Work-From-Home Productivity and Job Satisfaction: A Double-Layered Moderated Mediation Model

: This study was conducted against the backdrop of the global-pandemic-induced change in work climate resulting in employees working from home (WFH). The current study investigated the relationship between work-from-home productivity (WFHP) and job satisfaction. Using a structured survey instrument, data were collected from 1158 respondents from a developing country, India. After checking the psychometric properties of the measures using the LISREL software of structural equation modeling (SEM), data were analyzed using Hayes’s PROCESS macros. The ﬁndings indicate that: (i) WFHP positively predicts (a) job satisfaction and (b) work–life balance (WLB), (ii) WLB positively predicts job satisfaction, and (iii) WLB mediates the relationship between WFHP and job satisfaction. The results also support that (i) work stress moderates the relationship between WFHP and WLB, (ii) work–personal life enhancement (WPLE) (second moderator) moderates the moderated relationship between WFHP and work stress (ﬁrst moderator) in inﬂuencing WLB, (iii) emotional exhaustion moderates the relationship between WLB and job satisfaction, and (iv) WPLE (second moderator) moderates the relationship between WLB and emotional exhaustion (ﬁrst moderator) in inﬂuencing job satisfaction. The ﬁrst three-way interaction between WFHP, work stress, and WLB and the second three-way interaction between WLB, emotional exhaustion, and job satisfaction have been investigated for the ﬁrst time, to the best of our knowledge. The conceptual double-layered moderated mediation model is a novel idea, and the results signiﬁcantly contribute to the literature on WLB and job satisfaction. The implications for theory and practice are discussed.


Introduction
The unprecedented COVID-19 global pandemic has resulted in a phenomenal metamorphosis in the lives of people and organizations worldwide [1][2][3]. To prevent the spread of the virus, governments imposed frequent lockdowns and social distancing. Organizations did not have any alternative but to request employees to work remotely [4,5]. This study was conducted against the backdrop of this global-pandemic-induced WFH of employees in a developing country, India.
The separation of work from family has remained an exciting topic for organizational behavior and psychology researchers for over three decades [6][7][8][9][10]. Often, employees struggle to balance their work demands and personal priorities and preferences, resulting in a work-life imbalance. To help employees effectively balance work and life, organizations offer flexible working hours, remote working, or work from home (WFH) [11][12][13]. Several researchers documented that, in order to maintain a work-life balance (WLB), remote

Emotional Exhaustion
The pioneers of the construct [43], contend that emotional exhaustion is an essential component of burnout. When employees are overburdened by work, they feel exhausted and realize that their energies are burning out. There has been substantial evidence that emotional exhaustion has deleterious consequences: a low performance, high turnover, and low commitment [44][45][46][47]. There is consensus among organizational scholars that the urgency of performance and pressures results in emotional exhaustion, adversely affecting satisfaction [48][49][50][51].

Work Stress
Also known as job stress, workplace stress comes from the psychosocial work environment, Sustainability 2022, 14, 11179 4 of 29 increased workload, role ambiguity, role conflict, and lack of supervisor support [52][53][54]. Stress also may stem from incompatibility of the worker ability and work demands, known as a person-environment misfit [22]. Sometimes, stressful situations from the environment or uncertainties do not allow employees to perform their duties [55,56]. An example of stress created by the environment is the global-pandemic-induced stress, where individuals undergo psychological stress to maintain health and keep themselves from contacting the virus.
Extant research has identified the antecedents of work stress: job conflict, role conflict [57,58], and non-collegiality and increasing demands from the supervisors [59]. A majority of researchers reported that work stress is negatively related to the physical health of employees and WLB [14,60,61].

WPLE
Also labeled as work-life enrichment, work-personal life enhancement (WPLE) is related to "how positive experiences in work and non-work (home) lives interact to productivity gains in satisfaction, health, and performance" [62]. While a supportive organizational climate, social support from the peers, and family-friendly human resources management strategies are the antecedents to WPLE, outcomes include work engagement, performance, commitment, and job and family satisfaction [63][64][65].

Job Satisfaction
One of the widely studied dependent variables in the literature on organizational behavior and personnel psychology is 'job satisfaction' [66]. Job satisfaction is defined as a "pleasurable or positive emotional state resulting from the appraisal of one's job or job experiences" [67] (p. 1304). Job satisfaction is a global measure that involves an individual's evaluation of the job situation and various aspects of the job [68][69][70]. In this research, the primary dependent variable is job satisfaction, which was studied in connection with the productivity of employees performing work from home.

Theoretical Background and Hypotheses Development
The happy-productive worker thesis (HPWT) [71,72] provides a theoretical background for this research. The basic tenet of HPWT is that employees who are happier or more satisfied with their jobs likely perform better in the jobs [73]. There is compelling evidence that the productivity-satisfaction relationship is bi-directional, as revealed by a metaanalysis conducted by [68], and that productivity is likely to lead to satisfaction [74][75][76][77]. Further, [72] demonstrated that the productivity-satisfaction relationship would be stronger if it is operationalized more broadly than job satisfaction. The employee well-being and work-life balance would boost the performance of employees on the job. When individuals can assess their satisfaction with work and life and maintain a balance between these two [34], it is more likely that WLB leads to job satisfaction. As documented by [77], employees maintaining WLB are more motivated and productive and less stressed, and an increased productivity results in higher levels of job satisfaction. This research uses the constructs productivity (WFHP), work stress, emotional exhaustion, and job satisfaction. The HPWT framework would help to explain the relationship between these variables [78,79].
Using the HPWT framework, we developed a double-layered moderated mediation model ( Figure 1) to explain the relationship between work-from-home productivity and job satisfaction. We will now discuss each connection in the model. The conceptual model is presented in Figure 1.

WFHP and Job Satisfaction
Organizations implement WFH policies to give the employees more flexibility in performing work and to increase productivity [80][81][82]. Extant research documented several benefits of WFH, including an improved performance, reduced employee turnover, and increased job satisfaction [24,26,27]. Since the productivity-satisfaction relationship has been well established in the organizational behavior and industrial psychology literature [66,83], it is more likely that WFHP leads to job satisfaction. Though some scholars argue that the productivity/performance and satisfaction relationship is spurious and that a definite cause-and-effect is unknown [84,85], a majority of the studies evidenced that, when employees are productive, they are likely to be satisfied [57]. Therefore, we offer the following hypothesis based on abundant empirical evidence.

WFHP and WLB
Ever since the concepts of telecommuting (or WFH) and flexible working hours have been introduced in organizations, there has been resounding evidence that employees are happy, productive, and maintain a healthy WLB [86][87][88]. Earlier, organizations used to give employees the option of using telecommuting or remote working on some days in a week so that they maintain a happy balance between their work and life. Now, considering the global pandemic, remote working has become a rule rather than an exception because of periodical lockdowns and social distancing practices worldwide [16,17]. As a flexible work arrangement, teleworking has been recognized as a viable solution in crisis situations, such as a global pandemic. Though, in developing countries such as India, WFH as a concept is relatively new and not regularly practiced, individuals and families are getting adjusted to the change in work culture. Scant research that focused on the relationship between WFHP and WLB indicates that employees are happy because they are able to protect their health from being affected by the deadly pandemic and also remain productive in order to get paid (instead of being laid off). The logic behind a positive association of WFHP with WLB stems from the fact that, when employees are more productive, the pressures from employers will be low and hence workers will be able to devote their time to meeting the demands from life. On the contrary, when the productivity is low, demands from employers will be high and thus will not allow the workers to focus on life-related demands. Anecdotal evidence reveals that employees who show high productivity in jobs are able to maintain a balance between their work and lives. Thus,

WFHP and Job Satisfaction
Organizations implement WFH policies to give the employees more flexibility in performing work and to increase productivity [80][81][82]. Extant research documented several benefits of WFH, including an improved performance, reduced employee turnover, and increased job satisfaction [24,26,27]. Since the productivity-satisfaction relationship has been well established in the organizational behavior and industrial psychology literature [66,83], it is more likely that WFHP leads to job satisfaction. Though some scholars argue that the productivity/performance and satisfaction relationship is spurious and that a definite cause-and-effect is unknown [84,85], a majority of the studies evidenced that, when employees are productive, they are likely to be satisfied [57]. Therefore, we offer the following hypothesis based on abundant empirical evidence.

WFHP and WLB
Ever since the concepts of telecommuting (or WFH) and flexible working hours have been introduced in organizations, there has been resounding evidence that employees are happy, productive, and maintain a healthy WLB [86][87][88]. Earlier, organizations used to give employees the option of using telecommuting or remote working on some days in a week so that they maintain a happy balance between their work and life. Now, considering the global pandemic, remote working has become a rule rather than an exception because of periodical lockdowns and social distancing practices worldwide [16,17]. As a flexible work arrangement, teleworking has been recognized as a viable solution in crisis situations, such as a global pandemic. Though, in developing countries such as India, WFH as a concept is relatively new and not regularly practiced, individuals and families are getting adjusted to the change in work culture. Scant research that focused on the relationship between WFHP and WLB indicates that employees are happy because they are able to protect their health from being affected by the deadly pandemic and also remain productive in order to get paid (instead of being laid off). The logic behind a positive association of WFHP with WLB stems from the fact that, when employees are more productive, the pressures from employers will be low and hence workers will be able to devote their time to meeting the demands from life. On the contrary, when the productivity is low, demands from employers will be high and thus will not allow the workers to focus on life-related demands. Anecdotal evidence reveals that employees who show high productivity in jobs are able to maintain a balance between their work and lives. Thus, based on the available empirical evidence before and during the global pandemic, we offer the following hypothesis: Hypothesis 2 (H2): WFHP positively predicts WLB.

WLB and Job Satisfaction
Though challenging, balancing work and life with individuals is indispensable to maintaining job and life satisfaction [89]. In a large study conducted by [37] on 1416 employees from seven distinct populations (Malaysian, Chinese, New Zealand Maori, New Zealand European, Spanish, French, and Italian), WLB is strongly and positively associated with job satisfaction and life satisfaction. Past researchers documented that happy individuals at home also tend to be satisfied at work [13,90]. In one of the recent studies conducted in Indonesia, researchers reported that WLB strongly and positively influenced job satisfaction [91]. Abundant research provided evidence that employees who are high on balancing life and work tend to derive more job satisfaction. The literature reviews of [10,40,42,92] advocated that WLB has positive outcomes: job satisfaction, job performance, organizational commitment, and life satisfaction. Based on available abundant empirical support, we offer the following hypothesis: Hypothesis 3 (H3): WLB positively predicts job satisfaction.

WLB as a Mediator
Organizational psychologists contend that employee productivity is related to hedonic and eudaimonic satisfaction [93,94]. However, most importantly, productivity indirectly affects satisfaction through employee well-being, which can be seen in terms of WLB. In other words, WFH enables the employees to attend to the demands stemming from non-work activities while meeting the employer's demands. This WLB is essential for the well-being of individuals, which, in turn, results in job satisfaction. That is to say, WFHP may have both direct and indirect effects through WLB on job satisfaction.
When employees can distribute time to be spent on work and their personal lives, work-family conflicts will be low, and satisfaction will be high [24,[95][96][97]. In this study, we contend that WFH increases productivity and performance. At the same time, it will enable the employees to balance the demands of life and work, contributing to satisfaction. Based on the logos and intuitive appeal, we offer the following exploratory hypothesis: Hypothesis 4 (H4): WLB mediates between WFHP and job satisfaction.

First Stage Moderation of Work Stress
Work stress, also known as job stress or occupational stress, occurs when there is a mismatch between an individual's capabilities and requirements and resources for completing a task [98,99]. Work stress is related to the psychological and physiological reactions to the conditions that employees face in the workplace and has adverse effects on employee well-being and health [100]. Researchers report that some of the antecedents of work stress include job conflict, role conflict [57,58], non-collegiality, and increasing demands from the supervisors [59]. In addition, extant research reported that work stress is negatively related to the physical health of employees and WLB [14,61].
Direct effects of work stress on job satisfaction and commitment aside, in this study, we argue that work stress changes the strength of the relationship between WFHP and WLB. Since the negative relationship between work stress and WLB has been established in some studies [57,61], it would be interesting to study the role of work stress as a moderator. In a recent study conducted on 369 employees working in banking and high-tech industries in Taiwan, researchers found that long working hours resulted in work stress adversely affecting the WLB and satisfaction [101]. As some studies during the global pandemic found that work stress adversely affected the performance of employees [3,102], we contend that work stress would significantly alter the positive association between WFHP and WLB. Abundant literature on the stress and mental health of the healthcare and frontline workers during the global pandemic reveals adverse consequences on anxiety [103] and WLB [104]. This study investigates work stress as a boundary condition influencing the relationship between WFHP. Based on the above, we offer the following exploratory moderation hypothesis: Hypothesis 2a (H2a): Work stress moderates between WFTP and WLB such that low (high) level work stress strengthens (weakens) the relationship.

Second Stage Moderation of WPLE
As work interfering with life (WIL) and life interfering with work (LIW) are inherent phenomena, these effects on WLB largely depend on how the employees work towards enhancing both work and life. Work's role in individuals' lives plays a crucial role because work is the source of income. Hence, employees generally consider work as essential and derive energy and motivation by performing work satisfactorily. At the same time, individuals tend to enjoy life because the goal of work is to spend life happily. Therefore, the spillover from work to life and life to work should be understood in order to see the effect of spillover on WLB. When a job gives energy to life and vice versa, employees tend to move towards enhancing both work and life. In a recent study conducted among 283 employees working in the information technology sector in India, WLB positively predicted employee retention and psychological empowerment [105].
Employees differ in managing their work and life domains [106]. However, some can effectively navigate by employing strategies to combat imbalance by focusing on the energy that they obtain from work and life. For example, though there is resounding evidence that working long hours increases work stress, employees feel that it is necessary to complete the demanding tasks in order to succeed professionally. Therefore, they attempt to rationalize their work and maintain sustainable WLB, which is rewarding. In other words, individuals may implement strategies to enhance the work-life interface to counter the negative effect of work stress on WLB. Following this logic, it would be interesting to see the impact of WPLE (second moderator) on the relationship between WFTP and work stress (first moderator) influencing WLB. Thus, we offer the following exploratory hypothesis: Hypothesis 2b (H2b): WPLE moderates the relationship between WFTP and work stress to influence WLB.

Third Stage Moderation of Emotional Exhaustion
It was evidenced by previous researchers, as discussed in the earlier sections, that WLB was positively associated with job satisfaction [74,107]. From the viewpoint of employees, it is essential to see how emotional exhaustion, which is related to physical fatigue and a personal drain of energy, affects job satisfaction [43]. Extant research reported that an unreasonable workload and time pressure in completing work might result in emotional exhaustion [108]. Previous studies examined the adverse effects of emotional exhaustion on employee turnover, commitment, and performance [44,46,109]. In a recent study conducted on Norwegian school principals, [110] found that emotional exhaustion was negatively related to job satisfaction.
While the adverse effects of emotional exhaustion on performance and satisfaction were well documented [111,112], some scholars studied the moderating role of emotional exhaustion in the relationship between role ambiguity and workload and performance [51]. During the global pandemic, employees worldwide have undergone severe emotional exhaustion, as has been reported by several scholars [113,114]. Exploring the moderating role of emotional exhaustion in the WLB-job satisfaction relationship would be interesting.
Tying emotional exhaustion to the consequences of WLB is another study objective. Individuals differ in how they feel exhausted at work; hence, the effect of WLB on job satisfaction depends on the strength of the emotional exhaustion that they experience. Since emotional exhaustion is a part of work-related stress, the greater the emotional exhaustion employees feel because of overburdened work, the more likely the positive relationship between WLB and job satisfaction would be affected. Conversely, under lower emotional exhaustion, the effect of WLB on job satisfaction would be lesser when compared to higher levels of emotional exhaustion. Though previous studies did not dwell on the moderating effect of emotional exhaustion, we offer the following exploratory hypothesis: Hypothesis 3a (H3a): Emotional exhaustion moderates between WLB and job satisfaction such that lower (higher) levels of emotional exhaustion interacting with WLB would result in higher (lower) job satisfaction.

Fourth Stage Moderation of WPLE
Past researchers documented that the spillover between work and personal life plays an important role [115,116]. To counter the negative effect of emotional exhaustion, we contend that positive energy derived from work and life helps to alleviate the negative impact of emotional exhaustion on job satisfaction. Over a decade ago, [117] empirically demonstrated that work-family enrichment was a mediator in the relationship between WLB and job outcomes. Sometimes, organizational interventions such as supportive work culture, human resource policies promoting family-friendly relations, and social support at the workplace contribute to WPLE. Previous researchers have established a positive relationship between organizational interventions and work-family enrichment [62,118].
Though positive outcomes of WPLE are understandable, it would be interesting to explore how WPLE helps to alleviate the ill effects of emotional exhaustion on job satisfaction. To the best of our knowledge, the moderating role of WPLE in influencing individual outcomes in terms of job commitment, performance, and satisfaction has not been extensively studied by past researchers. In the context of employees working from the home because of the global pandemic, WPLE plays a vital role in smoothening the relationship between WLB and job satisfaction by reducing the negative effect of emotional exhaustion. The three-way interaction between WLB, emotional exhaustion, and WPLE is not simply an academic exercise, and we argue that it would benefit the individuals by enhancing job satisfaction. Based on the positive effects of WLB on job satisfaction, adverse effects of emotional exhaustion on job satisfaction, and positive influence of WPLE, we offer the following exploratory moderated moderated-mediation hypothesis: Hypothesis 3b (H3b): WPLE moderates the relationship between WLB and emotional exhaustion to influence job satisfaction.

Sample
Since the focus of this research is to investigate the impact of productivity of employees working remotely, especially during the COVID-19 pandemic, the respondents consisted of employees working from home. In India, as in other countries worldwide, many organizations resorted to work from home (WFH) to avoid spreading the virus. Therefore, we preferred to use Google forms to collect the data because of social distancing norms and periodical lockdowns. This has been followed by several researchers collecting surveys during the global pandemic [119,120].
According to the sample size tables prescribed by [121], the minimum required sample size when the population is over 100,000 is 384. Further, [122] recommends that a sample size of over 500 is considered very good (100 = poor; 200 = fair; 300 = good; 500 = very good; 1000 or more = excellent). Considering the post-pandemic scenario, we contacted the respondents personally and asked them to fill out the surveys using Google forms to collect the data. It took nearly six months to collect completed surveys from 1158 respondents. Thus, the final analysis consisted of 1158 completed surveys, considered 'excellent'.

Data Collection
The respondents were from the southern part of India. Four research students were involved in collecting data, and visited the houses of these respondents, following the COVID-19 restrictions of wearing face masks and taking necessary precautions. First, we contacted the supervisors of organizations and secured information about the employees. We made sure that the organizations asked the employees to work from home. We asked a qualifying question, "Did you work from home during the global pandemic?", and suggested that they proceed only if the answer was 'Yes'. We followed convenience sampling, as some respondents were not filling out surveys, and, during the global pandemic, it was challenging to have a probability-based sample. Many researchers use snowball sampling to collect data by submitting Google forms [16,123,124]. In order to have representativeness, we concentrated on obtaining as many respondents as possible. We tested for non-response bias by comparing the first 100 responses to the final 100 responses and found no statistically significant difference between these groups.

Measures
This study used six constructs, and the measures were adapted from the previously tested and validated sources. A five-point Likert scale ('5' = strongly agree; '1' = strongly disagree) was used to measure the constructs. The constructs, indicators, and sources of these constructs are presented in Table 1. WFHP was measured with ten items adapted from [125]. The sample items read as: "I am very productive while working from home" and "I feel that the quality of the work I do during working from home is better". The reliability coefficient of WFHP was 0.93. WLB was measured with eight items adapted from [126][127][128]. The sample items read as: "I have an adequate time to spend with the family even if I work in the organization or at home" and "I have sufficient time to take care of my children even if I work in organization or at home". The reliability coefficient Cronbach's alpha of WLB was 0.91.
Work stress was measured with six items adapted from [129]. The sample items read as: "I am discouraged about my work" and "I feel many things are beyond my control and ability while working from home". The reliability coefficient Cronbach's alpha of work stress was 0.89.
Job satisfaction was measured with five items adapted from [130]. The sample items read as "I am satisfied with my current job", and "I am satisfied with my current coworkers". The reliability coefficient of job satisfaction was 0.88. WPLE was measured with four items adapted from [96]. The sample item reads as: "Personal life gives me energy for my job" and the reliability coefficient of WPLE was 0.84.
Emotional exhaustion was measured with four items adapted from [131]. The sample item reads as: "I have felt emotionally drained from my work", and the reliability coefficient of emotional exhaustion was 0.85.

Measurement Model and Confirmatory Factor Analysis (CFA)
As [132] suggested, we followed a two-step approach to checking the measurement and structural models. We used the LISREL software of structural equation modeling (SEM) to check the measurement model and performed a confirmatory factor analysis (CFA). The results of CFA are presented in Table 1.
As shown in Table 1, the factor loadings for all of the indicators are over 0.70, except for one indicator for job satisfaction, which is 0.67. Since this is a well-established measure, we retained this indicator in the analysis. The reliability coefficient (Cronbach's alpha) for all six constructs was over 0.70, and the composite reliability (CR) values were over 0.70. Further, the average variance-extracted (AVE) estimates for all of the six constructs was greater than 0.50, and these statistics vouch for the measures' reliability, convergent validity, and internal consistency [133,134].

Convergent Validity, Discriminant Validity, and Common Method Bias
To establish discriminant validity, the square root of AVEs of the variables must exceed the correlations between the variables [135]. In this research, discriminant validity is said to be established when the square root of AVEs of the variables exceeds the correlations between the variables [136,137].
The correlation between WFHP and WLB is 0.56, and the square root values of AVE are 0.77 and 0.75, respectively. Similarly, the correlation between emotional exhaustion and WLPE is −0.60, and the square roots of AVE are 0.77 and 0.76, respectively. For all other variables, the correlations between the variables were less than the square root of AVEs, thus providing support for discriminant validity between the variables.
We also compared the baseline six-factor model with five other alternative models, and the results are presented in Table 2.
A comparison of measurement models reveals that the six-factor model fit the data well (χ 2 = 2160.36; df = 650; χ 2 /df = 3.32; root mean square error of approximation (RMSEA) = 0.045; root mean square residual (RMR) = 0.042; standardized RMR = 0.030; comparative fit index (CFI) = 0.946; goodness of fit index (GFI) = 0.942). A comparison of five alternative models with the baseline model (see Table 2) shows that the goodness of fit indices (RMSEA < 0.08; CFI > 0.90; and other indices) provided a good fit of the model to the data [138].

Descriptive Statistics and Multicollinearity
The descriptive statistics consisting of means, standard deviations, and zero-order correlations are presented in Table 3. An assessment of multicollinearity requires observing correlations between the variables because correlations exceeding 0.75 signify the presence of multicollinearity [139]. The highest correlation was −0.74 (between WLB and work stress) and the lowest correlation was −0.47 (between work stress and job satisfaction). As an additional check, we verified the variance inflation factor (VIF) and noted that the values were less than 5 for all of the variables, suggesting that multicollinearity is not a problem with the data [140].

Common Method Variance
We checked the common method variance (CMV) in three ways. First, we performed Harman's single-factor test, as suggested by [141], and found that a single factor accounted for less than a 25% variance (less than 50%), signifying that CMV is not a problem with the data. Second, we compared the goodness of statistics of the one-factor model with the six-factor model (see Table 3) and found that the one-factor model was a poor fit of the data (χ 2 = 9562.98; df= 665; χ 2 /df = 14.38; RMSEA = 0.107; RMR = 0.123; standardized RMR = 0.089; CFI = 0.682; GFI = 0.528). Third, we performed a latent-factor method wherein all of the indicators were loaded into a single factor each time and checked the VIF values and found that these values were less than 3.3. If the VIF values are more than 3.3 the data indicate the existence of pathological collinearity and the model is inferred to be contaminated by common method variance [142]. All three tests indicate that CMV is not a problem in this research.

Hypotheses Testing
The structural model was tested using [143] PROCESS macros. We used model #4 for testing H1-H4 and present the results in Table 4.

DV = Job Satisfaction DV = WLB H2 DV = Job Satisfaction
Step 1 Step 2 Step 3 Step 1 from Table 4 shows that the regression coefficient of WFHP on job satisfaction was positive and significant (β = 0.619, t = 26.14; p < 0.001). The results based on 20,000 bootstrap samples show that the 95 percent bias-corrected confidence interval (BCCI) was 0.5726 (LLCI) and 0.6566 (ULCI). These results support H1 in that WFHP positively predicts job satisfaction.
Hypothesis 3 posits that WLB positively predicts job satisfaction.
Hypothesis 4 states that WLB mediates the relationship between WFHP and job satisfaction. The indirect effect (as shown in the bottom of the Table 4) was 0.1230 (Boot se = 0.0221; Boot LLCI = 0.0829; Boot ULCI = 0.1692) and, since zero was not contained in the Boot LLCI and Boot ULCI, the results support the mediation hypothesis (i.e., H4).
We also double checked the results by adding the direct effect (0.4961) and indirect effect (0.1230) to obtain the total effect (0.6191) (i.e., 0.4961 + 0.1230 = 0.6191). Further, the indirect effect is a product of the regression coefficient of WFHP on WLB (0.5194) and that of WLB on job satisfaction (0.2369) [i.e., (0.5194 × 0.2369 = 0.2369). The indirect effect of WFHP WLB job satisfaction was significant, thus corroborating support of H4.

Testing the Moderated and Moderated Mediation Hypotheses (H2a and H2b)
We used Model #11 of [143] PROCESS macros to check the two-way and three-way interactions and present the results in Table 5. As shown in Table 5, the regression coefficient of the interaction term (WFHP × work stress) was significant (β WFHP × work stress = 0.1164; t = 2.2956; p < 0.05; Boot LLCI (0.0169); Boot ULCI (0.2160). These results support H2a. The two-way interaction is presented in Figure 2.
As shown in Figure 2, at high levels of work stress, WFHP is associated with low WLB. Further, the slope of the line representing high work stress is negative and the slope of the line representing low work stress is positive, thus supporting the hypothesis that work stress moderates the relationship between WFHP and work stress (H2a).
Hypothesis 2b posits that WPLE (second moderator) and work stress (first moderator) interact with WFHP to influence WLB. The regression coefficient of the three-way interaction was significant (β WHFP × work stress × WPLE =−0.047; t = −3.42; p < 0.01). The conditional effects of the WLB at values of moderators work stress and WPLE are presented at the bottom of Table 5. The indirect effect of WFHP on job satisfaction is mentioned in Table 6.  As shown in Figure 2, at high levels of work stress, WFHP is associated with low WLB. Further, the slope of the line representing high work stress is negative and the slope of the line representing low work stress is positive, thus supporting the hypothesis that work stress moderates the relationship between WFHP and work stress (H2a).
The visual presentation of the three-way interaction is shown in two panels of Figure 3. The relationship between WFHP, work stress, and WLB at low levels of WPLE (in Panel A), and at high levels of WPLE (Panel B) is shown. When WPLE is low, an increase in work stress decreases WLB as the curve has a negative slope. On the contrary, when WPLE is high, even when at higher levels of work stress, WLB increases, as the curve is positively sloped. These results corroborate the support for H2b. The visual presentation of the three-way interaction is shown in two panels of Figure  3. The relationship between WFHP, work stress, and WLB at low levels of WPLE (in Panel A), and at high levels of WPLE (Panel B) is shown. When WPLE is low, an increase in work stress decreases WLB as the curve has a negative slope. On the contrary, when WPLE is high, even when at higher levels of work stress, WLB increases, as the curve is positively sloped. These results corroborate the support for H2b.

Testing the Second Moderated Moderated-Mediation Hypotheses (H3a and H3b)
In order to test H3a and H3b, we used model #18 of [143] PROCESS macros and present the results in Table 7.

Testing the Second Moderated Moderated-Mediation Hypotheses (H3a and H3b)
In order to test H3a and H3b, we used model #18 of [143] PROCESS macros and present the results in Table 7.
Hypothesis 3a posits that emotional exhaustion moderates the relationship between WOB and job satisfaction. The regression coefficient of the two-way interaction term was significant (β WLB × emotional exhaustion = 0.131; t = 1.99; p < 0.05). The bootstrapping results were based on 20,000 bootstrap samples and 95% Boot LLCI (0.0019) and BOOT ULCI (0.2595), and since 'zero' was not contained in the lower and upper limits, H3b is supported. The visual presentation of the interaction effect is shown in Figure 4.
As shown in Figure 4, lower levels of emotional exhaustion interacting and low levels of WLB are associated with a higher job satisfaction than higher levels of emotional exhaustion and lower levels of WLB. Further, when WLB increases from 'low' to 'high', job satisfaction falls when emotional exhaustion is high, as the curve representing high emotional exhaustion is negative. On the contrary, high levels of WLB combined with lower level of emotional exhaustion results in a higher job satisfaction, as the curve representing a low level of exhaustion is positive. These results corroborate the support for H3a. Hypothesis 3b is related to the interaction between WLB, emotional exhaustion (first moderator) and WPLE (second moderator) influencing job satisfaction. The regression coefficient of the three-way interaction was significant (β WLB × emotional exhaustion × WPLE = −0.049; t = −2.87; p < 0.01). Conditional effects of the job satisfaction at values of emotional exhaustion and WPLE are presented at the bottom of Table 7. The indirect effect of WFHP on job satisfaction is mentioned in Table 8. As shown in Figure 4, lower levels of emotional exhaustion interacting and low levels of WLB are associated with a higher job satisfaction than higher levels of emotional exhaustion and lower levels of WLB. Further, when WLB increases from 'low' to 'high', job satisfaction falls when emotional exhaustion is high, as the curve representing high emotional exhaustion is negative. On the contrary, high levels of WLB combined with lower level of emotional exhaustion results in a higher job satisfaction, as the curve representing a low level of exhaustion is positive. These results corroborate the support for H3a.
Hypothesis 3b is related to the interaction between WLB, emotional exhaustion (first moderator) and WPLE (second moderator) influencing job satisfaction. The regression coefficient of the three-way interaction was significant (β WLB × emotional exhaustion × WPLE = −0.049; t = −2.87; p < 0.01). Conditional effects of the job satisfaction at values of emotional exhaustion and WPLE are presented at the bottom of Table 7. The indirect effect of WFHP on job satisfaction is mentioned in Table 8.   The index of moderated moderated-mediation from 20,000 bootstrap samples was −0.0254, and Boot SE (0.0090), Boot LLCI (−0.0429), and BOOT ULCI (−0.0075) present significant values, thus supporting H2b.
The visual presentation of the three-way interaction is shown in two panels of Figure 5. The relationship between WLB, emotional exhaustion, and job satisfaction at a low level of WPLE is shown in Panel A, and at a high level of WPLE in Panel B. When WPLE is low, an increase in emotional exhaustion decreases job satisfaction as the curve has a negative slope. On the contrary, when WPLE is high, even lower levels of WLB and lower levels of emotional exhaustion result in a higher job satisfaction when compared to higher levels of emotional exhaustion. When we move to Panel B, all three curves, representing low, middle, and higher levels of emotional exhaustion, are positive when WPLE is high. These results corroborate the support for H3b. The visual presentation of the three-way interaction is shown in two panels of Figure  5. The relationship between WLB, emotional exhaustion, and job satisfaction at a low level of WPLE is shown in Panel A, and at a high level of WPLE in Panel B. When WPLE is low, an increase in emotional exhaustion decreases job satisfaction as the curve has a negative slope. On the contrary, when WPLE is high, even lower levels of WLB and lower levels of emotional exhaustion result in a higher job satisfaction when compared to higher levels of emotional exhaustion. When we move to Panel B, all three curves, representing low, middle, and higher levels of emotional exhaustion, are positive when WPLE is high. These results corroborate the support for H3b.

Discussion
As many organizations worldwide have embarked on employing strategies of shifting the work mode from office to home, this study aims to investigate the relationship between employee productivity of WFH and job satisfaction. A conceptual model was developed, and hypotheses were tested using a large sample (N = 1158) in the context of a developing country, India. First, the psychometric properties of the survey instrument were tested, and the structural model was tested using [143] PROCESS macros. In this study, all of the hypotheses were supported.
First, the results revealed that WFHP is positively associated with job satisfaction (Hypothesis 1), which is consistent with findings from the literature [23,25,80]. This result has intuitive appeal because, when employees are productive, they are likely to obtain monetary rewards and extrinsic satisfaction, and dedication to work gives them intrinsic satisfaction. Second, the positive relationship between WFHP and WLB (Hypothesis 2) was supported in this study. Again, these results are in line with the previous studies in the literature [15,86]. When employees remain productive while working from home, the pressures from employers would be low, and employees will have adequate time to devote to family activities. On the contrary, when the productivity is low, constant pressure from the employers would not enable the employees to divert their energies and time for non-work activities.
Third, this study reported a positive relationship between WLB and job satisfaction (Hypothesis 3); the finding corroborates with previous studies [12,91]. When employees can maintain a happy balance between work and life, they will more likely be satisfied both at work and at home. The results also support the indirect effect of WFHP on job satisfaction through WLB (Hypothesis 4); the finding is consistent with logic and predictions from the previous studies [23,95,96].
Fourth, a key finding from this research is the moderating effect of work stress on the relationship between WFHP and WLB (Hypothesis 2a). Though there were no previous studies to vouch for this interaction effect, the negative impact of work stress on individual outcomes (e.g., performance) supports the moderation hypothesis [61,101]. Fifth, as predicted, WPLE reduced the negative effect of work stress on WLB (Hypothesis 2b) and found support in this research. When employees derive energy from non-work activities and use that energy at work, and, at the same time, energy derived from work is applied to non-work activities, contributing to the enhancement of both work and personal life, it helps to reduce the negative effect of work stress [105].
Sixth, this research supports that emotional exhaustion experienced by employees has a deleterious effect on job satisfaction (Hypothesis 3a). Previous researchers have investigated the moderating role of emotional exhaustion in other contexts [51,110,111,113], and, hence, the finding from this study is consistent with the literature.
Seventh, another critical contribution from this study is the role of WPLE in reducing the negative impact of emotional exhaustion on job satisfaction (Hypothesis 3b). Though no previous studies were available to vouch for this result, the relationship is expected as some other researchers have documented the positive effect of WPLE in different contexts [115,118].
This research, riding on the HPWT [72], documents that worker productivity is a precursor for WLB and job satisfaction. Further, despite the negative effect of job stress and emotional exhaustion, which were predicted to have a negative impact on job satisfaction and a positive effect on turnover, WPLE acts as an antidote and helps employees to achieve a higher performance and satisfaction.

Theoretical Implications
The present study contributes to both the literature on WLB and job satisfaction. First, to the best of our knowledge, the conceptual model developed and tested in this study is the first of its kind in the Indian context, making a significant contribution to the growing literature on WLB. It is essential to observe that, though this study is conducted in the context of the developing country, India, the results are consistent with the findings from research on WLB in developed countries [32,115].
Second, considering the global pandemic scenario, when WFH has become a rule rather than an exception, the study found that employees' productivity plays a vital role in enhancing WLB and job satisfaction. Third, the present study supported the relationship between productivity (performance) and satisfaction, which has been exhaustively studied for the last five decades, thus adding to the literature.
Fourth, the effect of WFHP on job satisfaction through balancing work and life (WLB as a mediator) contributes to the literature on WLB in a modest way.
Fifth, an essential contribution is the moderating effect of work stress in the relationship between WFHP and WLB, which previous researchers have not investigated. Most importantly, the role of WPLE (second moderator) in alleviating the negative effect of work stress (first moderator) in the relationship between WHFP and job satisfaction mediated through WLB makes a unique contribution to the literature. Furthermore, the moderated mediation hypothesis, denoting the three-way interaction between WFHP, work stress, and WPLE, is a novel idea explored in this research for the first time.
Sixth, as mentioned in the title, regarding the double-moderation model, we investigated the role of WFHP (second moderator) in influencing the relationship between WLB and emotional exhaustion (first moderator) in predicting job satisfaction. The threeway interaction between WLB, emotional exhaustion, and WPLE is another fundamental contribution of this research.
In summary, the two two-way interactions (WFHP × work stress; WLB × emotional exhaustion) and two three-way interactions (WFHP × work stress × WPLE; WLB × emotional exhaustion × WPLE) provide a new dimension of research that makes a substantial contribution to the literature on organizational behavior and industrial psychology. Most importantly, the present study shows the benefits of WPLE in reducing the negative impact of work stress and emotional exhaustion in the global pandemic environment.
Riding on the HPWT framework, the results from this research extend the theoretical base of productivity relationship by adding the work-life balance as a significant contributor to satisfaction and performance. Though the notion of a satisfied worker need not be a productive worker and, vice versa, was in academic circles for some time, the latest scholars have a fresh look at this theory by identifying the bi-directional nature of satisfaction and performance [66,[74][75][76]. Furthermore, the double-layered moderated mediation also adds to the approach to stress and emotional exhaustion by providing empirical evidence that supports alleviating the impact of stress and emotional exhaustion through work and personal life enhancement.

Practical Implications
The findings from this study have several implications for managers in organizations where employees are performing duties remotely. First, as the study documented, employee productivity plays a vital role in maintaining WLB; the managers must provide information about employee performance. As expected, a happy balance of work and life leads to job satisfaction; employees must devote time between work and home appropriately.
Second, as several researchers found that work stress has escalated because of the sudden shift from on-site to off-site (remote), which adversely affected WLB, employees need to focus on the strategies to enhance personal life and work (WPLE). To reduce the negative effect of stress on job satisfaction and WLB, managers need to see that employees are not under severe work pressure. This is because, in the global pandemic environment, workers already undergo the psychological stress of maintaining health while protecting themselves and their families from the virus.
The third important implication of this study is the role of WLB as a mediator in the relationship between WFHP and job satisfaction. As WLB is a precursor to job satisfaction, managers need to focus on the employees who cannot reconcile the demands from work and home and take corrective steps. If the resources required to complete given jobs assigned to the employees are inadequate, managers need to grant them resources. Further overburdening the employees with work to compensate for the accrued loss because of lockdowns and closure may result in emotional exhaustion that may adversely affect productivity and job satisfaction. The managers are, therefore, advised to give work that employees can shoulder.
Fourth, the study suggests that the employers conduct periodic surveys to see how the employees perceive the work pressure and whether they can maintain a balance between work and life. Employee satisfaction surveys can help the managers to identify the problems associated with work and life so that they can take appropriate measures to restore WLB. In summary, this study underscores the importance of maintaining WPLE in alleviating the ill effects of emotional exhaustion and work stress on job satisfaction and WLB.

Limitations
As with any social science research, this study has some limitations. This study was conducted in a global-pandemic-induced environment where employees work from home. Although the concept of WFH was in vogue two decades ago when employees had the option of remote working some days a week (especially in developed countries such as the USA, UK, and Australia), organizations worldwide have resorted to WFH during the pandemic, which is continuing. Even after the availability of vaccination, many employees continue to WFH for safety, and many employers continue encouraging the workers to work remotely. The conceptual model developed was focused on the productivity of workers engaging in WFH. When the situation becomes near normal, the dynamics of WFH and on-site productivity may be slightly different. Working in physical locations in some organizations is mandatory; hence, this study may not apply to organizations working on-site.
The second major limitation is that the study is conducted from the perspective of a developing country (i.e., India). The results are therefore generalizable across many developing countries. However, since the work culture, administrative apparatus, and technological infrastructure in developing countries are different, the results should be interpreted carefully when we apply these to developed countries.
Third, though the larger sample size in this study is a positive point, focusing on a variety of industries (technology, education, manufacturing, and services) in this study may be a limitation because the nature of work in these industries is different. However, the results are generalizable to the extent where the employees find similar problems (e.g., network).

Future Research
This study offers several avenues for future research. First, in addition to the variables in this research, future studies may involve the influence of work-family conflict (WFC) and family-work conflict (FWC) on WLB and job satisfaction. Second, work-to-family enrichment (WFE) and family-to-work enrichment (FEW), as suggested by [144], can be included in future studies on WLB and job satisfaction. Third, as organizational citizenship behavior (OCB) plays a vital role in individual and corporate performance, it would be of interest to future researchers to include OCB in affecting the WLB-job-satisfaction relationship. Fourth, a cross-country comparison of developed versus developing countries may help to identify the cultural differences in influencing the relationships between the variables in this study. Finally, future researchers may be encouraged to conduct longitudinal studies to see how the dynamics of the interaction of variables change over time.

Conclusions
In this research, we attempted to underscore the border conditions of alleviating the ill effects of work stress and emotional exhaustion in the global-pandemic-induced environment worldwide. As developed in this study, a double-layered moderated mediation model emphasizes the importance of maintaining work-life enrichment in order to counter emotional exhaustion and work stress. Further, this study highlights the importance of WLB and sustaining workers' productivity at home to enhance job satisfaction. However, the pandemic is slowly becoming endemic before the organizations come to a near-normal situation. When the employees return to the usual way of functioning, the study provides direction to both the employer and employees on what to carry out to maintain WLB. It is concluded that, in addition to being productive, how employees perceive their work and family lives is vital in enhancing job satisfaction. Even after the restoration of normalcy, the research on WLB continues to be on the agenda of future researchers.