Page 3218
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
The Impact of Flexible Working Arrangements on Turnover
Intentions as mediated by Job Control among Selected IT-BPM
Employees in Cebu
Gensbergh G. Rago, Melanie R. Banzuela and Paolo Louis Manghihilot
School of Business and Economics, University of San Carlos, Cebu City, Philippines
DOI:
https://doi.org/10.47772/IJRISS.2025.914MG00247
Received: 04 December 2025; Accepted: 12 December 2025; Published: 29 December 2025
ABSTRACT
Employee turnover is a persistent challenge in the Information Technology–Business Process Management (IT-
BPM) sector, particularly in the Philippines where attrition rates remain among the highest in Asia. Although
flexible working arrangements (FWAs) are increasingly used to improve retention, previous studies offer
inconsistent findings and provide limited evidence on the mediating role of job control. This study examined the
effect of FWAs on turnover intention, with job control as a mediating variable, among 589 IT-BPM employees
in Cebu City. A quantitative, correlational, and cross-sectional design was employed, and data were analyzed
using Partial Least Squares Structural Equation Modelling (PLS-SEM) in SmartPLS 4.0. The measurement
model demonstrated strong reliability and validity, while the structural model showed acceptable fit and
predictive relevance. The findings indicated that FWAs significantly increased job control but unexpectedly
showed a positive direct effect on turnover intention, suggesting that flexibility may heighten intentions to leave
when not paired with adequate support or clear work–life boundaries. Job control exhibited a significant negative
effect on turnover intention, confirming its role as a protective factor against attrition. Mediation analysis
revealed a competitive mediation effect in which the autonomy generated by FWAs offset, but did not fully
eliminate, their positive direct influence on turnover intention. The study contributes to Social Exchange Theory
by clarifying that the benefits of FWAs depend on employees’ perceptions of fairness, autonomy, and
organizational support. Practically, the results suggest that IT-BPM firms must strengthen job control
mechanisms to ensure that FWAs translate into meaningful autonomy. Future research may integrate additional
psychological, organizational, and contextual variables using longitudinal and comparative designs.
Keywords - Flexible Working Arrangements, Job Control, Turnover Intentions, IT-BPM Industry
INTRODUCTION
High employee turnover is a common problem in the IT-BPM industry across many developing countries. In
India, for example, turnover rates of about 14–15% continue to challenge companies that rely on stable and
skilled workers (Salunkhe et al. 2024; Patil 2025). Similar issues appear in Pakistan, Sri Lanka, Malaysia, and
Bangladesh, where employees often leave because of weak management support, low or uncompetitive pay,
poor work–life balance, and limited career growth opportunities (Farooq et al. 2022; Kanchana & Jayathilaka
2023; Seneviratna et al. 2024; Sidike & Zulkifly 2025; Rahman et al. 2023). In the Philippines, the situation is
even more serious. Turnover in the IT-BPM sector has reached as high as 38% in previous years and, although
it has dropped to about 19%, it is still higher than many neighboring countries (Repaso et al. 2022; Kurata et al.
2023; Bernardo et al. 2023). By comparison, turnover rates in India and other Asian IT-BPM hubs generally fall
between 15–20%, suggesting that the Philippines faces greater difficulty in keeping its employees (Presbitero et
al. 2021; Patil 2025; Sethar et al. 2022). These patterns show that while high turnover is a shared regional
concern, the challenge is especially pronounced in the Philippine IT-BPM industry.
There is a clear research gap in understanding how flexible working arrangements (FWAs) affect employees’
turnover intentions when job control is the main factor that explains this relationship, especially in the IT-BPM
Page 3219
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
industry. Most existing studies show that FWAs help reduce turnover, but they usually explain this through other
factors such as job satisfaction, organizational commitment, or employee engagement (Haines et al. 2024; Yang
et al. 2024; Berber et al. 2022). A few studies discuss job control and suggest that FWAs can increase employees’
autonomy, which may lead to lower turnover. However, these studies often look at job control together with
other variables and do not focus on its unique mediating role, particularly in IT-BPM settings (Haines et al. 2024;
Tsen et al. 2021). Recent reviews also highlight the need to understand how FWAs work differently across
industries and cultural contexts, including IT-BPM (Yang et al. 2024). Very few studies directly test job control
as the only mediator between FWAs and turnover intentions among IT-BPM employees. Most research comes
from general service or high-tech industries, leaving limited evidence specific to IT-BPM firms, especially in
developing countries. Scholars also call for more context-specific studies to understand how job control shapes
the effects of FWAs across different job roles, cultures, and organizational systems (Yang et al. 2024; Tsen et
al. 2021). These gaps show the need for more focused research on the role of job control in the FWA–turnover
relationship in the IT-BPM sector. Based on the foregoing, the study examined the impact of flexible working
arrangements on turnover intentions as mediated by job control among selected IT-BPM employees in Cebu.
Specifically, the study answered the following questions:
1. What is the profile of the respondents in terms of age, sex, parental status and job level?
2. What are the variable levels of the flexible working arrangement, job control and turnover intention?
3. Do flexible working arrangements and job control impact turnover intention?
4. Does flexible working arrangements impact job control?
5. Does job control mediate the impact of flexible working arrangements on turnover intention?
6. What are the implications of the findings to the theory, practice and future research direction?
The study provides practical insights for IT-BPM firms by showing how flexible work arrangements and job
control influence turnover intentions. The findings can guide organizations in refining workforce policies,
strengthening retention efforts, and implementing flexibility practices that support productivity without
increasing attrition. For employees, the results highlight the value of autonomy and supportive management in
reducing turnover intention. For researchers, the study offers updated evidence from the Philippine IT-BPM
sector, clarifies the mediating role of job control, and presents a validated model for future research on flexible
work and employee outcomes.
REVIEW OF LITERATURE
Turnover intention refers to an employee’s conscious and deliberate willingness or preparedness to leave their
current organization within a certain period, even if they have not yet taken concrete steps to resign (Lazzari et
al. 2022; Narwaria et al. 2024; Budin 2024; Bernardo et al. 2023). It is widely recognized as the strongest
predictor of actual employee turnover, especially in industries like IT-BPM where direct quit data may be
unavailable or difficult to track (Lazzari et al. 2022; Narwaria et al. 2024; Budin 2024; Bernardo et al. 2023).
Turnover intention is typically measured through employee self-reports, such as survey questions asking whether
they are considering or planning to leave their job, or if they are actively seeking alternative employment (Lazzari
et al. 2022; Park et al. 2024). High turnover intention is a chronic challenge in the IT-BPM sector, leading to
disruptions in project timelines, increased recruitment and training costs, and loss of organizational knowledge
(Narwaria et al. 2024; Bernardo et al. 2023; Özkan 2021). Factors influencing turnover intention among IT-BPM
employees include job satisfaction, compensation, career growth opportunities, work-life balance, organizational
culture, and management support (Farooq et al. 2022; Narwaria et al. 2024; Kanchana and Jayathilaka 2023;
Budin 2024; Bernardo et al. 2023; Özkan 2021).
Page 3220
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
Turnover intention is used as a proxy for actual turnover in research and HR practice, as it allows organizations
to identify at-risk employees and intervene before they actually leave (Lazzari et al. 2022,; Narwaria et al. 2024;
Budin 2024; Hur 2024).
Flexible Working Arrangements
In general, flexible working arrangements (FWAs) are employer-provided benefits that give employees some
level of control over when and where they work, outside the standard workday. In the IT-BPM industry, FWAs
typically include flex time, flex place and combined flexibility Flextime pertains to flexibility in work schedules
(choosing start/end times). Moreover, flexplace refers to flexibility in work location (remote work,
telecommuting, hybrid models) and Combined Flexibility pertains to both schedule and location flexibility,
sometimes negotiated individually (i-deals) (Harrop et al. 2025; Soga et al. 2022). FWAs are designed to allow
employees autonomy to complete their work outside the traditional temporal (time) and spatial (place)
boundaries of standard office jobs (Harrop et al. 2025; Soga et al. 2022).
Multiple studies and reviews find that FWAs such as flextime, telecommuting, and hybrid schedules—are
associated with lower turnover intentions in IT and high-tech industries, including IT-BPM. This effect is largely
attributed to increased job satisfaction, organizational commitment, autonomy, and improved work-life balance
(Yang et al. 2024; Hemavathi and T. 2023; Berber et al. 2022; Murti and Martdianty 2021; Gašić and Berber
2023; Permatasari and Setiyawan 2024; Rosita et al. 2024). Some studies note that the impact of FWAs can vary.
For instance, the benefits are strongest when employees have high job independence or when FWAs are well-
aligned with job roles. In some cases, if FWAs increase work-family conflict (e.g., poorly managed
telecommuting), they may not reduce turnover intentions or could even increase them (Haines et al. 2024; Tsen
et al. 2021; Tsen et al. 2021). Nevertheless, the study hypothesized that:
H1: flexible working arrangements negatively and significantly impact turnover intentions.
Job Control
Job control refers to the degree of autonomy and discretion employees have over how, when, and where they
perform their work. Research shows that FWAs are strongly associated with greater job autonomy and job
control, as they allow employees to tailor their work schedules and locations to fit their needs (Harrop et al.
2025; Haines et al. 2024). The positive impact on job control is a key mechanism by which FWAs improve job
satisfaction, organizational commitment, and work-life balance (Harrop et al. 2025; Haines et al. 2024) The
benefits are most pronounced when both flextime and flexplace are available, and when employees have genuine
discretion in using these arrangements (Harrop et al. 2025). With this, the study hypothesized that:
H2: Flexible working arrangements positively and significantly impact job control.
Research in the IT-BPM industry suggests that job control generally has a negative impact on turnover intentions,
meaning higher job control tends to reduce employees' intentions to leave. Job control enhances employees’
sense of autonomy and empowerment, which fosters organizational commitment and job satisfaction, thereby
lowering turnover intentions (Chu et al. 2022; Chen et al. 2023). For example, job control can substitute for
leadership influence by increasing employees’ felt obligation for constructive change, which reduces turnover
intention (Chu et al. 2022). Overall, Social Exchange Theory supports that when employees perceive job control
as a valuable resource provided by the organization, they reciprocate with lower turnover intentions due to
perceived fairness and support. Thus, in IT-BPM, enhancing job control is an effective strategy to reduce
turnover intentions by strengthening employees’ psychological attachment to their work and organization (Chu
et al. 2022; Chen et al. 2023). With this, the study hypothesized that:
H3: Job control has a negative and significant impact on turnover intention
Flexible working arrangements (FWAs) in the IT-BPM sector reduce turnover intentions primarily through
increasing job control, which acts as a key mediating mechanism consistent with Social Exchange Theory. FWAs
enhance employees’ perceived job control and work engagement, which in turn lower turnover intentions by
Page 3221
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
fostering a sense of autonomy and reciprocity toward the organization (Haines et al. 2024). With this, the study
hypothesized that
H4: Job control mediate the negative impact of flexible working arrangements on turnover intention.
Theoretical Framework
The study was anchored on the Social Exchange Theory which explains that negative impacts of flexible working
arrangements and job control on turnover intentions in IT-BPM arise when employees perceive an imbalance in
the exchange relationship.
Social Exchange Theory posits that when organizations offer positive practices—like flexible working
arrangements (FWAs) and job control—employees feel valued and reciprocate with loyalty and reduced turnover
intentions. However, if FWAs or job control are poorly implemented, leading to increased work-family conflict,
role ambiguity, or lack of support, employees may perceive the exchange as unfair or burdensome, increasing
their intention to leave (Berber et al. 2022; Yang et al. 2024; Tsen et al. 2021). FWAs and job control can enhance
organizational commitment and job satisfaction, which mediate the reduction in turnover intentions. If these
arrangements fail to deliver perceived benefits or create new stressors, the expected positive reciprocity is
undermined, and turnover intentions may rise (Berber et al. 2022; Yang et al. 2024; George and Poluru 2024).
When flexible arrangements blur work-life boundaries or create isolation, employees may feel the costs outweigh
the benefits, disrupting the social exchange. This can result in higher turnover intentions, especially if employees
do not feel adequately supported or if job control leads to increased responsibility without corresponding
resources (Tsen et al. 2021; Yang et al. 2024). As it is, Social Exchange Theory explains that the effectiveness
of flexible working arrangements and job control in reducing turnover intentions depends on employees’
perceptions of fairness and support. When these practices are seen as genuine organizational investments, they
foster loyalty and when perceived as burdensome or inequitable, they can backfire and increase turnover
intentions.
Conceptual Framework
Figure 1. Research Model
Note. Anchored on the Social Exchange Theory, the study identified Flexible working arrangement as
independent variable, job control as mediating variable and turnover intention as dependent variable.
METHODS
The study employed a quantitative, correlational, and cross-sectional research design to examine the effect of
flexible working arrangements on turnover intentions, with job control as a mediating variable. Data were
collected from October 17 to November 4, 2025, using an online survey administered to employees of IT–BPM
firms in Cebu City. A total of 589 respondents were selected through systematic sampling based on an unknown
population, exceeding the minimum required sample of 385. All constructs were measured using a 7-point Likert
scale ranging from 1 (Strongly Disagree) to 7 (Strongly Agree), and formal permissions were secured for the
Page 3222
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
adaptation of scale items to ensure ethical compliance. Structural Equation Modelling was conducted using
SmartPLS 4.0.
Table 1. Mean Range, Interpretation and Verbal Descriptions
Scale
Range
Interpretation
Verbale Description
7
6.50 – 7.00
Very High
Respondents expressed very high agreement, indicating a highly
favorable and strongly positive perception of the variable
measured.
6
5.50 – 6.49
High
Respondents generally agreed, reflecting a clearly positive
perception or condition related to the variable.
5
4.50 – 5.49
Moderately
High
Respondents slightly to moderately agreed, suggesting a
favorable perception, although not very strong.
4
3.50 – 4.49
Neutral
Respondents showed neither agreement nor disagreement,
indicating neutrality toward the variable.
3
2.50 – 3.49
Moderately
Low
Respondents slightly disagreed, signifying a moderately
unfavorable perception or limited support.
2
1.50 – 2.49
Low
Respondents generally disagreed, reflecting a negative
perception or dissatisfaction with the variable.
1
1.00 – 1.49
Very Low
Respondents strongly disagreed, indicating a very low level of
agreement and a highly unfavorable perception of the variable.
Prior to estimating structural relationships, the measurement model was assessed to establish indicator reliability
and construct validity. Following current guidelines by Hair et al. (2021–2024), the evaluation included internal
consistency reliability using Cronbach’s alpha and composite reliability (acceptable ≥0.70), convergent validity
via the Average Variance Extracted (≥0.50), and discriminant validity using the HTMT criterion (<0.85, and up
to 0.90 for closely related constructs). After confirming that all constructs met the required standards, the
structural model was examined to determine the significance and predictive relevance of the hypothesized
relationships. This assessment involved testing for multicollinearity through VIF values (all within acceptable
≤3.0 to ≤5.0 thresholds), estimating path coefficients and their significance using bootstrapping, and evaluating
predictive accuracy using the coefficient of determination (R² and adjusted R²). Effect size (f²) was assessed
using the benchmarks of 0.02 (small), 0.15 (medium), and 0.35 (large). Model fit was evaluated using the
Standardized Root Mean Square Residual (SRMR), with results indicating an acceptable fit for both the saturated
model (0.066, ≤0.08) and estimated model (0.081, ≤0.10). Collectively, these procedures ensured that the model
demonstrated adequate measurement quality and empirical support for the theorized structural relationships.
RESULTS AND DISCUSSION
This section presents the empirical results of the study based on the data collected from employees in the IT-
BPM sector. The analyses include descriptive statistics, assessment of the measurement model, and evaluation
of the structural model to test the hypothesized relationships among flexible working arrangements, job control,
and turnover intention. Descriptive statistics summarize the respondents’ demographic characteristics and
provide an overview of the central tendencies of the study variables. The measurement model results establish
the reliability, validity, and measurement quality of the constructs, ensuring that the indicators appropriately
capture the theoretical dimensions they represent. The structural model then examines the direct, indirect, and
mediating effects specified in the research framework. Collectively, these findings provide a comprehensive
Page 3223
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
understanding of how flexible work practices and job control influence turnover intention among IT-BPM
employees.
Table 2. Demographics of the Respondents
Demographic
n
%
Sex
274
47%
315
53%
Parental Status
403
68%
186
32%
Job Level
88
15%
501
85%
The demographic profile shows that the sample consists of slightly more female (53%) than male (47%)
respondents, indicating a relatively balanced gender distribution. Most participants (68%) reported having no
children, while 32% had at least one child, suggesting that a majority of the workforce is either single or without
parental responsibilities. In terms of job level, the sample is predominantly composed of non-managerial
employees (85%), with only 15% occupying managerial roles. This distribution reflects the structure of typical
IT-BPM organizations, where operational and technical roles make up the largest portion of the workforce.
Overall, the demographic characteristics provide a representative overview of employees in the IT-BPM sector,
with implications for understanding differences in flexibility needs, job control, and turnover intentions across
employee groups.
Table 3. Level of Variables
Construct
Mean
Standard Deviation
Interpretation
Flexible Working
Arrangements
4.05
2.13
Neutral
Job Control
4.94
1.76
Moderately
High
Turnover Intention
3.50
1.86
Neutral
The respondents reported neutral perceptions toward both flexible work arrangements (FWA) and turnover
intention (TI). The mean score for FWA (M = 4.05) indicates that employees neither agreed nor disagreed about
the availability or effectiveness of flexible work practices, suggesting that FWAs may be present but are not
strongly perceived or consistently implemented within the organization. Similarly, turnover intention (M = 3.50)
also falls within the neutral range, reflecting that employees are not actively planning to leave but are also not
strongly committed to staying, indicating a state of uncertainty regarding their future with the organization. In
contrast, job control (JC) received a moderately high mean score (M = 4.94), showing that employees generally
feel a favorable level of autonomy and discretion in how they perform their work. This suggests that while job
control is perceived positively, the neutral perceptions of both FWAs and turnover intention may indicate
opportunities for organizations to further strengthen flexibility practices to enhance employee retention and
overall work experience. Having established the overall levels of the study variables, the analysis proceeds to
the measurement model assessment to evaluate the reliability and validity of the constructs before testing the
structural relationships.
Page 3224
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
Measurement Model Assessment
Before estimating the structural relationships, the measurement model was rigorously assessed to ensure that the
constructs were represented with adequate reliability and validity. Establishing measurement quality is a
necessary prerequisite in PLS-SEM, as the accuracy of the structural paths depends on the soundness of the
underlying indicators. In this study, all indicators were specified as reflective, implying that they function as
observable manifestations of their respective latent variables. The assessment followed the standard PLS-SEM
procedure, which is appropriate for predictive, complex models and datasets that do not conform to normality
assumptions. The evaluation focused on internal consistency reliability, convergent validity, and discriminant
validity to confirm that the measurement model met accepted psychometric standards prior to proceeding with
hypothesis testing.
Table 2. Internal Consistency and Validity
Variables
Cronbach's
alpha
Composite
reliability (rho_a)
Composite
reliability (rho_c)
Average variance
extracted (AVE)
FWA
0.936
0.938
0.951
0.796
JC
0.902
0.921
0.928
0.723
TI
0.897
0.928
0.922
0.704
All constructs showed high internal consistency, with Cronbach’s alpha, Dijkstra–Henseler’s rho (ρₐ), and
composite reliability (ρc) values exceeding the recommended threshold of 0.70. Specifically, the reliability
coefficients for Flexible Working Arrangements (FWA), Job Control (JC), and Turnover Intentions (TI) ranged
from 0.897 to 0.951, indicating strong internal consistency across all indicators. Convergent validity was also
established, as all constructs reported Average Variance Extracted (AVE) values above 0.70, demonstrating that
the items adequately captured their respective latent constructs.
Table 3. Outer Loading Relevance Testing
Indicators
FWA
JC
TI
FWA1
0.888
FWA2
0.885
FWA3
0.908
FWA4
0.910
FWA5
0.869
JC1
0.688
JC2
0.801
JC4
0.923
JC5
0.904
JC3
0.912
T1
0.811
T2
0.863
Page 3225
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
Indicators
FWA
JC
TI
T3
0.859
T4
0.811
T5
0.849
Most indicators demonstrated satisfactory outer loadings, with values above the recommended 0.70 threshold.
All FWA items (0.869 to 0.910) and all TI items (0.811 to 0.863) exhibited strong loadings. For JC, four
indicators (JC2, JC3, JC4, JC5) showed high loadings above 0.80, while JC1 presented a marginally lower
loading at 0.688, slightly below the preferred benchmark. Although this value does not necessitate removal, it
suggests comparatively weaker contribution and may be considered for refinement in future research. Overall,
the indicators support adequate construct representation.
Table 4. Discriminant Validity
Variables
FWA
JC
TI
Heterotrait-monotrait ratio (HTMT) – Matrix
JC
0.576
TI
0.123
0.161
Fornell Locker Criterion
FWA
0.892
JC
0.543
0.850
TI
0.116
-0.148
0.839
Discriminant validity was examined using the Heterotrait–Monotrait Ratio (HTMT) and the Fornell–Larcker
criterion. HTMT values were substantially below the conservative threshold of 0.85, with construct pairs ranging
from 0.123 to 0.576. These results indicate that the constructs are empirically distinct. The Fornell–Larcker
criterion further confirmed discriminant validity, as the square roots of the AVE values for FWA (0.892), JC
(0.850), and TI (0.839) exceeded the correlations with other constructs. Together, these tests confirm that each
construct measures a unique conceptual domain within the model.
Structural Model Assessment
After establishing the reliability and validity of the measurement model, the next step in PLS-SEM involves
evaluating the structural model to determine the strength, significance, and relevance of the hypothesized
relationships among the latent constructs. Structural model assessment in SmartPLS 4.0 follows the updated
guidelines of Hair et al. (2021, 2022, 2024), focusing on key criteria such as collinearity diagnostics, path
coefficients, coefficient of determination (R²), effect sizes (f²) and the significance of direct, indirect, and
moderating effects through bootstrapping. This evaluation allows researchers to determine how well the model
explains the variance in the dependent variables and whether the theoretical assumptions are empirically
supported.
Table 5. Collinearity Diagnostics
Indicators
VIF
FWA1
3.264
Page 3226
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
FWA2
3.282
FWA3
3.892
FWA4
3.687
FWA5
2.650
JC1
1.566
JC2
1.866
JC3
4.761
JC4
5.431
JC5
3.660
T1
2.313
T2
2.171
T3
2.476
T4
2.157
T5
2.224
Variance Inflation Factor (VIF) values were used to assess collinearity among indicators. All FWA and TI
indicators exhibited VIF values between 2.157 and 3.892, falling well within acceptable thresholds. For JC, three
indicators (JC1, JC2, JC5) reported low VIF values, while JC3 (4.761) and JC4 (5.431) showed comparatively
higher values, with JC4 slightly exceeding the recommended upper limit of 5. Despite this, the collinearity levels
remain within a range that does not compromise the validity of the measurement model. Overall,
multicollinearity does not pose a substantive concern.
Figure 2. Structural Model
Page 3227
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
Table 6. Explanatory Power
Indicators
R-square
R-square adjusted
JC
0.295
0.293
TI
0.077
0.074
The explanatory power of the model varies across the endogenous constructs. Job Control (JC) obtained an R²
value of 0.295 (adjusted R² = 0.293), which meets the weak-to-moderate threshold in PLS-SEM, where values
of 0.25, 0.50, and 0.75 respectively indicate weak, moderate, and substantial explanatory power (Hair et al.,
2019). This suggests that approximately 29.5% of the variance in JC is explained by its predictors, representing
a meaningful level of explanatory strength.
Turnover Intention (TI) recorded an R² value of 0.077 (adjusted R² = 0.074), which falls below the 0.10 level
typically considered weak in PLS-SEM. However, when interpreting R² values in the turnover intention
literature, prior studies commonly report R² values ranging from 0.10 to 0.32, and these levels are still considered
theoretically meaningful and empirically acceptable. For instance, Alawi (2017) found that transformational
leadership explained 17% of turnover intention (R² = 0.17), while Kusuma et al. (2023) reported an R² of 0.126
for a model including leadership, talent management, and motivation—both of which were deemed valuable
contributions in understanding turnover intention. Additionally, several studies in similar fields report R² values
between 0.27 and 0.33 (Ntseke et al., 2022; Utami, 2025), demonstrating that modest explanatory power is
common in behavioral and organizational models predicting turnover intentions. Given this context, the current
study’s R² value for TI, although lower, remains consistent with the broader empirical pattern indicating that
turnover intention is influenced by multiple unobserved or external factors. As such, the model still offers
important explanatory insights, and future research may incorporate additional predictors to further strengthen
the model’s explanatory capacity.
Table 7 . Predictive Power
Indicators
JC
TI
FWA
0.418
0.059
JC
0.069
The predictive relevance of the model was assessed using cross-validated redundancy values. For Job Control
(JC), the model yields a Q² value of 0.418, which exceeds the recommended threshold of 0.25 for medium
predictive power and falls well above the 0.50 benchmark for large predictive relevance in behavioral research
contexts. This indicates that the predictors in the model provide strong predictive accuracy for JC. For Turnover
Intentions (TI), the Q² values associated with its predictors are lower, with FWA showing 0.059 and JC showing
0.069. These values fall just above the minimum acceptable level of Q² > 0, which indicates small but meaningful
predictive relevance. Although modest, such levels are consistent with prior turnover intention research, where
prediction is typically influenced by multiple external or unobserved variables. Overall, the model demonstrates
strong predictive power for JC and small yet meaningful predictive relevance for TI, aligning with patterns
commonly reported in employee turnover studies.
Table 8. Model Fit Assessment
Criteria
Saturated model
Estimated model
SRMR
0.053
0.053
Page 3228
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
d_ULS
0.333
0.333
d_G
0.143
0.143
Chi-square
506.975
506.975
NFI
0.925
0.925
The model fit indices indicate that the structural model meets the recommended thresholds for acceptable model
fit in PLS-SEM. The Standardized Root Mean Square Residual (SRMR) value of 0.053 for both the saturated
and estimated models is below the recommended cutoff of 0.08, indicating a good fit between the model-implied
and observed correlations. Similarly, the discrepancy measures d_ULS (0.333) and d_G (0.143) fall within
acceptable ranges, suggesting that the empirical data align well with the model’s predictions The Chi-square
value (χ² = 506.975) is reported for completeness but is not typically emphasized in PLS-SEM due to its
sensitivity to sample size. The Normed Fit Index (NFI) value of 0.925 exceeds the 0.90 threshold, indicating
strong incremental model fit. Collectively, these indices confirm that the proposed model demonstrates an
acceptable and robust fit to the data
Table 9. Direct Effects
Hypothesis
Mean
STDEV
P values
Interpretation
Decision
H1: FWA -> TI
0.279
0.045
0.000
Significant
Not supported
H2: FWA-> JC
0.543
0.033
0.000
Significant
Supported
H3: JC -> TI
-0.299
0.045
0.000
Significant
Supported
The results show that flexible working arrangements (FWA) have a significant effect on turnover intention (TI),
but in the opposite direction of what was hypothesized. Although the relationship is statistically significant, the
positive coefficient indicates that FWAs do not directly reduce turnover intention in this sample, leading to the
rejection of H1. This finding aligns with recent literature suggesting that FWAs may not always lower turnover
intentions and can sometimes increase role ambiguity, work–life conflict, or feelings of isolation if not well
supported (Haines et al., 2024; Tsen et al., 2021). Studies in IT-BPM and high-tech environments similarly note
that FWAs only reduce turnover when implemented effectively and accompanied by managerial support and
clear boundaries (Harrop et al., 2025; Soga et al., 2022). In contrast, the significant positive relationship between
FWA and job control (JC) supports H2, confirming that FWAs enhance employees’ discretion over how, when,
and where they work. This is consistent with extensive evidence showing that FWAs are strong drivers of
perceived autonomy and job control (Harrop et al., 2025; Haines et al., 2024). In the IT-BPM industry, where
work tasks are highly structured but often asynchronous, FWAs expand employees’ sense of control over their
work routines and scheduling, reinforcing findings from earlier research that flexibility enhances autonomy and
work–life fit (Berber et al., 2022; Yang et al., 2024).
Finally, the significant negative effect of job control on turnover intention supports H3. This means that higher
job control reduces employees’ intention to leave, which aligns with prior research showing that autonomy,
discretion, and control over work processes enhance engagement, job satisfaction, and organizational
commitment (Chu et al., 2022; Chen et al., 2023). Job control has been shown to buffer stress, improve well-
being, and strengthen employees’ psychological attachment to their organization—key mechanisms consistently
associated with lower turnover intentions in IT-BPM and related knowledge-based industries (Narwaria et al.,
2024; Bernardo et al., 2023).Overall, the findings suggest that FWAs influence turnover intention indirectly
rather than directly, working primarily through enhanced job control. This pattern reflects broader evidence in
the literature indicating that the success of FWAs in reducing turnover depends on employees’ perceptions of
autonomy, fairness, and managerial support (Berber et al., 2022; Yang et al., 2024). When FWAs improve job
Page 3229
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
control, turnover intention decreases; when they create ambiguity or are poorly implemented, they may fail to
reduce turnover or even heighten employees’ intention to leave.
Table 10. Mediation Analysis
Hypothesis
Direct
Indirect
Total
Interpretation
Decision
β
p
β
p
β
p
H4: FWA -> TI
0.279
0.000
-
0.163
0.000
2.597
0.009
Competitive
Supported
The mediation analysis shows a competitive mediation effect for H4, indicating that flexible working
arrangements (FWA) exert both a positive direct effect and a negative indirect effect (via job control) on turnover
intention (TI). The direct path from FWA to TI is positive and significant (β = 0.279, p = 0.000), suggesting that
FWAs, on their own, may increase turnover intention. This aligns with studies showing that poorly supported or
inconsistently implemented FWAs can create work–life boundary conflicts, role ambiguity, or social isolation,
which may heighten employees’ intention to leave (Haines et al., 2024; Tsen et al., 2021). However, the indirect
path from FWA → JC → TI is negative and significant (β = –0.163, p = 0.000), confirming that FWAs reduce
turnover intention when they effectively increase job control.
This supports extensive literature showing that job autonomy and discretion are central mechanisms through
which FWAs improve well-being, organizational commitment, and retention in IT-BPM and knowledge-
intensive environments (Harrop et al., 2025; Chu et al., 2022; Chen et al., 2023). When employees perceive that
FWAs genuinely enhance their autonomy and control over work timing and location, turnover intention tends to
decline. The presence of competitive mediation means that the direct and indirect effects operate in opposite
directions. As observed in recent studies, FWAs can have mixed outcomes: they may increase autonomy and
reduce burnout for some employees, yet create stressors or unmet expectations for others (Yang et al., 2024;
Berber et al., 2022; George & Poluru, 2024). In this study, the total effect remains statistically significant,
indicating that job control plays a meaningful role in counterbalancing the potential negative consequences of
FWAs. Overall, the findings support H4 and reinforce Social Exchange Theory, which posits that employees
respond positively when FWAs translate into genuine autonomy and support—but may react negatively if the
flexibility increases workload pressures or lacks organizational support. Thus, FWAs reduce turnover intention
only when paired with sufficient job control, making job control a critical pathway through which FWAs
influence employee retention.
CONCLUSION
The study examined the impact of flexible working arrangements (FWAs) on turnover intention in the Philippine
IT-BPM sector, with job control as a mediating mechanism grounded in Social Exchange Theory. Using data
from 589 employees and a structural equation model, the findings reveal a complex and nuanced relationship
between workplace flexibility, autonomy, and employees’ intention to stay or leave. Although FWAs
significantly increased job control, and job control significantly reduced turnover intention, FWAs alone did not
directly lower turnover intention. Instead, FWAs demonstrated a positive direct association with turnover
intention, indicating that flexibility—when poorly supported or inconsistently implemented—may inadvertently
heighten employees’ inclination to leave. The significant negative indirect effect of FWAs on turnover intention
via job control confirms that job control is a crucial pathway through which flexibility becomes beneficial. Thus,
job control operates as a competitive mediator, counteracting the potentially adverse direct effects of FWAs on
employee retention.
These results contribute to theory by clarifying the conditional nature of the FWA–turnover link and reinforcing
Social Exchange Theory’s emphasis on employee perceptions of fairness, support, and reciprocal value. FWAs
do not uniformly generate positive outcomes; their success depends on whether they translate into genuine
autonomy rather than additional burden, ambiguity, or work–life boundary conflict. Empirically, the study
Page 3230
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
advances IT-BPM–specific evidence from a developing country context—an area where scholarship remains
limited—by demonstrating that job control is a key explanatory mechanism underlying employees’ reactions to
flexible work practices.
From a practical perspective, the findings underscore that flexibility initiatives must be paired with supportive
managerial systems, clear implementation guidelines, and mechanisms that protect employee autonomy. Merely
offering FWAs is insufficient; organizations must ensure consistent application, communication clarity, and
adequate resources so that flexibility enhances rather than undermines employee well-being. Strengthening job
control—through participatory decision-making, autonomy in scheduling, and empowerment practices—can
substantially reduce turnover intention and improve retention outcomes in IT-BPM firms.
Future research may expand the model by incorporating additional psychological and organizational variables
such as work engagement, organizational support, job embeddedness, or perceived job independence, which
prior studies identify as salient predictors of turnover. Longitudinal designs may also clarify how employees’
perceptions of FWAs and job control evolve over time. Overall, this study highlights that FWAs are most
effective in reducing turnover intention when embedded within a broader system that reinforces autonomy,
fairness, and supportive work conditions, thereby offering actionable guidance for IT-BPM organizations
seeking to enhance employee retention in increasingly flexible work environments
REFERENCES
1. Al Bernardo, A. L., Lacap, J. P., Talon, C., Bolante, P., Aumentado, Z. J., Capalao, M. W., Llado, M. C.
E., & Dubrea, E. M. (2023). The Mediating Role of Employee Engagement on the Link Between Person–
Organization Fit and Turnover Intention: Evidence from BPO Companies in the Philippines. Journal of
Entrepreneurship and Business.
https://doi.org/10.17687/jeb.v11i2.1025
2. Ali, A. A. M., Chen, X., Hussain, W., Jingzu, G., Yang, Q., & Al Shami, S. S. A. (2023). Envisaging the
Job Satisfaction and Turnover Intention Among the Young Workforce: Evidence from an Emerging
Economy. PLOS ONE, 18. https://doi.org/10.1371/journal.pone.0287284
3. Berber, N., Gašić, D., Katić, I., & Borocki, J. (2022). The Mediating Role of Job Satisfaction in the
Relationship Between FWAs and Turnover Intentions. Sustainability.
https://doi.org/10.3390/su14084502
4. Berber, N., & Gašić, D. (2023). The Mediating Role of Employee Engagement in the Relationship
Between Flexible Work Arrangements and Turnover Intentions. Behavioral Sciences, 13.
https://doi.org/10.3390/bs13020131
5. Budin, K. A. G. (2024). Determinants of Employees’ Turnover Intention: A Conceptual Paper. ISC-
BEAM. https://doi.org/10.21009/isc-beam.011.65
6. Chu, X., Ding, H., Zhang, L., & Li, Z. (2022). Strengths-Based Leadership and Turnover Intention.
Frontiers in Psychology, 13. https://doi.org/10.3389/fpsyg.2022.786551
7. Farooq, H., Janjua, U., Madni, T. M., Waheed, A., Zareei, M., & Alanazi, F. (2022). Identification and
Analysis of Factors Influencing Turnover Intention of Pakistan IT Professionals. IEEE Access.
https://doi.org/10.1109/access.2022.3181753
8. Gašić, D., & Berber, N. (2023). Employee Engagement as a Mediator Between Flexible Work
Arrangements and Turnover Intentions. Behavioral Sciences, 13. https://doi.org/10.3390/bs13020131
9. George, J. B., & Poluru, N. V. (2024). Exploring the Impact of Flexible Work Arrangements on Turnover
Intention. Journal of Accounting, Business and Management.
https://doi.org/10.31966/jabminternational.v32i1.1467
10. Harrop, N., Jiang, L., & Overall, N. (2025). A Meta‐Analysis of Antecedents and Outcomes of Flexible
Working Arrangements. Journal of Organizational Behavior. https://doi.org/10.1002/job.2896
11. Haines, V. Y., Guerrero, S., & Marchand, A. (2024). Flexible Work Arrangements and Employee
Turnover Intentions: Contrasting Pathways. The International Journal of Human Resource Management,
35, 1970–1995.
https://doi.org/10.1080/09585192.2024.2323510
12. Itesa, U., & Burchell, J. M. (2025). Leadership Styles and Flexible Work Arrangements as Determinants
of Turnover Intention for IT Professionals. Business Management Research and Applications.
https://doi.org/10.54093/bmra.v4i1.8194
Page 3231
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
13. Kanchana, L., & Jayathilaka, R. (2023). Factors Impacting Employee Turnover Intentions Among
Professionals in Sri Lankan Startups. PLOS ONE, 18.
https://doi.org/10.1371/journal.pone.0281729
14. Kurata, Y., Belisario, K. V. S., Mescala, M. M. D., & Valenzuela, R. L. L. (2023). Factors Affecting
Employee Retention in the Philippine BPO Industry: Integrating Job Embeddedness Theory. Proceedings
of the International Conference on Industrial Engineering and Operations Management.
https://doi.org/10.46254/an13.20230674
15. Lazzari, M., Saiz-Alvarez, J. M., & Ruggieri, S. (2022). Predicting and Explaining Employee Turnover
Intention. International Journal of Data Science and Analytics, 14, 279–292.
https://doi.org/10.1007/s41060-022-00329-w
16. Murti, A., & Martdianty, F. (2021). Workload, Work–Life Balance, and Flexible Working in Relation to
Turnover Intention of IT Workers. Contemporary Research on Business and Management.
https://doi.org/10.1201/9781003196013-18
17. Narwaria, T., Dwivedi, R., Vashisht, A., Mishra, A., Saxena, P., Gupta, V., Singhal, A., & Raghuwanshi,
S. (2024). A Bibliometric Analysis of Employee Turnover Intention Among IT Sector Employees.
International Review of Management and Marketing.
https://doi.org/10.32479/irmm.17067
18. Patil, A. H. (2025). The IT Workforce Dilemma: Unraveling Turnover in the Indian Tech Sphere.
International Journal of Innovative Science and Research Technology.
https://doi.org/10.38124/ijisrt/25may535
19. Permatasari, D. W., & Setiyawan, S. (2024). The Effect of Flexible Work Arrangement and Work–Life
Balance on Turnover Intention of Y-Generation Employees. International Journal of Latest Technology
in Engineering, Management & Applied Science.
https://doi.org/10.51583/ijltemas.2024.130201
20. Rahman, H., Jayashree, S., Agamudai, C., & Malarvizhi, N. (2023). Impact of HRM Practices on
Turnover Intention Through Employee Loyalty in the Bangladesh ICT Industry. Russian Law Journal.
https://doi.org/10.52783/rlj.v11i9s.1581
21. Repaso, J. A. A., Capariño, E. T., Hermogenes, M. G. G., & Perez, J. G. (2022). Determining Factors
Resulting to Employee Attrition Using Data Mining Techniques. International Journal of Education and
Management Engineering. https://doi.org/10.5815/ijeme.2022.03.03
22. Rosita, F., Noermijati, N., Margono, M., & Susilowati, C. (2024). The Role of FWA, Job Embeddedness,
and Work–Life Balance in Reducing Turnover Intention. Journal of Community Development in Asia.
https://doi.org/10.32535/jcda.v7i2.2987
23. Salunkhe, H. A., Jain, D., Hinge, P., & Boralkar, M. (2024). Impact of HR Practices on Work
Engagement and Turnover Intention in IT Companies. SA Journal of Human Resource Management.
https://doi.org/10.4102/sajhrm.v22i0.2723
24. Seneviratna, D., Chathuranga, L., Kithulwatta, W. M. C. J. T., & Rathnayaka, R. K. T. (2024). Impact of
Soft Productivity Factors on Employee Turnover Among IT Employees. Sri Lankan Journal of Applied
Statistics. https://doi.org/10.4038/sljas.v25i1.8113
25. Shilpakar, N., Giri, B., & Pokhrel, S. (2024). Flexible Working Arrangements and Employee Turnover
Intention: Mediating Role of Employee Engagement. SAIM Journal of Social Science and Technology.
https://doi.org/10.70320/sacm.2024.v01i01.003
26. Sidike, R., & Zulkifly, N. A. (2025). A Qualitative Exploration of Employee Turnover in a Malaysian
BPO Company. International Journal of Academic Research in Business and Social Sciences.
https://doi.org/10.6007/ijarbss/v15-i2/24674
27. Soga, L. J., Bolade-Ogunfodun, Y., Mariani, M. M., Nasr, R., & Laker, B. (2022). Unmasking the Other
Face of Flexible Working Practices: A Systematic Literature Review. Journal of Business Research.
https://doi.org/10.1016/j.jbusres.2022.01.024
28. Tsen, M. K., Gu, M., Tan, C. M., & Goh, S. (2021). Does Flexible Work Arrangements Decrease or
Increase Turnover Intention? International Journal of Sociology and Social Policy.
https://doi.org/10.1108/ijssp-08-2021-0196
29. Tsen, M. K., Gu, M., Tan, C. M., & Goh, S. (2021). Effect of Flexible Work Arrangements on Turnover
Intention: Does Job Independence Matter? International Journal of Sociology, 51, 451–472.
https://doi.org/10.1080/00207659.2021.1925409
30. W. Sethar, Channar, H., & Jatoi, S. A. (2022). Turnover Intentions of IT Professionals: Case of Software
Houses in Pakistan. Pakistan Journal of International Affairs. https://doi.org/10.52337/pjia.v5i3.601
Page 3232
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue XIV November 2025 | Special Issue on Management
www.rsisinternational.org
31. Yang, J., Arshad, M. A. B., & Zhao, M. (2024). Flexible Work Arrangements and Employee Turnover
Intentions: A Comprehensive Review. International Journal of Academic Research in Business and
Social Sciences.
https://doi.org/10.6007/ijarbss/v14-i12/24030
32. Yang, Y.-T., Feng, Y., & Jeong, S.-P. (2024). Developing an Advanced Prediction Model for Employee
Turnover Intention Using Machine Learning. Scientific Reports, 14. https://doi.org/10.1038/s41598-023-
50593-4