, overall, respondents’ perceptions are relatively good, particularly regarding work motivation and managerial performance. However, variables related to the recruitment of administrators reveal moderate perceptions, necessitating attention to policy implementation. The variables in this study were measured using a structured Likert scale. Although these are often used to measure perceptions and attitudes, the results are sometimes biased toward social desirability and personal interpretation. This is considered a data limitation. Variables related to the recruitment of administrators and local politics are highly sensitive, so some respondents expressed concerns about potential social or political risks. However, the researchers endeavored to maintain the confidentiality of the respondents’ answers to the questionnaire. Generalization of the findings of this study should be carried out carefully, taking into account the specific local context.

The sample size met the criteria for SEM-LISREL analysis, allowing for the accurate analysis of complex covariance structures with SEM-LISREL, which can identify correlations between complicated variables. This method examines how factors such as local politics (PL), village facilitators (PD), recruitment of administrators (RP), training and education (DL), and organizational culture (OC) influence work motivation (WM) and management performance (MP). Latent variables cannot be measured directly, so the authors used measurable indicators (observed variables) to estimate their values, which were obtained from previous studies. Latent and indicator variables are shown in

Table 2. Latent variables and measurement indicators.

No.	Latent Variables	Indicator Variables and Symbols	Term Error	No.	Latent Variables	Indicator Variables and Symbols	Term Error
1	Management Performance (KN)	BUMDES Income (KN1)	ε1	4	Village Facilitator (PD)	Facilitator (PD1)	δ5
		Asset Addition (KN2)	ε2			Motivator (PD2)	δ6
		Increase in Management Salary (KN3)	ε3			Supervisor (PD3)	δ7
		Improving the Quality of Human Resources of the Management (KN4)	ε4			Communicator (PD4)	δ8
2	Work Motivation (MT)	Willingness to Work (MT1)	ε5	5	Recruitment of Administrators (RP)	Human Resource Needs (RP1)	δ9
		Desire for Achievement (MT2)	ε6			Source of Human Resources (RP2)	δ10
		Perseverance at Work (MT3)	ε7			Open Recruitment (RP3)	δ11
		Forward-Oriented (MT4)	ε8			Transparency (RP4)	δ12
3	Local Politics (PL)	Local Leader Intervention (PL1)	δ1	6	Training and education (DL)	Knowledge (DL1)	δ13
		Kinship Relationship (PL2)	δ2			Skills (DL2)	δ14
		BUMDES Development Priority (PL3)	δ3			Attitude (DL3)	δ15
		Political Elite Intervention (Pl4)	δ4			Cooperative Relationship (DL4)	δ16
7	Organizational Culture (BU)	Risk Tolerance (BU1)	δ17	Result-Oriented (BU3)			δ19
		Supervision (BU2)	δ18	Reward (BU4)			δ20

and

Table 3. Literature review of variables and their measuring units.

No.	Latent Variables	Measurable Variables		Unit of Measurement
		Indicator Variables and Symbols	Source	Base Data	Scale Range *	Data Entered **
1	Management Performance (KN)	The BUMDES Revenue (KN1)	[12,13]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Asset Addition (KN2)	[99]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		The BUMDES Management Income (KN3)	[99,100]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Improvement of Human Resource Quality of the Management (KN4)	[101]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
2	Work Motivation (MT)	Willingness to Work (MT1)	[77]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Desire for Achievement (MT2)	[77]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Diligence to Work (MT3)	[102]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Forward-Oriented (MT4)	[103,104]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
3	Local Politics (PL)	Local Leader Intervention (PL1)	[105]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Kinship Relationship (PL2)	[106]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Prioritization of BUMDES Development (PL3)	[105]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Political Elite Intervention (PL4)	[107]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
4	Village Facilitator (PD)	Facilitator (PD1)	[17,99]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Motivator (PD2)	[99,108]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Supervisor (PD3)	[52,99]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Communicator (PD4)	[99,109]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
5	Recruitment of Administrators (RP)	Human Resource Needs (RP1)	[110,111]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Source of Human Resource (RP2)	[110,111]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Open recruitment (RP3)	[110,111]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Transparency (RP4)	[110,111]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
6	Training and Education (DL)	Knowledge (DL1)	[87,112]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Skill (DL2)	[113]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Attitude (DL3)	[87,112]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Cooperative Relationship (DL4)	[113]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
7	Organizational Culture (BU)	Risk Tolerance (BU1)	[69,114]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Supervision (BU2)	[115,116]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Result-Oriented (BU3)	[69,104]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”
		Reward (BU4)	[69,104]	Likert Scale	5 PLS	“SD = 1, D = 2, N = 3, A = 4, SA = 5”

Description: * 5 PLS = 5-Point Likert Scale. ** SD = Strongly Disagree (1), D = Disagree (2), N = Neutral (3), A = Agree (4), SA = Strongly Agree (5).

.

There is some theoretical similarity between LISREL and AMOS. Both are based on covariance. Before testing the model, all data instruments were run through CFA (Confirmatory Factor Analysis) to ensure their authenticity [117]. One should consider whether the data are standard by running them through a multivariate or univariate test. If the skewness and kurtosis p-values are more significant than 0.05, then the data can be considered to follow a normal distribution [118]. Verifying reliability is the next step. The variance extracted (VE) and the reliability of the construction process are to be examined in particular. Scores of at least >0.7 for construct reliability [119] and 50% for the variance extracted are considered cutoff values [120]. Furthermore, this test measures GoF (goodness of fit) to evaluate the overall quality of model fit. Nine indicators from the GoF summary were utilized in this study. According to Heir, four or five goodness-of-fit requirements must be achieved for a model to fit [121,122]. You can see the results of nine GoF tests in

Table 4. Goodness of fit.

Goodness of Fit	Cutoff Value	Goodness of Fit	Cutoff Value
Chi-Square (χ2)	Small	CMIN/DF	≤2.00
Significance	≥0.05	CFI	≥0.95
RMSEA	≤0.08	IFI	≥0.90
GFI	≥0.90	NFI	≥0.90
AGFI	≥0.90	RMSR	≤0.08

.

The t-statistic and probability values can be used to assess the results of the hypothesis testing process once the goodness of fit (GoF) has been satisfied. According to the hypothesis testing criteria, at a significance level of 5% (α: 0.05), the t-statistic must be greater than 1.96. Only then can the hypothesis be accepted or rejected [123]. As an illustration, according to research by Putra [124] published in the South Asian Journal of Social Studies and Economics, a t-statistic value of ≥1.96 indicates significance at the 5% level (α = 0.05), meaning that the relationship in the model is statistically significant. This statement aligns with the findings of Suryani et al. [125], who used the Partial Least Squares SEM approach and emphasized that a t-statistic value above 1.96 is the primary indicator supporting hypothesis acceptance. Examining the relationship between variables in-depth within a theoretical framework is a fundamental element in building a solid empirical research base. In the face of increasingly complex dynamics related to organizations, social behavior, and development, the preparation and testing of hypotheses is an essential stage in evaluating the validity of conceptual models and making a real contribution to the development of science and policy practice.

4. Results

4.1. Data Normality

One uses the normality test to determine if the data used in research are normally distributed. The most crucial part of data description in structural equation modeling is ensuring that the assumptions of multivariate and univariate data normality are satisfied.

Table 5. Univariate normality test.

No.	Indicator Variables	Skewness		Kurtosis		Skewness and Kurtosis
		Z-Score	p-Value	Z-Score	p-Value	Chi-Square	p-Value
1	Local Leader Intervention (PL1)	0.098	0.922	−4.220	0.000	17.818	0.142
2	Kinship Relationship (PL2)	0.153	0.878	−1.852	0.064	3.454	0.178
3	BUMDES Development Priority (PL3)	−0.302	0.763	−2.586	0.010	6.778	0.034
4	Political Elite Intervention (PL4)	−0.672	0.502	−0.690	0.490	0.927	0.629
5	Facilitator (PD1)	−0.891	0.373	−0.298	0.765	0.883	0.643
6	Motivator (PD2)	−0.703	0.482	−0.963	0.336	1.421	0.491
7	Supervisor (PD3)	−0.831	0.406	−1.318	0.188	2.428	0.297
8	Communicator (PD4)	−1.028	0.304	−1.241	0.215	2.598	0.273
9	Human Resource Needs (RP1)	0.806	0.420	−8.853	0.000	79.017	0.102
10	Source of Human Resources (RP2)	0.974	0.330	−8.329	0.000	70.327	0.107
11	Open Recruitment (RP3)	0.761	0.446	−8.303	0.000	69.514	0.110
12	Transparency (RP4)	−0.940	0.347	−6.226	0.000	39.652	0.132
13	Knowledge (DL1)	−0.467	0.640	−2.702	0.007	7.522	0.123
14	Skills (DL2)	−0.625	0.532	−5.013	0.000	25.517	0.221
15	Attitude (DL3)	−0.794	0.427	−3.951	0.000	16.245	0.314
16	Cooperative Relationship (DL4)	−0.489	0.625	−4.049	0.000	16.635	0.305
17	Risk Tolerance (BU1)	−0.592	0.554	−2.280	0.023	5.551	0.362
18	Supervision (BU2)	−0.608	0.543	−3.349	0.001	11.587	0.103
19	Result-Oriented (BU3)	−0.928	0.354	−3.590	0.000	13.750	0.101
20	Reward (BU4)	−0.957	0.338	−3.444	0.001	12.777	0.102
21	Willingness to Work (MT1)	−1.062	0.288	−2.655	0.008	8.177	0.117
22	Desire for Achievement (MT2)	−1.185	0.236	−2.845	0.004	9.499	0.109
23	Perseverance at Work (MT3)	−1.136	0.256	−1.458	0.145	3.417	0.181
24	Forward-Oriented (MT4)	−1.118	0.264	−0.976	0.329	2.201	0.333
25	BUMDES Income (KN1)	−1.130	0.259	−1.912	0.056	4.932	0.085
26	Asset Addition (KN2)	−1.355	0.176	−1.931	0.053	5.565	0.062
27	Increase in Management Salary (KN3)	−1.186	0.236	−0.473	0.636	1.629	0.443
28	Improving the Quality of Human Resources of the Management (KN4)	−1.391	0.164	−0.619	0.536	2.318	0.314

presents the results of the data normality test conducted on the screening results of the 250 respondents.

The outcomes of the one-way test for normality are displayed in Table 5. Indicators of data normality, including the p-value, chi-square, skewness, and kurtosis, are presented in the table. The preceding indicators suggest that the p-value for chi-square skewness and kurtosis is more prominent than 0.05. For indicators to be considered univariate normal, their p-values for chi-square skewness and kurtosis must be greater than 0.05 [118]. In addition, SEM analysis also presumes multivariate normality through the use of maximum likelihood (ML) estimation. The results of the multivariate normality test are presented in

Table 6. Multivariate normality test.

Skewness			Kurtosis			Skewness and Kurtosis
Value	Z-Score	p-Value	Value	Z-Score	p-Value	Chi-Square	p-Value
195.721	35.385	0.000	902.03	9.137	0.000	1335.558	0.000

.

The results of the multivariate normality test, as shown in Table 6, were 1335.558 for chi-square skewness and 0.000 for kurtosis. It can be concluded that the data reveal multivariate abnormality, as the p-value of 0.000 is smaller than 0.05. Despite this, many researchers use SEM analysis even when the data are not perfectly normal. If this occurs, the researcher employs an assumption, which entails adjusting the standard error and a portion of the goodness of fit due to data distribution anomalies, as stated in the study results by Ghozali and Fuad [126]. To mitigate its effects, researchers assess the model to ensure that it satisfies four or five goodness-of-fit criteria specific to the study. Data anomalies have a lesser impact on CFI (Comparative Fit Index) and RMSR (root mean square residual), two of the criteria mentioned.

4.2. Data of Confirmatory Factor Analysis (CFA)/Factor Loading

The first step in testing SEM before running the full model test is to use the confirmatory factor analysis (CFA) model. Using CFA, one can determine if the indicators accurately measure the latent variable.

Table 7. Value loading factors of CFA indicators.

No.	Indicators	Local Politics (PL)	Village Facilitator (PD)	Recruitment of Management (RP)	Education Training (DL)	Organizational Culture (BU)	Work Motivation (MT)	Management Performance (KN)
1	Local Leader Intervention (PL1)	0.69	—	—	—	—	—	—
2	Kinship Relationship (PL2)	0.82	—	—	—	—	—	—
3	BUMDES Development Priority (PL3)	0.75	—	—	—	—	—	—
4	Political Elite Intervention (PL4)	0.74	—	—	—	—	—	—
5	Facilitator (PD1)	—	0.64	—	—	—	—	—
6	Motivator (PD2)	—	0.71	—	—	—	—	—
7	Supervisor (PD3)	—	0.74	—	—	—	—	—
8	Communicator (PD4)	—	0.65	—	—	—	—	—
9	Human Resource Needs (RP1)	—	—	0.72	—	—	—	—
10	Source of Human Resources (RP2)	—	—	0.80	—	—	—	—
11	Open Recruitment (RP3)	—	—	0.80	—	—	—	—
12	Transparency (RP4)	—	—	0.73	—	—	—	—
13	Knowledge (DL1)	—	—	—	0.74	—	—	—
14	Skills (DL2)	—	—	—	0.74	—	—	—
15	Attitude (DL3)	—	—	—	0.83	—	—	—
16	Cooperative Relationship (DL4)	—	—	—	0.67	—	—	—
17	Risk Tolerance (BU1)	—	—	—	—	0.70	—	—
18	Supervision (BU2)	—	—	—	—	0.70	—	—
19	Result-Oriented (BU3)	—	—	—	—	0.77	—	—
20	Reward (BU4)	—	—	—	—	0.80	—	—
21	Willingness to Work (MT1)	—	—	—	—	—	0.70	—
22	Desire for Achievement (MT2)	—	—	—	—	—	0.78	—
23	Perseverance at Work (MT3)	—	—	—	—	—	0.81	—
24	Forward-Oriented (MT4)	—	—	—	—	—	0.84	—
25	BUMDES Income (KN1)	—	—	—	—	—	—	0.73
26	Asset Addition (KN2)	—	—	—	—	—	—	0.74
27	Increase in Management Salary (KN3)	—	—	—	—	—	—	0.74
28	Improving the Quality of Human Resources of the Management (KN4)	—	—	—	—	—	—	0.73
Chi-square = 743.23, df = 329, p-value = 0.0000, RMSEA = 0.071

displays the comprehensive results of the model CFA tests, which revealed a p-value of 0.0000, a degree of freedom (DF) value of 329, and a chi-square value of 743.23. The cmin/df value has a chi-square value of 2.259 (743.23/329). This number is significantly higher than the threshold of 2. The obtained RMSEA value is 0.071. As a result, this value is less than or equal to 0.08. Table 7 also displays the standardized factor loading value for each indicator concerning its corresponding hidden variable.

For the latent variable (LV) of management performance (KN), two indicators, the BUMDES income (KN₁) indicator, with a factor loading value of 0.73, and the indicator for improving the quality of human resources of the management (KN₄) had the lowest factor loading values. On the other hand, two indicators, the asset addition (KN₂) indicator, with a factor loading value of 0.74, and the increase in management salary (KN₃) indicator, with a factor loading value of 0.74, were the highest factor loading values for the latent variable management performance (KN). The indicator with the lowest factor loading value for LV work motivation (MT) was willingness to work (MT1), with a value of 0.70. The indicator with the greatest factor loading value was forward-oriented (MT₄), with a value of 0.84.

The local leader intervention (PL₁) indicator has the lowest factor loading value of 0.69 for the LV local politics (PL). In contrast, the kinship relationship (PL₂) indicator has the most significant factor loading value of 0.82. The facilitator indicator (PD1) has the lowest factor loading value of 0.64 for the LV village facilitator (PD), while the supervisor indicator (PD₃) has the most significant factor loading value of 0.74. There is one indicator for the LV recruitment of administrators (RP) with the lowest factor loading value at 0.72. Two indicators with the highest loading values, at 0.80, are the source of the human resource indicator (RP₂) and open recruitment (RP₃). The cooperative relationship indicator (DL₄) has the lowest factor loading value of 0.67 for the LV training and education (DL). In contrast, the attitude indicator (DL₃) has the highest factor loading value of 0.83. The risk tolerance indicator (BU₁) and the supervision indicator (BU₂) both have loading factor values of 0.70 for the LV organizational culture (BU), whereas the reward indicator (BU₄) has the highest value at 0.80. Factor loading values greater than 0.5 are considered valid in theory [127]. All of the aforementioned indicators of latent variables are legitimate indicators of their respective latent variables, as shown in Table 7.

Since the chi-square value remained high in the initial CFA test, the model was adjusted to achieve a value that met the criterion for goodness of fit. To make this adjustment, we examined the model’s output for information about the covariance errors of the indicators and then correlated them. The relationship between the covariance errors is evident in

Table 8. Modification indices.

Between	And	Decrease in Chi-Square	New Estimate
Prioritization of BUMDES Development (PL3)	Local Leader Intervention (PL1)	13.20	−0.12
Prioritization of BUMDES Development (PL3)	Kinship Relationship (PL2)	31.20	0.16
Political Elite Intervention (PL4)	Kinship Relationship (PL2)	16.40	−0.10
Source of Human Resource (RP2)	Motivator (PD2)	9.20	−0.08
Open Recruitment (RP3)	Source of Human Resource (RP2)	8.20	0.11
Transparency (RP4)	Kinship Relationship (PL2)	8.00	−0.08
Risk Tolerance (BU1)	Open Recruitment (RP3)	12.30	0.10
Supervision (BU2)	Transparency (RP4)	8.20	−0.09
Result-Oriented (BU3)	Transparency (RP4)	11.40	0.10
Result-Oriented (BU3)	Supervision (BU2)	12.50	0.11
Reward (BU4)	Knowledge (DL1)	9.70	−0.08
Willingness to Work (MT1)	Communicator (PD4)	8.00	−0.08
Willingness to Work (MT1)	Supervision (BU2)	20.60	−0.12
Willingness to Work (MT1)	Reward (BU4)	9.70	0.08
Desire for Achievement (MT2)	Knowledge (DL1)	10.40	−0.08
Diligence to Work (MT3)	Desire for Achievement (MT2)	26.60	0.11
BUMDES Income (KN1)	Desire for Achievement (MT2)	10.30	−0.07
BUMDES Income (KN1)	Forward-Oriented (MT4)	18.20	0.08
Asset Addition (KN2)	Forward-Oriented (MT4)	13.00	0.07
Improvement in Human Resource Quality of the Management (KN4)	Communicator (PD4)	8.20	0.06
Chi-square = 532.33

.

Twenty correlations between the covariance errors in the updated model can be observed in Table 8. Since the chi-square value drops to 532.33, which falls into the tiny category, due to the correlation between PL₃ and PL₂ errors and MT₃ and MT₂ errors, it can be considered that the model is a good fit.

4.3. Measurement Model Evaluation

The second phase of the research design involved evaluating the measurement model. Convergent validity (CV), composite reliability (CR), Cronbach’s alpha (CA), and average variance extraction (AVE) tests were conducted in this step. In the concurrent validity test, the following requirement must be met: the loading factor must be significant (loading factor > 0.50) [128]. The indicator values were all above 0.5, indicating valid factor loadings [94]. The indicators mentioned above are valid measures of latent variables, according to these results. In addition,

Table 9. Correlations between latent variables based on correlation values, error values, and t-counts.

Latent Variables	Symbols	PL	PD	RP	DL	BU	MT
Local Politics	PL	1.00
Village Facilitator	PD	0.83	1.00
		(0.04)
		21.80
Recruitment of Administrators	RP	0.60	0.75	1.00
		(0.05)	(0.04)
		11.30	17.25
Training and Education	DL	0.76	0.78	0.83	1.00
		(0.04)	(0.04)	(0.03)
		18.31	18.64	25.04
Organizational Culture	BU	0.72	0.73	0.79	0.73	1.00
		(0.05)	(0.05)	(0.04)	(0.04)
		16.00	15.56	21.21	16.58
Work Motivation	MT	0.75	0.77	0.73	0.78	0.78	1.00
		(0.04)	(0.04)	(0.04)	(0.04)	(0.04)
		18.23	18.16	17.61	20.09	19.70
Management Performance	KN	0.75	0.85	0.76	0.83	0.79	0.92
		(0.04)	(0.04)	(0.04)	(0.03)	(0.04)	(0.02)
		17.51	23.68	19.23	24.01	19.87	38.23
Description:		Correlation Value			Error		t-Count

displays the relationships (correlations) between the latent variables in the entire CFA model.

Table 9 presents the outcomes of the correlation value, error value, and t-count assessments of the relationships between the latent variables. A score above 1.96 indicates a statistically significant link between the latent variables. Additionally,

Table 10. Correlations between latent variables.

No.	Correlation Relationships	Symbols	Values	Errors	t-Count	Description
1	Local Politics and Village Facilitator	PL and PD	0.83	−0.04	21.80	Significant
2	Local Politics and Recruitment of Administrators	PL and RP	0.60	−0.05	11.30	Significant
3	Local Politics and Training and Education	PL and DL	0.76	−0.04	18.31	Significant
4	Local Politics and Organizational Culture	PL and BU	0.72	−0.05	16.00	Significant
5	Local Politics and Work Motivation	PL and MT	0.75	−0.04	18.23	Significant
6	Local Politics and Management Performance	PL and KN	0.75	−0.04	17.51	Significant
7	Village Facilitator and Recruitment of Administrators	PD and RP	0.75	−0.04	17.25	Significant
8	Village Facilitator and Training and Education	PD and DL	0.78	−0.04	18.64	Significant
9	Village Facilitator and Organizational Culture	PD and BU	0.73	−0.05	15.56	Significant
10	Village Facilitator and Work Motivation	PD and MT	0.77	−0.04	18.16	Significant
11	Village Facilitator and Management Performance	PD and KN	0.85	−0.04	23.68	Significant
12	Recruitment of Administrators and Training and Education	RP and DL	0.83	−0.03	25.04	Significant
13	Recruitment of Administrators and Organizational Culture	RP and BU	0.79	−0.04	21.21	Significant
14	Recruitment of Administrators and Work Motivation	RP and MT	0.73	−0.04	17.61	Significant
15	Recruitment of Administrators and Management Performance	RP and KN	0.76	−0.04	19.23	Significant
16	Training and Education and Organizational Culture	DL and BU	0.73	−0.04	16.58	Significant
17	Training and Education, and Work Motivation	DL and MT	0.78	−0.04	20.09	Significant
18	Training and Education, and Management Performance	DL and KN	0.83	−0.03	24.01	Significant
19	Organizational Culture and Work Motivation	BU and MT	0.78	−0.04	19.70	Significant
20	Organizational Culture and Management Performance	BU and KN	0.79	−0.04	19.87	Significant
21	Work Motivation and Management Performance	MT and KN	0.92	−0.02	38.23	Significant

includes the analysis related to the results presented in Table 9. Table 10 shows that there were no signs of multicollinearity relationships, except for the latent variables (LVs) of work motivation (MT) and management performance (KN). In Table 10, a high correlation (r = 0.92) was found between the LV of work motivation (MT) and the LV of management performance (KN). This figure addresses the concerns raised about discriminant validity and the potential for multicollinearity. When interpreting the correlation between the two constructs, it would be beneficial to consider potential issues of construct overlap. However, based on a strong theoretical foundation, the relationship between work motivation and management performance is indeed predicted to be closely related. Therefore, the high correlation observed in this model reflects consistency between the empirical findings and the theoretical foundations. Furthermore, the Average Variance Extracted (AVE) values were also evaluated for each construct, and the results were adequate. Modification indices and error covariances were also examined to ensure that there were no indications of multicollinearity that could disrupt model stability.

4.4. Variance Extracted (VE) and Construct Reliability

The construct reliability test quantifies how well each indicator represents a shared form variable by measuring its internal consistency. Two types of reliability testing exist: variance extracted (VE) and Cronbach’s alpha, also known as construct reliability. The construct reliability and variance extracted cutoff values were 0.7 and 0.5, respectively.

Table 11. Construct reliability and variance extracted.

Latent Variables	Indicators	Symbols	SLF *	Errors	SLF2	Construct Reliability	Variance Extracted
Local Politics (PL)	Local Leader Intervention	PL1	0.66	0.48	0.4356	0.835687	0.507621
	Kinship Relationship	PL2	0.71	0.25	0.5041
	BUMDES Development Priority	PL3	0.68	0.35	0.4624
	Political Elite Intervention	PL4	0.58	0.28	0.3364
Village Facilitator (PD)	Facilitator	PD1	0.48	0.33	0.2304	0.778897	0.469907
	Motivator	PD2	0.54	0.29	0.2916
	Supervisor	PD3	0.60	0.30	0.3600
	Communicator	PD4	0.52	0.38	0.2704
Recruitment of Administrators (RP)	Human Resource Needs	RP1	0.78	0.55	0.6084	0.849284	0.585679
	Source of Human Resources	RP2	0.85	0.40	0.7225
	Open Recruitment	RP3	0.85	0.41	0.7225
	Transparency	RP4	0.74	0.48	0.5476
Training and Education (DL)	Knowledge	DL1	0.67	0.37	0.4489	0.82963	0.550254
	Skills	DL2	0.72	0.44	0.5184
	Attitude	DL3	0.77	0.29	0.5929
	Cooperative Relationship	DL4	0.64	0.51	0.4096
Organizational Culture (BU)	Risk Tolerance	BU1	0.60	0.38	0.3600	0.841817	0.572027
	Supervision	BU2	0.69	0.36	0.4761
	Result-Oriented	BU3	0.70	0.34	0.4900
	Reward	BU4	0.72	0.30	0.5184
Work Motivation (MT)	Willingness to Work	MT1	0.61	0.40	0.3721	0.85852	0.602912
	Desire for Achievement	MT2	0.66	0.28	0.4356
	Perseverance at Work	MT3	0.64	0.22	0.4096
	Forward-Oriented	MT4	0.65	0.18	0.4225
Management Performance (KN)	BUMDES Income	KN1	0.59	0.31	0.3481	0.826431	0.544193
	Asset Addition	KN2	0.63	0.32	0.3969
	Increase in Management Salary	KN3	0.58	0.27	0.3364
	Improving the Quality of Human Resources of the Management	KN4	0.54	0.25	0.2916

Note: * SLF = Standardized Loading Factor.

displays the outcomes of the construct reliability tests and the variance extracted (VE) tests.

Table 11 shows that the value of the construct reliability test for the LV of local politics (PL) is 0.835, while that for the LV of the village facilitator (PD) is 0.778, that for the LV of recruitment of administrators (RP) is 0.849, that for the LV of training and education (DL) is 0.829, that for the LV of organizational culture (BU) is 0.841, that for the LV of work motivation (MT) is 0.858, and that for the LV of management performance (KN) is 0.826. While the variance extracted value of the LV for local politics (PL) is 0.507, that for village facilitator (PD) is 0.469 and that for recruitment of administrators (RP) is 0.585. Training and education (DL) has a value of 0.550, organizational culture (BU) one of 0.572, work motivation (MT) one of 0.602, and management performance (KN) one of 0.544. It can be concluded that the indicators of PL, PD, RP, DL, BU, MT, and KN are legitimate and reliable in generating their latent variables because the construct reliability value obtained is greater than 0.7, and the variance extracted value is greater than 0.5. This is optional because it is valid from a reliability standpoint, even though the variance extracted PD behavior is still low at less than 0.05. Once the CFA model had been analyzed, the full model test was conducted using the model constructed for this study.

4.5. SEM Model Evaluation

4.5.1. Goodness of Fit

In the third stage, a thorough evaluation of the structural equation model (SEM) was conducted. This evaluation began with goodness-of-fit (GoF) measurements, which are crucial for assessing the extent to which the proposed model aligns with the empirical data. Various fit indices were used in this assessment, including absolute fit indices such as the chi-square statistic and the root mean square error of approximation (RMSEA). These indices were interpreted based on recommended cutoff values; for instance, RMSEA values less than 0.08 and CFI/TLI values above 0.90 are generally considered indicative of good model fit, according to Hu and Bentler [129]. The overarching goal of the model fit test was to determine how well the data match the model.

A significant probability value (p) was produced by a modest chi-square value (χ2), an essential indicator of overall fit; this indicated no substantial difference between the input covariance matrix of forecasts and actual data. However, the test’s chi-square (χ2) value was low. Thus, the significance threshold was greater than 0.05, indicating that the data covariance matrix and the calculated covariance matrix are not significantly different.

Figure 3 displays all of the model’s outputs. With a p-value of 0.0000, 310 degrees of freedom (DF), and a chi-square value of 532.53, Figure 3 displays the model estimation results. The created model satisfies the requirements for a model with a moderate fit having an RMSEA value of 0.054, as the RMSEA is less than 0.08.

In

Table 12. Goodness of fit of the model.

Goodness of Fit	Cut Off Value	Estimation	Description
Chi-Square (χ2)	Small	532.53	Fit
Signifikansi	≥0.05	0.000	Not fit
RMSEA	≤0.08	0.054	Fit
GFI	≥0.90	0.870	Marginal Fit
AGFI	≥0.90	0.830	Marginal Fit
CMIN/DF	≤2.00	1.717	Fit
CFI	≥0.95	0.990	Fit
IFI	≥0.90	0.990	Fit
NFI	≥0.90	0.970	Fit
RMSR	≤0.08	0.036	Fit

, the results of the measures used, based on the type of absolute fit measures, are presented. The test results yielded a chi-square value of 532.53, which is considered small or low, indicating that the model fits. Then it produced a CMIN/DF of 1.717; this value is smaller than two, so the model is categorized as a good fit. Furthermore, the root mean squared error of approximation (RMSEA) value is the most informative model measurement, measuring the deviation of the parameter values in a model from the population covariance matrix. This measurement is independent of the number of samples. The RMSEA value obtained was 0.054, which is 0.08–0.1, indicating a model fit. Then, the goodness-of-fit index (GFI) value measured the model’s accuracy in producing the observed covariance matrix. The GFI value obtained was 0.870, which is close to 0.900, indicating a marginal-fit model. At the same time, the adjusted goodness-of-fit index (AGFI) value is also an exact measurement, but it has been adjusted to reflect the effect of the degree of freedom (df) on a model. A fit model is a model that has an AGFI value greater than 0.90. The obtained AGFI value was 0.830, which is close to the criterion of 0.900 and the marginal-fit model. Then, the Comparative Fit Index (CFI) value of this index falls within the range of 0 to 1, where a value closer to 1 indicates the highest level of fit or a perfect fit. The recommended CFI value is 0.95 or greater. The results of this study suggest that the CFI value is 0.99 > 0.95, meaning that the model fits. Furthermore, the incremental fit index (IFI) value is used to overcome the problem of parsimony and sample size. The IFI value obtained was 0.99, which exceeds the cutoff limit of 0.90, indicating that the model is a good fit. Then, the NFI value cut for this index is close to 0.90. An NFI of 1.0 indicates that the model is a perfect fit. In this study, the NFI value was 0.97, which is greater than 0.9, indicating that the model is a good fit. Meanwhile, the RMSR cutoff value for this index is below 0.08. In this study, the RMSR value of 0.036 ≤ 0.08 means that the model fits.

After reviewing all the criteria, it can be concluded that seven show a good fit, two a moderate, and one a low fit. Having more than one model fit test is preferable, but researchers are not required to employ all criteria [130]. According to Hair [131], Researchers should not rely on a single measure to evaluate model, should involve more than one key factor, only four or five key factors. The data and sample validity of the model are enhanced when more GoF requirements are satisfied. Accordingly, the model mentioned earlier is suitable and valid. A fit model is shown by the CFI and RMSR, in addition to the other seven criteria, even though the multivariate normality test indicates abnormality [130].

4.5.2. Determinant Coefficient (Structural Model Test)

Structural models, such as the inner model, predict the causal relationship between latent variables. Reviewing the R² value, which represents the goodness-of-fit test, enables one to evaluate the structural model. With an R² value of 0.74, it can be deduced that the latent variables of local politics (PL), village facilitator (PD), recruitment of administrators (RP), training and education (DL), and organizational culture (BU) account for 74% of the variance in work motivation (MT), with other variables accounting for the remaining 26%. An R² value of 0.82 for management performance (KN) indicates that the variables under investigation—local politics (PL), village facilitator (PD), recruitment of administrators (RP), training and education (DL), organizational culture (BU), and work motivation (MT)—explain 82% of the variance in management performance (KN), with other variables accounting for the remaining 18%. In

Table 13. R-Square values.

	R-Square (R2)	R-Square Adjusted
Work Motivation (MT)	0.74	0.74
Management Performance (KN)	0.92	0.82

, the R² values are presented.

4.6. Hypothesis Testing and Interpretation

A statistical method for interpreting data collected through both controlled experiments and uncontrolled observation is hypothesis testing [132]. To evaluate a model, one must determine whether exogenous variables significantly affect endogenous variables by assessing the causal links between latent variables. A statistically significant coefficient is one with a t-value of 1.96 or higher. In

Table 14. The direct relationships between latent variables.

No.	Direct Influences	Coefficients	Error Values	t-Statistic Values	Descriptions
1	Local Politics (PL) → Work Motivation (MT)	0.23	0.18	1.97	Significant
2	Village Facilitator (PD) → Work Motivation (MT)	0.15	0.19	1.98	Significant
3	Recruitment of Administrators (RP) → Work Motivation (MT)	0.14	0.21	1.96	Significant
4	Training and Education (DL) → Work Motivation (MT)	0.15	0.16	2.09	Significant
5	Organizational Culture (BU) → Work Motivation (MT)	0.31	0.13	2.56	Significant
6	Local Politics (PL) → Management Performance (KN)	0.12	0.19	1.54	Insignificant
7	Village Facilitator (PD) → Management Performance (KN)	0.49	0.21	2.67	Significant
8	Recruitment of Administrator (RP) → Management Performance (KN)	0.10	0.22	0.98	Insignificant
9	Training and Education (DL) → Management Performance (KN)	0.41	0.17	2.16	Significant
10	Organizational Culture (BU) → Management Performance (KN)	0.32	0.17	1.97	Significant

, the test outcomes for direct relationships between latent variables are presented.

The following ten points, drawn from Table 14, will shed light on the hypothesis’s outcomes:

(1)

With a t-statistic value of 1.97 and an error value of 0.18, the coefficient of influence of the LV local politics (PL) on the LV work motivation (MT) was 0.23. The strong positive effect of the LV local politics (PL) on the LV of work motivation (MT) was supported by the t-statistic value of 1.97, which is greater than the t-table value of 1.96, so it has a significant effect.

(2)

The coefficient of influence of the LV village facilitator (PD) on the LV of work motivation (MT) was 0.15, with an error value of 0.19 and a t-statistic value of 1.98, where this value was greater than the t-table value of 1.96, so it has a significant effect.

(3)

The coefficient of influence of the LV of recruitment of administrators (RP) on the LV work motivation (MT) was 0.14, with an error value of 0.21 and a calculated t-value of 1.96. This indicates that the LV of recruitment of administrators (RP) has a significant positive effect on the LV of work motivation (MT), as the calculated t-value of 1.96 was greater than the t-table value of 1.96.

(4)

The coefficient value of the effect of the LV training and education (DL) on the LV work motivation (MT) was 0.15, with an error value of 0.16 and a t-statistic value of 2.09, indicating a significant positive relationship between the two variables. This was supported by the t-statistic value of 2.09, greater than the t-table value of 1.96.

(5)

The coefficient of influence of the LV organizational culture (BU) on the LV work motivation (MT) was 0.31, with an error value of 0.13 and a t-statistic value of 2.56. It can be concluded that the LV of organizational culture (BU) has a positive effect on the LV of work motivation (MT) because the t-statistic value of 2.56 was greater than the t-table value of 1.96.

(6)

The absence of a significant relationship between the LV local politics (PL) and the LV of management performance (KN) was supported by the t-statistic value of 1.54 < the t-table value of 1.96.

(7)

The coefficient of influence of the LV of village facilitator (PD) on LV management performance (KN) was 0.49, with an error value of 0.21 and a t-statistic value of 2.67. Because the t-statistic value was 2.67 > the t-table value of 1.96, it was concluded that the LV of village facilitator (PD) has a significant positive effect on the LV of management performance (KN).

(8)

The value of the influence coefficient of the LV of recruitment of administrators (RP) on the LV of management performance (KN) was 0.10, the error value was 0.22, and the t-statistical value was 0.98. Because the t-statistical value was 0.98 < the t-table value of 1.96, it was concluded that the LV of recruitment of administrators (RP) was not significant in relation to the LV of management performance (KN).

(9)

The value of the coefficient of influence of the LV training and education (DL) variable on the LV management performance (KN) was 0.41, the error value was 0.17, and the t-statistical value was 2.16. Training significantly improves the LV management performance (KN), as the t-statistical value of 2.16 > the t-table value of 1.96.

(10)

With a t-statistical value of 1.97 and an error value of 0.17, the influence coefficient of the LV of organizational culture (BU) on the LV of management performance (KN) was 0.32. It may be inferred that the LV of organizational culture (BU) significantly influences the LV of management performance (KN), as the t-statistical value of 1.97 was greater than the t-table value of 1.96. All the test results for indirect relationships between latent variables can be found in

Table 15. Indirect influence.

No.	Indirect Influences	Coefficients	Error Values	t-Statistic Values	Descriptions
1	Local Politics (PL) → Work Motivation (MT) → Management Performance (KN)	0.23	0.22	3.14	Significant
2	Village Facilitator (PD) → Work Motivation (MT) → Management Performance (KN)	0.27	0.15	2.63	Significant
3	Recruitment of Administrators (RP) → Work Motivation (MT) → Management Performance (KN)	0.19	0.15	2.43	Significant
4	Training and Education (DL) → Work Motivation (MT) → Management Performance (KN)	0.19	0.19	2.46	Significant
5	Organizational Culture (BU) → Work Motivation (MT) → Management Performance (KN)	0.19	0.08	2.34	Significant

.

The following five points, derived from Table 15, will shed light on the hypothesis’s outcomes:

(1)

With a t-statistical value of 3.14 and an error value of 0.22, the effect coefficient of the LV local politics (PL) was 0.23. The LV of local politics (PL) has a positive effect on the LV of work motivation (MT) through the LV of management performance (KN) (t-statistic = 3.14 > t-table = 1.96).

(2)

The LV of village facilitator (PD) had an influence coefficient of 0.27, an error value of 0.15, and a t-statistic value of 2.63 on the LV of management performance (KN) through the LV of work motivation (MT).

(3)

The coefficient of influence of the LV recruitment of administrator (RP) was 0.19, with an error value of 0.15 and a t-statistic value of 2.43. This indicates that the LV recruitment of administrators (RP) positively affects the LV of management performance (KN) through the LV work motivation (MT) variable, as evidenced by the t-statistic value of 2.43, which is greater than the t-table value of 1.96.

(4)

The coefficient of influence of the LV training and education (DL) was 0.19, with an error value of 0.19 and a t-statistic value of 2.46. This indicates that LV training and education (DL) have a positive and significant impact on LV management performance (KN) through the LV of work motivation (MT). The t-table value of 1.96 is smaller than the t-statistic value of 2.46. The LV organizational culture (BU) had an influence coefficient of 0.19, an error value of 0.14, and a t-statistical value of 2.34.

(5)

This shows that the LV organizational culture (BU) significantly affects the LV management performance (KN) through the LV work motivation (MT) (t-table = 1.96 vs. 2.34). We can conclude that LV organizational culture (BU), significantly affects LV work motivation (MT), a latent variable, through management performance (KN), as the t-statistic value was 2.34 > t-table value of 1.96.

5. Discussion

5.1. Direct Effect Between Variables

5.1.1. The Effect of the Latent Variables Local Politics (PL), Village Facilitator (PD), and Recruitment of Administrators (RP) on the Latent Variable Work Motivation (MT)

A t-statistic of 1.97, an error of 0.18, and a coefficient of 0.23 (Table 14) indicated that the latent variable local politics (PL) influences the latent variable work motivation (MT). There was a positive and statistically significant relationship between the latent variables work motivation (MT) and local politics (PL), as the t-statistic value of 1.97 > the t-table value of 1.96. The results of this study indicate that local politics plays a significant role in influencing the work motivation of BUMDES administrators; thus, it is confirmed that the research hypothesis can be accepted. These findings demonstrate that institutional dynamics at the village level shape work motivation not only by technical factors such as training or organizational culture, but also significantly influence power structures and local political dynamics. This aligns with the theory of political embeddedness [133] and institutional theory [134], which emphasizes that organizational behavior exists within a complex web of socio-political structures. BUMDES administrators are often motivated by factors such as political loyalty, social status, and expectations of access to village resources controlled by local elites [135].

Although its political significance at the local level is somewhat low, the reality on the ground is that influence often exceeds the influence of other variables, such as training and education or organizational culture [136,137], indicating that work motivation in BUMDES environments is highly susceptible to political interests. From a policy perspective, these findings have crucial implications. Efforts to increase work motivation are insufficient if they rely solely on technical or administrative interventions; they must also be accompanied by reforms that make village political governance more open and accountable. Those in charge of BUMDES will be more motivated to do a good job when they realize that local politics is invested in the program’s success. They will engage with stakeholders to monitor its implementation, help spread the word, and utilize it together. The political elite helps and supports BUMDES in developing by marketing or utilizing BUMDES. As a result, the orientation of work motivation can transform from one based on political loyalty to institutional professionalism that is sustainable, in line with the principles of inclusive and capacity-based local economic development [133,134,137]. From a scientific perspective, these findings underscore the importance of incorporating political variables as a significant dimension in analyzing work motivation in the rural public sector. As long as political factors are not regulated systematically and transparently, interventions aimed at improving individual capacity will be unable to enhance motivation and performance sustainably and optimally. This finding is aligned with the findings of Kristanti et al. [138], who demonstrated the regional context-dependent relationship between the latent variables local politics (PL) and work motivation (MT) [139,140].

Then, the village facilitator (PD) had a coefficient of 0.15, an error of 0.19, and a t-statistic of 1.98 (Table 14), as it influences the latent variable of work motivation (MT). There was a positive and statistically significant relationship between the latent variables of village facilitator (PD) and work motivation (MT), as indicated by the t-statistic value of 1.98, which exceeds the t-table value of 1.96. The finding that village facilitators have a significant influence on the work motivation of BUMDES administrators highlights the importance of their role in driving local institutional strengthening. Thus, it is confirmed that the research hypothesis can be accepted. A good village facilitator in BUMDES management has a positive impact on the work motivation of BUMDES administrators through their role as a motivator, supervisor, communicator, and facilitator. Based on the contextual work motivation theory framework, the presence of a mentor can strengthen perceptions of work meaning, autonomy, and positive feedback, all of which contribute to increased intrinsic motivation [137]. Compared to other factors such as local politics and recruitment of administrators, this relationship tends to be stronger because the interaction between mentors and mentees is direct, intensive, and continuous. Social learning theory explains that work behavior is influenced by the processes of observation and social reinforcement in the work environment [141], while the embedded development facilitation approach emphasizes that facilitators function as mediators between institutional structures and village development actors [142]. This finding is also consistent with the contingency theory of motivation, which emphasizes that the effectiveness of motivation is determined by the fit between individual characteristics, organizational support, and job demands [143]. The policy implication is the need for rigorous selection, ongoing training, and performance-based evaluation of village facilitators, ensuring that their role is transformative and not merely administrative. According to Fyniel’s research [80], village facilitators play a crucial role in the financial and administrative management of village development and in the implementation of long-term community empowerment. In addition, according to Susanti [144], the village facilitator plays a crucial role in fostering community activities, raising awareness, and involving the community in the process of creating self-sufficient villages that can serve as development subjects.

Work motivation (MT) is a latent variable influenced by the recruitment of administrators (RP) with a coefficient of 0.14, an error of 0.21, and a t-statistic of 1.96 (Table 14). It can be inferred that the latent variable work motivation (MT) is significantly affected by the recruitment of administrators (RP) because the t-statistic value of 1.96 > the t-table value of 1.96. The results, which show that the recruitment of administrators has a significant influence on work motivation, indicate that selection based on competence, transparency, and organizational value alignment plays a vital role in shaping the intrinsic motivation of BUMDES administrators. Thus, it is confirmed that the research hypothesis can be accepted. This finding is consistent with Self-Determination Theory, which highlights the importance of autonomy and competence as key factors driving long-lasting work motivation [78]. Conversely, a recruitment process that ignores the aspects of quality and individual commitment has the potential to reduce motivation due to a weak perception of the meaning of work and contribution to the organization, as explained in the theory of motivation [137]. From a theoretical perspective, these findings also support Human Capital Theory, which asserts that the quality of human resources is a key factor in driving organizational performance [145]. Therefore, the formulation of BUMDES management recruitment policies should prioritize the principles of meritocracy and institutional needs to strengthen work motivation and operational effectiveness on an ongoing basis. Supporting this view is a study by Alfiansyah [146] and Haryani [147], which found that a highly motivated worker and an effective recruitment process for the BUMDES board are critical success factors in empowering communities at the village level and implementing sound governance principles [148]. Based on the research in Muizu [21], it is crucial to establish an effective system of recruitment, mentoring, compensation, and monitoring processes to revive BUMDES-initiated firms. Because of this, BUMDES management may be able to boost employee morale and productivity without sacrificing human resource quality [149].

5.1.2. The Effect of the Latent Variables Training and Education (DL) and Organizational Culture (BU) on the Latent Variable Work Motivation (MT)

The latent variables of training and education (DL) and work motivation (MT) have an impact value of 0.15, with a 0.16 margin of error and a t-statistic of 2.09 (Table 14). The results indicate that latent variable training and education (DL) have a significant positive influence on latent variable work motivation (MT), as the t-statistic value of 2.09 is greater than the t-table value of 1.96. Research findings indicating that education and training have a significant influence on the work motivation of BUMDES administrators, suggesting that capacity building through systematic learning programs contributes directly to intrinsic motivation. Thus, it is confirmed that the research hypothesis can be accepted. This is consistent with Self-Determination Theory, which emphasizes the importance of fulfilling psychological needs such as competence, autonomy, and relatedness [78]. Contextual training that aligns with job requirements enriches skill variety, task meaningfulness, and understanding of work outcomes, as outlined in the Job Characteristics Model [143], ultimately enhancing individual engagement and performance. When compared to other variables, such as the role of village facilitators, recruitment of administrators, education, and training, it shows a more substantial influence because it has more direct characteristics and can be internalized through increased self-efficacy and the relevance of the material [150,151]. Support for this is also evident from the fact that effectively designed training can improve job satisfaction, organizational commitment, and performance, particularly in the public sector [152]. Therefore, BUMDES need to develop sustainable, contextual, and integrated training programs that incorporate institutional strategies to strengthen capability-based governance, foster strategic partnerships with educational institutions, and employ reflective learning approaches [58]. The findings are consistent with those of Amin [153], who found that training and an effective leadership style can enhance the motivation and performance of BUMDES personnel. When people have a better understanding of their role as BUMDES administrators, they can feel more invested in the growth of their villages and, consequently, be more motivated to perform their duties effectively [19,154].

According to the results, the latent variable organizational culture (BU) affects work motivation (MT), a latent variable with a coefficient of 0.31, an error of 0.13, and a t-statistic of 2.56 (Table 14). The latent variable organizational culture (BU) has a significant positive influence on the latent variable work motivation (MT), as indicated by a t-statistic value of 2.56, which exceeds the t-table value of 1.96. The findings of this study suggest that organizational culture has a significant influence on the work motivation of BUMDES administrators. Thus, it is confirmed that the research hypothesis can be accepted. This suggests that shared values, collective norms, and consistent internal practices have a more profound impact on intrinsic motivation than other factors, such as education, training, and village assistance. Organizational culture operates latently and continuously, shaping daily work attitudes and behaviors, in line with Self-Determination Theory, which emphasizes the importance of fulfilling psychological needs for autonomy, competence, and relatedness [78]. When the prevailing culture is participatory and performance-oriented, this reinforces perceptions of the meaning of work and increases the active involvement of managers, as explained in organizational climate theory [155]. Conversely, if elitist local power dynamics influence organizational culture, this can lead to a distortion of professional values and a decline in work motivation, as argued in political embeddedness theory [133]. Therefore, the formation of an organizational culture that is in line with the principles of participatory and accountable governance is key in designing institutional interventions that are not only technical but also transformative [156,157]. Consistent with this finding, Turmudhi [158] suggests that BUMDES should adopt a more positive corporate culture and increase employee dedication by facilitating better dialogue between village institutions and instituting more frequent planning sessions. In this way, the company’s dedication serves as its incentive for the leadership. Therefore, BUMDES must establish a good organizational culture for sustainability.

5.1.3. The Effect of the Latent Variables Local Politics (PL), Village Facilitator (PD), and Recruitment of Administrators (RP) on the Latent Variable Management Performance (KN)

The latent variables of local politics (PL) and management performance (KN) have an effect value of 0.12, with a margin of error of 0.19 and a t-statistic of 1.54 (Table 14). There is no significant relationship between the latent variables of local politics (PL) and management performance (KN), since the t-statistic value of 1.54 is less than the t-table value of 1.96. The finding is that local politics has no significant influence on the performance of BUMDES administrators. Thus, it is confirmed that the research hypothesis can be rejected. This suggests that political dynamics at the village level do not always have a direct impact on organizational performance, particularly in the context of village economic institutions that are increasingly prioritizing professionalism and accountability. BUMDES is unaffected by the actions of political elites and the policies of local governments. This is different from variables such as village assistants, education and training, and organizational culture, which have been shown to have a significant effect on motivation and performance because they are directly related to increasing individual capacity and strengthening collective work [78,137]. The absence of a substantial influence from local politics can be explained using the institutional decoupling approach, which describes how bureaucratic practices and operational decisions within organizations often operate independently of external pressures to maintain formal legitimacy while still ensuring internal efficiency [159]. At the policy level, these findings suggest that technocratic approaches, such as competency-based training and improvements to recruitment systems, have a greater impact on performance than strategies based on political affiliation. Therefore, strengthening the capacity of BUMDES should be directed towards developing a meritocracy-based organizational governance system that is free from political influence, thereby creating an institutional system that focuses on results.

The latent variable village facilitator (PD) was found to influence management performance (KN) with a coefficient of 0.49, an error of 0.21, and a t-statistic of 2.67 (Table 14). The conclusion was drawn that the latent variable of management performance (KN) is significantly affected by the latent variable of village facilitator (PD), as the t-statistic value of 2.67 > the t-table value of 1.96. Research findings indicating that village facilitators have a significant impact on the performance of BUMDES administrators, underscoring the importance of external actors in enhancing institutional capacity and organizational governance at the village level. Thus, it is confirmed that the research hypothesis can be accepted. This correlation tends to be more prominent than for other variables because village facilitators not only act as intermediaries between national policies and local implementation, but also serve as sources of information, supervision, and catalysts in encouraging community participation and public accountability [144]. The active involvement of facilitators in the planning, reporting, and evaluation processes of BUMDES programs directly contributes to enhancing the managerial and administrative capabilities of administrators, ultimately resulting in a positive impact on organizational performance. This finding is consistent with the capacity-building approach in the framework of institutional development theory, which emphasizes the importance of technical support and continuous assistance to form adaptive and effective local organizations [160]. However, the strength of this relationship can be weakened if the quality of the partner is low, inconsistent, or influenced by local political dynamics that cause conflicts of interest or dependency [101]. Therefore, from a policy perspective, there is a need for a more rigorous, competency-based recruitment and training mechanism for village facilitators, as well as a measurable evaluation system to ensure that their presence has a real impact in supporting professional and sustainability-oriented BUMDES governance. This is similar to the findings of Febrianti and Hayati [161], who also found that helping out with tasks such as training, education, or the legal process significantly impacted the effectiveness of BUMDES. Both the operational and formal aspects of BUMDES have been enhanced, which will have a positive effect on community economic empowerment at the village level.

The t-statistic of 0.98, the error of 0.22, and the coefficient for the influence of the latent variable of recruitment of administrators (RP) on the latent variable management performance (KN) was 0.10 (Table 14). The latent variable of recruitment of administrators (RP) does not significantly affect the latent variable of management performance (KN) because the t-statistic value of 0.98 is less than the t-table value of 1.96. Research findings indicating that the recruitment of administrators has no significant effect on the performance of BUMDES administrators suggest that the human resource selection process, although essential, is insufficient to ensure maximum performance if a system of coaching, ongoing training, and adequate incentives is not complemented. Thus, it is confirmed that the research hypothesis can be rejected. Administrators are chosen openly without formal recruitment in Bola District, Wajo Regency, South Sulawesi Province, and most BUMDES administrators have political or familial ties to the village chief. Since experience and education are less important, this can lead to lower-quality hiring. Based on the Human Capital Theory perspective [145], recruitment should ideally be the first step in developing quality human resources. However, in the practice of village institutions, this process is often influenced by non-technical factors such as kinship or political interests, which tend to hinder the application of meritocratic principles [162]. This condition explains why other variables such as training and education, as well as organizational culture, play a more significant role in driving performance, as they contribute directly to improving capacity both individually and collectively [78,156]. The gap between theoretical expectations and empirical findings shows that improving the quality of recruitment in BUMDES must be accompanied by governance reform and institutional professionalism. Therefore, relevant policies need to emphasize the implementation of competency-based recruitment standards and post-placement training in an effort to bridge the gap between input and work output. To improve the performance of BUMDES administrators, it is crucial to enhance the recruitment process, as research by Zainuri [124] suggests, so that new administrators will be prepared to implement excellent governance and accountability practices. Nevertheless, success is contingent upon people’s ability to shoulder their assigned responsibilities.

5.1.4. The Effect of the Latent Variables of Training and Education (DL) and Organizational Culture (BU) on the Latent Variable of Management Performance (KN)

The latent variables of training and education (DL) and management performance (KN) are influenced by one another, with a coefficient of 0.41, an error of 0.17, and a t-statistic of 2.16 (Table 14). It can be inferred that the latent variable of training and education (DL) significantly influences the latent variable of management performance (KN), as the t-statistic value of 2.16 exceeds the t-table value of 1.96. The finding that education and training have a significant impact on the performance of BUMDES administrators suggests that enhancing competencies through formal and informal learning pathways can support the effective implementation of institutional functions. Thus, it is confirmed that the research hypothesis can be accepted. This correlation appears to be stronger than for other variables, such as recruitment or local political dynamics, because education and training directly strengthen the capacity of administrators through mastery of technical knowledge, managerial skills, and understanding of relevant regulations in the management of BUMDES. Administrators at BUMDES can benefit from additional training and education in areas such as marketing, finance, human resources, and relevant laws that impact the administration of BUMDES. Based on the Human Capital Theory perspective [145], training is a form of investment that improves the quality of individuals and has a positive impact on the overall performance of an organization. Meanwhile, the weak link between recruitment and performance may be due to a selection process that is not merit-based or a lack of post-recruitment training programs. The low influence of local politics suggests that the operational functioning of BUMDES is not always determined by political intervention, particularly when governance is established on the principles of professionalism and accountability. Therefore, these findings have important policy implications, namely, the need for strengthened continuous training tailored to local needs and institutional contexts to promote more strategic and sustainable human resource development in rural areas [163]. This finding is in agreement with the findings of Ariska et al. [164], who found that investing in the education and growth of BUMDES administrators can have a significant impact on the system’s efficiency and effectiveness in the long run, as well as on the reliability of its financial reports.

The coefficient of 0.32, error of 0.17, and t-statistic of 1.97 (Table 14) indicate the influence of organizational culture on management performance. The latent variable management performance (KN) is positively and significantly impacted by organizational culture (BU), as indicated by the t-statistic value of 1.97, which exceeds the t-table value of 1.96. Research findings suggesting that organizational culture has a significant impact on the performance of BUMDES administrators, highlighting the importance of collective values, shared norms, and structured internal practices in enhancing work effectiveness at the village institutional level. Thus, it is confirmed that the research hypothesis can be accepted. When compared to other factors, such as local political dynamics and recruitment processes, the significant influence of organizational culture can be explained through the organizational culture theory framework [149], which posits that a solid internal value system can foster team cohesion, increase emotional commitment, and facilitate efficient coordination among members. Suppose that organizational cultures are adaptive and oriented towards achieving institutional goals. In such cases, managers tend to have clearer roles, a shared vision, and a higher intrinsic motivation to achieve optimal performance. This finding aligns with the view in organizational behavior theory, which posits that a positive and value-based work environment is a significant driver of productivity and innovation [165]. Conversely, variables such as local politics, which is external and not directly involved in the organization’s daily operational activities, tend to have a weaker influence on the performance of administrators. Therefore, the policy implications of these results indicate the importance of a targeted strategy to strengthen organizational culture through inclusive leadership, open communication, and systematic internalization of institutional values. Thus, maintaining these cultural aspects becomes the primary foundation for creating sustainable, professional, and accountable BUMDES governance. This highlights the significance of a supportive company culture in motivating BUMDES administrators to perform well. This finding aligns with the findings of Olakunle [166], who found that a compelling company culture can boost employee productivity. Professionalism, discipline, honesty, productivity, and punctuality are all essential components of a healthy work culture that can significantly enhance an organization’s ability to motivate and retain its employees.

5.2. Indirect Effects Between Latent Variables

5.2.1. The Effect of the Latent Variable Local Politics (PL) on the Latent Variable Management Performance (KN) Through the Latent Variable Work Motivation (MT)

Local politics (PL) is a latent variable with an effect coefficient of 0.23, an error value of 0.22, and a t-statistic of 3.14 (Table 15). The strong positive effect of the latent variable of local politics (PL) on the latent variable of management performance (KN) through the latent variable of work motivation (MT) is supported by the t-statistic value of 3.14 > the t-table value of 1.96. Thus, it is confirmed that the research hypothesis can be accepted. Based on the previous data, it can be concluded that the latent variable local politics (PL) does not significantly affect the latent variable management performance (KN). However, the value becomes directly significant when the work motivation latent variable is used as a link between PL and KN. To what extent do the community, political elites, village policies, and the village chief all rally behind the BUMDES’s leadership? This indicates the extent to which local politics affects management’s effectiveness. The finding that local politics has a significant influence on the performance of BUMDES administrators through work motivation but does not show a significant direct impact on performance illustrates the complexity of the relationship between local power structures and village organizational dynamics. Conceptually, this pattern is consistent with the mediated moderation approach in work motivation theory and the open organizational system theory framework, which states that the influence of politics on performance is not direct but mediated by psychological aspects such as perceptions of fairness, managerial support, and incentives [78,167]. In the context of BUMDES, the presence of local political figures can create normative pressure or open access to resources and social legitimacy, which in turn affects the level of motivation of administrators and impacts their performance. The absence of direct relations can also be interpreted as a manifestation of efforts to maintain institutional professionalism in the face of external political influence, in line with the principles of institutional autonomy and performance-based governance [134,160]. Therefore, policies aimed at improving the performance of BUMDES should not only focus on reducing the role of local politics but also be directed at strengthening work motivation through mechanisms of transparency, achievement-based rewards, and systematic training. Considering the plurality of village political arenas, this approach is more adaptive in supporting sustainable institutional performance. Stakeholders provide attention in the form of supervision and guidance, as well as assisting in the marketing process of BUMDES, so that the administrators feel appreciated and motivated to work. This will improve the performance of the administrators. Herman [168] identifies several factors that impact the performance of BUMDES administrators. Some of these factors are external, such as local politics, while others are internal, including community expectations and support, which ultimately influence work motivation.

5.2.2. The Influence of the Latent Variables of Village Facilitators (PD) and Recruitment of Administrators (RP) on the Latent Variable of Management Performance (KN) Through the Latent Variable of Work Motivation (MT)

The t-statistic was 2.63, the error value was 0.15, and the coefficient of effect for the latent variable village facilitator (PD) was 0.27 (Table 15). A favorable impact of the latent variable village facilitator (PD) on the latent variable management performance (KN) via the latent variable work motivation (MT) was inferred from the t-statistic value of 2.63 > the t-table value of 1.96. Research findings showing that village facilitators have a significant influence on the performance of BUMDES administrators through work motivation confirm the strategic position of facilitators as external actors in strengthening institutional capacity and the effectiveness of village organizations. Thus, it is confirmed that the research hypothesis can be accepted. The strength of this relationship is relatively more prominent than for other variables, because village facilitators not only perform technical policy implementation functions, but also act as empowerment agents capable of fostering self-confidence, a sense of capability, and meaning in work for BUMDES administrators through a participatory and solution-oriented facilitation approach [17]. From a theoretical perspective, these findings are consistent with the Self-Determination Theory framework [78], which emphasizes that social support that promotes individual autonomy can enhance intrinsic motivation, which in turn has a positive impact on performance. In the institutional reality of villages, which often face resource constraints and local socio-political challenges, the presence of competent and responsive facilitators is a key factor in driving sustainable institutional change.

Therefore, village assistance policies should not only focus on administrative aspects, but also on strengthening the practical dimension and interpersonal relationships of the assistants. In practical terms, the government needs to review the selection and training mechanisms for facilitators to place greater emphasis on mastering facilitation skills, building participatory relationships, and gaining a deep understanding of local social dynamics, which have been proven to contribute to the improvement of community-based institutions’ performance. Nurkhamid et al.’s [169] findings that village facilitators play a crucial role in enhancing BUMDES administrators’ performance and overcoming development hurdles are consistent with this finding. Effective mentoring has a positive effect on administrators’ motivation and performance.

The t-statistic was 2.43, the error value was 0.15, and the coefficient of impact for the latent variable recruitment of administrators (RP) was 0.19 (Table 15). The results show that that the latent variable recruitment of administrators (RP) significantly influences the latent variable management performance (KN) via the latent variable work motivation (MT), as the t-statistic value of 2.43 > the t-table value of 1.96. Previous test results indicated that the latent variable of administrator recruitment had a significant direct effect on work motivation. Thus, it is confirmed that the research hypothesis can be accepted. However, the latent variable of management recruitment did not have a significant direct effect on management performance. These two test results are contradictory. Subsequently, a test was conducted to examine the indirect relationship between the latent variable of administrator recruitment and the latent variable of management performance, mediated by the latent variable of work motivation. The results of the study show that the recruitment of BUMDES administrators has a significant effect on performance through work motivation, even though it does not have a direct effect. This suggests that motivation serves as a psychological mediator linking human resource quality to organizational performance achievement. This finding is consistent with the concept of mediated moderation in work motivation theory [78], which emphasizes that individual success in the workplace does not only depend on initial stages such as the selection process, but is also influenced by the extent to which intrinsic motivation drives emotional and cognitive engagement in performing tasks. The insignificant direct effect of recruitment on performance is likely due to the weakness of competency-based selection systems, meaning that recruited individuals may not necessarily perform optimally unless motivated by strong incentives. This finding explains that recruiting qualified administrators will result in individuals who are ready to work and needed by the organization. Administrators will feel motivated because they are working with the right people. They will cooperate, exchange ideas, and share experiences with fellow team members. This will increase their work motivation and have an impact on performance.

In terms of policy, these findings emphasize the importance of developing a recruitment system that not only assesses administrative qualifications but also includes strategies for fostering motivation, such as value-based training, ongoing mentoring, and incentive systems that foster commitment to organizational goals. From the perspective of Human Capital Theory [145] and Self-Determination Theory [78], improving the performance of BUMDES administrators requires an integrative approach that combines structural aspects with internal psychological processes in order to produce consistent performance oriented towards village development. The board’s performance will improve after receiving encouragement. That is why it is crucial to involve and empower local elites and stakeholders so that they can help advance BUMDES together [170].

5.2.3. The Effect of the Latent Variables of Training and Education (DL) and Organizational Culture (BU) on the Latent Variable of Management Performance (KN) Through the Latent Variable Work Motivation (MT)

Training and education (DL) was a latent variable with an effect coefficient of 0.19, an error value of 0.19, and a t-statistic of 2.46 (Table 15). It can be inferred that the latent variable of training and education significantly influences the latent variable of management performance (KN) through the latent variable of work motivation (MT), as the t-statistic value of 2.46 is greater than the t-table value of 1.96. Research results indicating that education and training have a significant impact on the performance of BUMDES administrators through increased work motivation emphasize the importance of human resource capacity-building strategies in promoting the effectiveness of village organizations. Thus, it is confirmed that the research hypothesis can be accepted. Consistent with previous findings, this finding confirms that training and education contribute to better performance; moreover, it demonstrates that employees report higher motivation and enthusiasm after receiving training and education. This suggests that intrinsic motivation at work can significantly contribute to increased productivity [63]. The strength of this relationship surpasses that of other variables because education and training directly contribute to the development of technical skills, a professional work ethic, and self-confidence, which underlie the formation of intrinsic motivation [78]. The motivation formed from this learning process functions as a mediating channel that transforms knowledge into productive work behavior, in line with the Human Capital Theory framework [145] and Self-Determination Theory. These findings also confirm that interventions in the form of training are easier to control and measure in terms of effectiveness than contextual factors such as organizational culture and management recruitment. Therefore, the policy implication is the importance of designing and institutionalizing sustainable training programs tailored to specific local needs to strengthen BUMDES governance. Additionally, these results provide a foundation for policymakers to allocate resources more strategically toward capacity development for BUMDES managers, thereby ensuring the institutional sustainability of BUMDES amid the social and political challenges faced by villages.

From a t-statistic of 2.34, an error of 0.08, and a coefficient of effect of 0.19 (Table 15) for the latent variable organizational culture (BU), it can be inferred that the latent variable significantly influences the latent variable management performance (KN) via the latent variable work motivation (MT), as the t-statistic value of 2.34 > the t-table value of 1.96. Research findings indicating that organizational culture has a significant impact on the performance of BUMDES administrators through work motivation suggest that collective values, shared norms, and consistent internal practices play an essential role in shaping committed, proactive, and achievement-oriented work behavior. Thus, it is confirmed that the research hypothesis can be accepted. This influence appears to be more dominant than for other variables because organizational culture provides a framework for thinking and an emotional foundation that facilitates the internalization of organizational goals by managers, thereby supporting the formation of stable and sustainable intrinsic motivation [156]. This finding is in line with the principles of Self-Determination Theory [78], which emphasizes the importance of a social environment that promotes shared values and work autonomy in strengthening intrinsic motivation, ultimately leading to improved performance.

In contrast, factors like recruiting, education, and training are typically technical and external; hence, their impact on motivation and performance is largely contingent upon the presence of a supportive organizational framework. Leaders who cultivate a happy workplace might identify methods to acknowledge their employees’ contributions, whether through public recognition, private commendations, or monetary rewards. The management is inspired and appreciated, which enables them to make a significant impact on enhancing the performance of BUMDES. Thus, the policy implications of these findings highlight the importance of a solid organizational culture development strategy that aligns with the collective mission through transformative leadership, a unifying vision, and the strengthening of collaborative work values at the BUMDES level. Such a strategy not only enhances sustained work motivation but also reinforces institutional effectiveness within the complex social and cultural context of the village, in line with previous studies by Lukman et al. [171]. These findings suggest that enhancing a positive organizational culture, providing adequate facilities, and offering appropriate motivation can significantly improve employee performance. The significance of organizational culture in shaping motivation and performance is demonstrated in this context. Consistent with the results of this study, Muis et al. [114] also discovered that organizational culture had a significant impact on business performance.

The application of digital technology and innovative management plays an important role in increasing the motivation of MSME and BUMDES actors, which in turn has a positive impact on overall business performance. Digital strategies, as explained by Díaz-Pelaez and Chura-Quispe [172] encourage innovation and efficiency in operations, which become a new source of motivation for MSME actors to develop. Accessible technology support and expanded market access, as demonstrated by Antoni et al. [173], can strengthen confidence and innovation spirit within rural communities. This is further reinforced by the findings of Fakhruddin et al. [174], which show that digital technology can empower women through digital entrepreneurship, creating strong internal motivation and performance. When motivation is combined with digitalization tools and innovative management, as outlined by Kurniasari and Suresh [175,176], a sustainable, adaptive, and high-performing local business ecosystem is created, particularly in the context of rural economic empowerment and BUMDES.

This kind of transformation also helps village organizations such as BUMDES build competitiveness on par with larger business entities, as they have standardized and measurable work systems [86,177]. Therefore, the integration of digital technology and managerial strategies in BUMDES not only enhances work effectiveness but also provides a concrete example of how community-based organizations can adopt global practices in a contextual and meaningful way.

6. Conclusions and Recommendations

6.1. Research Conclusions

This study aimed to analyze the influence of local politics (PL), village facilitators (PD), recruitment of administrators (RP), training and education (DL), and organizational culture (BU) on work motivation (MT) and management performance (KN) in Wajo Regency, South Sulawesi Province, Indonesia. Primary data were collected using structured interviews from 250 respondents. The respondents included administrators of village-owned enterprises (BUMDES), community leaders, and representatives from the private sector. The primary data collected were analyzed using structural equation modeling (SEM).

The SEM analysis results showed that the indicators used can effectively measure all latent variables. Then, it was found that the LVs of local politics (PL), village facilitator (PD), recruitment of administrators (RP), training and education (DL), and organizational culture (BU) had a significant and positive direct effect on work motivation (MT) in the BUMDES. At the same time, the LVs of local politics (PL) and recruitment of administrators (RP) had an insignificant effect on the management performance variable (KN). In addition, it was also found that the latent variables of local politics (PL), village facilitator (PD), recruitment of administrators (RP), training and education (DL), and organizational culture (BU) had a significant indirect effect on the management performance (KN) of BUMDES through work motivation (MT). Therefore, work motivation (MT) acts as a central mediating variable that connects external and internal organizational factors with BUMDES performance. These results reinforce the validity of the conceptual model that has been formulated and highlight the urgency of policy interventions aimed at increasing work motivation to maximize BUMDES management performance.

This study expands the application of SEM (structural equation modeling) by integrating institutional, individual, and cultural dimensions into a single integrated model of work motivation and performance, as these variables have often been studied separately in the past. This study demonstrates that SEM remains reliable and relevant for analyzing complex interactions between latent variables in the context of village governance, although it is often considered unsuitable for complex multivariate approaches.

6.2. Relevant Recommendations

The findings above provide several evidence-based policy recommendations:

(1)

Given that local politics only indirectly influences management performance through work motivation, local governments need to ensure that political processes do not hinder the management of BUMDES. Instead, local political actors need to be involved in strengthening governance, marketing, and supervision, including the use of digital tools to enhance transparency and oversight, rather than in operational decision-making, which has the potential for bias.

(2)

The significant role of village facilitators in motivating work emphasizes the need to strengthen their functions through technical training, regular supervision, and performance-based incentives. Digital skill development and innovative management tools will enhance their capacity to support BUMDES efficiency and work ethics.

(3)

Since recruitment of administrators did not have a direct impact on management performance, a merit- and competency-based selection system needs to be established. Implementing digital recruitment systems, open and objective recruitment mechanisms can improve accountability and attract individuals with high work motivation and relevant expertise. Management will be motivated to work and performance will be enhanced because they have the right team members.

(4)

Education and training, along with a conducive organizational culture, are key drivers of work motivation and management performance. Therefore, it is essential to develop sustainable training policies that are based on local needs and cultural transformation, supporting innovation, collaboration, and accountability. Using digital learning platforms and innovation-focused content ensures relevance, sustainability, and continuous improvement.

(5)

Since work motivation acts as an intermediary in the relationship between structural variables and management performance, policies related to motivation, such as rewards, career development, and the implementation of participatory decision-making processes, need to be applied in the village administration system. Digital participatory tools to promote inclusive, accountable decision-making in village administration should be utilized.