In the era of post-pandemic workplace transformation, the operational landscape of companies has undergone a significant paradigm shift. One of the primary adaptations made by various organizations to maintain productivity and business continuity is the adoption of more flexible working systems. In principle, this flexibility is expected to be an ideal solution for improving both employee well-being and productivity (Sidauruk et al., 2025). However, in reality, this shift in work culture often introduces new challenges. The implementation of adaptive policies does not always correlate directly with performance improvement, especially if such freedom is not accompanied by mental readiness, solid self-management, and clear boundaries between professional and personal lives (Fauzi et al., 2025). Data indicates performance fluctuations leading to a significant 12% decrease in the final quarter of 2025, proving that work flexibility has not yet delivered optimal performance impacts.

This phenomenon is clearly reflected in the operational dynamics at PT Kareb Alam Sejahtera. As a company that has adopted work flexibility, the organization faces a reality where these policies have not yielded optimal results. Based on internal data for the 2024–2025 period (see Table 1), a concerning fluctuation in employee performance was found. This is evidenced by a downward trend in Key Performance Indicator (KPI) achievements, with a sharp decline of 12% in the fourth quarter of 2025 (from 82% to 70%). Furthermore, the work system expected to provide convenience has instead shown indications of problems regarding work engagement. Based on Table 2, this is visible from the spike in absenteeism reaching 8.4% and a 14.5% latency rate in flexible system logins by the end of 2025. The increasing turnover rate (4.5%) also aligns with complaints related to burnout amidst a supposedly flexible system. These conditions suggest that employees are struggling to separate professional and personal time (Work-Life Balance), while also indicating a lack of individual initiative (Self-Leadership) as the root cause of declining work engagement, which ultimately erodes overall corporate performance.

Theoretically, the drive to conduct this research is motivated by a research gap consisting of inconsistencies in previous findings regarding the influence of these three factors on performance. Regarding FWA, (Mujanah, 2020) state it has a positive impact because it reduces commuter stress; however, Pratama & Budiono (2026) found insignificant or even negative impacts because FWA can trigger professional isolation and communication disruptions. Similar debates occur with the Self-Leadership variable; (Solihah & Mujanah, 2026) prove its positive influence through increased innovation, while Nur et al. (2026) found insignificant results in employees with excessive workloads because the ability to lead oneself is hindered by fatigue. Likewise, with WLB, (Harun et al., 2023) found a significant influence on increasing performance, whereas (Istanti et al., 2025) argue that excessive WLB can actually decrease target achievement because employees over-prioritize leisure time.

Given these inconsistencies in the literature, it is crucial to include psychological variables to bridge the relationship between working conditions and performance. Therefore, this study positions Work Engagement as a mediating variable to clarify the mechanism of this influence. Practically, this research is highly urgent for PT Kareb Alam Sejahtera to formulate targeted HR policies and ensure whether current flexibility truly engages employees or merely loosens discipline. Academically, this research is imperative to fill the literature gap concerning how physical working conditions (FWA) and internal employee conditions (Self-Leadership and WLB) interact to form engagement in the manufacturing sector during the hybrid work era. Based on the background and research gaps described, this study aims to analyze and prove the interactions between these variables comprehensively. Specifically, this research aims to test the influence of Flexible Working Arrangement, Self-Leadership, and Work-Life Balance on Employee Performance at PT Kareb Alam Sejahtera, both directly and indirectly through Work Engagement as a mediating variable.

Employee Performance

Employee performance is the accumulation of work results in terms of quality and quantity achieved by an employee in carrying out their duties according to the responsibilities given by the organization. Robbins and Judge (2017) emphasize that performance is influenced by the interaction between ability, motivation, and opportunity. Based on Robbins and Judge (2017), performance is measured through five main indicators:

Flexible Working Arrangement (FWA)

Flexible Working Arrangement refers to the autonomy and discretion granted to employees regarding the time and location of their work. Armstrong and Taylor (2020) explain that FWA is a modern HR management strategy to increase employee satisfaction and retention by accommodating personal needs without sacrificing organizational goals. According to Nastiti and Irawan (2022), FWA indicators are generally measured through:

Self-Leadership

Self-Leadership is the process of self-direction and self-motivation to achieve goals through the mastery of cognitive and behavioral strategies. Neck, Manz, and Houghton (2020) define this concept as an individual's ability to take control of their own behavior, particularly in work environments that minimize direct supervision (less-supervised environments). It is measured using the Revised Self-Leadership Questionnaire (RSLQ) with three strategic dimensions (Kamelane & Tuanaya, 2025):

This study utilizes a quantitative approach with an explanatory research design to test and explain the causal relationships between variables. The research population specifically focuses on head office (non-production) employees whose work systems are relevant to the implementation of Flexible Working Arrangement policies, including the managerial, administrative, HR, and finance divisions at PT Kareb Alam Sejahtera, totaling 83 individuals. Given the size of the population, sampling was conducted through a saturated sampling (total sampling) technique, where all 83 employees in these divisions were designated as research respondents to maximize data representation and accuracy. Primary data were collected using an online questionnaire instrument based on a 5-point Likert scale.

After these requirements were met, the data were analyzed descriptively to map the profiles and distribution of respondents' answers. Before proceeding to the main analysis stage, the collected data underwent instrument testing, which included validity tests (to ensure the accuracy of measurement indicators) and reliability tests using Cronbach's Alpha (to ensure questionnaire consistency). Furthermore, to fulfill the statistical prerequisites for regression, classical assumption tests were performed including normality, multicollinearity, and heteroskedasticity tests to ensure the resulting estimation model is valid and unbiased. Subsequently, for the hypothesis testing stage, considering that the conceptual model involves testing direct and indirect effects simultaneously, inferential analysis was operated using the PROCESS Macro Model 4 by Andrew F. Hayes via IBM SPSS software. This hypothesis test serves to evaluate the significance of the formulated influences between variables. Specifically, the significance of the role of Work Engagement as a mediating variable was evaluated using the Bootstrapping technique at a 95% confidence level with 5,000 bootstrap samples. The mediation effect is declared significant only if the range between the Lower Level and Upper Level Confidence Interval (LLCI-ULCI) does not include zero (0).

Classical Assumption Test

Based on the table above, the normality test results for the influence of Flexible Working Arrangement, Self-Leadership, and Work-Life Balance on Employee Performance through Work Engagement with a sample size of 83 show a significance value (Asymp. Sig. 2-tailed) of 0.173. Since this value is greater than 0.05, it can be concluded that the residual data in this regression model are normally distributed. Consequently, the statistical requirements for normality are fulfilled, and the subsequent classical assumption tests can proceed.

Based on the table above, the multicollinearity test results demonstrate that none of the independent variable Flexible Working Arrangement, Self- Leadership, and Work-Life Balance nor the mediating variable Work Engagement, have a Tolerance value below 0.10. The tolerance values, ranging from 0.345 to 0.959, indicate that there is no problematic correlation among the variables. Similarly, the Variance Inflation Factor (VIF) results for all variables are well below the threshold of 10, with the highest value being only 1.650. These results confirm that the regression model is free from multicollinearity issues. Thus, the model is considered valid and reliable for further inferential analysis.

Based on the results of the heteroscedasticity test in the image above, it can be observed that the scatterplot shows a random and even distribution of points around the number 0 on the Y axis. This implies that there is no indication of heteroscedasticity in the regression model.

Path Analysis

The mediation hypothesis testing in this study was conducted using the PROCESS Macro Model 4 developed by Andrew F. Hayes via IBM SPSS software. This analysis investigates the relationships between Flexible Working Arrangement (FWA), Self-Leadership, and Work-Life Balance (WLB) as independent variables (X), Employee Performance as the dependent variable (Y), and Work Engagement as the mediating variable (M). The analysis was performed with a total sample size of 83 respondents (N=83). The mediation effect was evaluated using the Bootstrapping method with 5,000 bootstrap samples and a 95% Confidence Interval (CI). This method is used to determine the significance of the indirect effect; the mediation role is considered statistically significant if the range between the Lower Level Confidence Interval (LLCI) and the Upper Level Confidence Interval (ULCI) does not contain the value zero (0). Conversely, if the interval crosses zero (includes both negative and positive values), the mediating variable is declared unable to bridge the influence of the independent variable on the dependent variable.

Since the significance value is less than 0.05 and the F-count is greater than the F-table, it can be concluded that Flexible Working Arrangement, Self- Leadership, Work-Life Balance, and Work Engagement simultaneously have a positive and significant effect on Employee Performance at PT Kareb Alam Sejahtera. This indicates that the regression model is a good fit (Goodness of Fit) and is reliable for predicting performance trends based on these variables.

Referring to the table, the Adjusted R Square value is 0.719 or 71.9%. his result indicates that 71.9% of the variation in Employee Performance can be explained by the four variables included in this study: Flexible Working Arrangement, Self-Leadership, Work-Life Balance, and Work Engagement. The remaining 28.1% (100% - 71.9%) is attributed to other variables or factors outside of this research model that were not examined in this study, such as organizational culture, compensation, or physical work environment.

1. The Influence of Flexible Working Arrangement on Work Engagement The results indicate that the implementation of flexible working arrangements (FWA) significantly encourages higher work engagement. When employees at PT Kareb Alam Sejahtera are given the autonomy to manage their work time and location, they show higher vigor and dedication. This is supported by Kumala (2025), who states that FWA provides the autonomy needed for employees to design an optimal work environment, thus fostering a deeper mental connection. Similarly, Jihadi et al. (2021) emphasize that flexibility in a hybrid work era acts as a catalyst for employee psychological attachment to their roles.

2. The Influence of Self-Leadership on Work Engagement Self-leadership is proven to be a vital driver of work engagement. Employees who possess strong self-direction and self-motivation strategies are better equipped to stay focused, especially under minimal supervision. According to Kamelane et al. (2025), individuals with high self-leadership use cognitive strategies to maintain enthusiasm. This aligns with findings by Neck, Manz, and Houghton (2020) in Hisam (2024), which suggest that the ability to self-reward and set personal goals prevents engagement from dropping during isolated working conditions.

3. The Influence of Work-Life Balance on Work Engagement Harmony between professional and personal life significantly boosts work engagement. When personal lives are not sacrificed, employees experience less burnout. Suhartini and Sary (2025) explain that work engagement thrives when employees have sufficient energy reserves derived from a balanced life. Furthermore, Saputra et al. (2024) found that subjective perceptions of harmony allow employees to be fully "absorbed" in their tasks without the distraction of unresolved personal pressures.

4. The Influence of Flexible Working Arrangement on Employee Performance FWA has a direct and positive impact on performance by reducing stressors like long commutes. Sasando and Harmastuti (2025) argue that FWA allows employees to align work tasks with their productivity peaks, leading to higher quality output. This supports the notion that at PT Kareb Alam Sejahtera, the physical environment and time sovereignty granted to employees directly translate into achieved KPIs.

5. The Influence of Self-Leadership on Employee Performance Strong self-leadership leads to better work outcomes because employees do not require constant monitoring. Kamelane and Tuanaya (2025) prove that self-leadership enhances innovative behavior and task efficiency. As noted by Indrayani et al. (2025), in modern work settings, the internal initiative of an employee is a more reliable predictor of performance than external supervision.

Based on the results of the research that has been conducted, the following conclusions can be drawn: Based on the research conducted at PT Kareb Alam Sejahtera, it can be concluded that Flexible Working Arrangements, Self-Leadership, and Work-Life Balance serve as vital pillars in enhancing both Work Engagement and Employee Performance. These three factors directly empower employees to manage their professional responsibilities more effectively while maintaining personal well-being. Furthermore, Work Engagement is proven to be a significant driver of performance, as employees who feel a deep emotional and mental connection to their roles are more likely to contribute positively to the company’s objectives. However, an interesting finding emerges regarding the mediation role of Work Engagement. While all independent variables successfully boost engagement, this engagement does not act as the primary bridge to performance. Instead, the flexibility provided, the strength of individual leadership, and a balanced life impact performance through more direct functional and physical pathways such as reduced stress, better time management, and increased discipline rather than solely through psychological attachment.

To optimize employee outcomes, PT Kareb Alam Sejahtera should prioritize the formalization of structured Flexible Working Arrangement (FWA) policies that align with individual productivity cycles, while simultaneously investing in Self-Leadership development to foster a proactive, independent workforce that thrives without excessive external supervision. Furthermore, the organization must actively cultivate a Work-Life Balance culture that respects personal boundaries, ultimately creating a high-performance environment where employees are not only productive but also resilient and mentally health