MODELLING THE IMPACT OF HEAVY WORK INVESTMENTS ON EMPLOYEE’S SATISFACTION

In this article, the authors develop an econometric model to determine the impact of the heavy work investment on employee’s satisfaction. The data was collected using a questionnaire addressed to employees with higher level of education and at least one year of experience, within the specialized departments of multinational companies that offer consultancy and audit services in Romania and analyzed using Smart-Pls statistical software 3.3.2. Based on the literature in the field, the article identifies the determinants of the heavy work investment. These indicators are the independent variables that shape the impact of the heavy work investment on employee’s performance and satisfaction. The findings confirm the theoretical aspects presented in the article, emphasizing that employee’s satisfaction is the direct result of the collaboration of several factors, such as personal needs, workload, time spent at work, attitude to work and physical and mental condition. Identifying the drivers that contribute to increasing organizational performance is essential in the human resources management process, to shed light on positive aspects, such as employee satisfaction, rewarded for their work and, at the same time, to prevent the risk of overwork. The article contributes to the research theory by modelling the investment in work according to elements specific to human typologies, bringing a unique contribution in the form of practical inferences for the management of human resources.


Introduction
Research studies have revealed the complexity of the phenomenon of the heavy work investment, closely correlated with workaholism. Research studies in the field of neuroscience (Andreassen, 2014) found that work addiction is a worrying phenomenon, which can have destructive consequences on the individual. There is a fine line between work enthusiasts and work addicts, analyzed in this paper by the Heavy Work Investment (HWI) indicator, which divides human typologies into two categories, depending on the HWI level, respectively the workaholics with a negative HWI index and those involved with enthusiasm in professional activities, with a positive HWI index. As highlighted in the next section, the HWI concept has attracted pros and cons opinions in the literature. Therefore, this article aims to find the drivers underlying the HWI indicator, to capitalize on human potential, in the benefit of the individuals and the organization.
The influence of HWI factors on job satisfaction and organizational performance, in sections 2 and 3, is based on economic modelling of the impact of the heavy work investment. Employee's satisfaction for the work performed becomes a priority in the context of human resource management. The questionnaire addressed to the interviewees reflects the importance of organizing the activity in the form of projects, in terms of autonomy given to coordinators, flexibility and enhanced communication within the teams. The econometric study performed with the method Partial Least Square -Structural Equation Modelling (PLS-SEM) allows a relevant assessment of the degree of dependence between outcome variables, such as job satisfaction, with independent variables, such as number of hours allocated to work or workload, mood and individual physical condition.
The conclusions of the research papers in the field lead to the idea of extrapolating the HWI study for large communities, which would have in common the desire to work, regardless of the income, through the importance attributed to the professional activities carried out, known as work centrality. However, in the current macroeconomic context of the developed countries, more and more companies are becoming adept to a healthy and sustainable way of doing businesses, where overwork is not a long-term option. Empirical studies dealing with HWI modelling remain, however, limited. Therefore, in order to respond to the need to capitalize on the HWI indicator, while removing the risk factors, a detailed analysis of each influencing factor, essential within the organization's management, is required.
In this article, the authors focus on the answer to the question "What is the impact of massive investment in labour on employee satisfaction?", and, to answer this question, a quantitative analysis was conducted based on a survey drawn up on the basis of replies to a questionnaire. The statistical hypotheses formulated in the methodology section were validated using structural equations modelling. The satisfaction of the work performed was analyzed in the light of the HWI indicator by Schaufeli, Shimazu and Taris (2009). Identifying work satisfaction as a determinant for HWI is relevant in the human resource management process. The study conducted in this article, based on the survey, finds that activities grouped by projects increase autonomy, task identity and communication within project teams.
The purpose of this paper is to validate the causal relationship between the massive investment in work and the degree of employee satisfaction, by examining the data collected from audit and consulting companies in Romania. As such, the causal relationship between the HWI model and job satisfaction is assessed by using the structural equations modelling, tested with the least partial squares method. The present paper continues with the review of the specialized literature, followed by the research methodology, where the research hypotheses are formulated and tested. The paper continues with a section of results and discussions, in which the main results of the article are discussed and compared with those of the literature. Discussions, limitations of the analysis and further research are presented in the conclusions section.

Literature review
The phenomenon called work "addiction" or "workaholism" was firstly observed by the American psychologist Wayne Oates (1971), who defines it as an uncontrollable process of continuous work. In academic language, work dependence is quantified by the Heavy Work Investment (HWI) indicator, which measures the effort and time allocated to work (Harpaz and Snir, 2003), as the main form of the behavioural model development, characterized by maximum involvement, with potential for physical and mental exhaustion (Fassel, 1990;Mazzetti, et al., 2014;Andreassen, 2015;Aziz and Burke, 2015;Shimazu, et al., 2015;Taris and Schaufeli, 2015;Shkoler, et al., 2017;Tziner, et al., 2018).
The importance given to work has been explored by many researchers in profile papers, and the finding that work plays a key role in personal life has been empirically supported in most industrialized countries (Mannheim, et al., 1997;Arvey, et al., 2004). In addition, other studies (Ishiyama and Kitayama, 1994; Barsade and Gibson, 2007) have shown that the emphasis on work is based on personal considerations, but also on external considerations, such as demographic, educational, industrial factors specific to the economic sector, related to the development of technology or the particularities of a culture (Tucker and Rutherford, 2005).
Starting with the importance given to work, Kaplan (1987) studied the causes that determine some lottery winners to continue working, despite the huge financial gains. Among these drivers, the main factors are education, job satisfaction and professional status. For example, among the people who remained involved in the activity are those with higher education, professionals or company managers. Moreover, the theory of the educational factor was developed in the research works of Harpaz (1990), while the psychological motivation of the desire to remain involved in the social-occupational field was the subject of sociological studies applied to a large community, representing the retired population, with a guaranteed source of income (Vecchio, 1980;Wang, et al., 2008;Highhouse, et al., 2010). Other author (Messenger, et al., 2007) made comparisons in terms of allocated working time between national models from Brazil, Chile, China, Czech Republic, Hungary, Korea, Jamaica, Malaysia, Mauritius, Mexico, Peru, Russia, Senegal and Tunisia, based on 14 studies. There is a tendency to reduce the number of hours, with significant regional differences, influenced by age and gender. Messenger (2004) offers solutions to meet the need to balance the work schedule, versus free time. The author proposes five interconnected criteria for establishing an adequate work schedule, namely the protection of health at work and employee safety, a friendly environment, favourable to the development of activities, gender equality, all these contextual factors working together to increase productivity.
Attempts to reduce working hours in developing countries have been undermined for practical perspective, due to the need for workers to work extra to ensure adequate earnings. Although the promotion of working time flexibility is often proposed as an alternative to the long work schedule, in policy documents, which governs contractual labour relations, these flexibility measures, as well as part-time work, are not used in practice, due to the fact that they reduce household income, in the context of low wages and standard of living, compared to the developed countries. According to OECD (2013) study, reconciling work and family life should be a prominent concern of the economic and social policies.
By the development of the conceptualization of HWI theory and research, the effects of the behaviour of the workaholics on themselves or in relation to those with whom they interact were underlined. The observations collected indicate two HWI typologies, which are influencing HWI in a different way, the first being characterized by a negative attitude of the individual towards him/herself, while the activity was carried out and a worsening of the interaction with the organizational or social environment (Dembe, et al., 2005;Iwasaki, et al., 2006;Raediker, et al., 2006), while the second typology is characterized by an attitude positive towards oneself and the organizational framework, manifested by enthusiasm for the team and the activity carried out (Snir and Zohar, 2008). At the same time, other studies (Feldman and Russell, 1998;Barsade and Gibson, 2007) have shown that employees with high levels of emotional intelligence and able to capitalize on their strengths are also better suited to develop and maintain a harmonious passion for their work. Consequently, these employees may experience a positively correlated HWI index compared to those who are dependent on work, but who show a hostile attitude towards the organization, the work carried out, as well as in interpersonal relationships. Therefore, an HWI filled with enthusiasm for work can benefit the individual and society at large (Spence and Robbins, 1992;Kanai, et al., 1996;McMillan, et al., 2002;Vallerand, et al., 2003;Ersoy-Kart, 2005).
Thus, the factors underlying the determination of HWI can be internal in nature, which are largely related to individual traits that define the type of personality (i.e. competitiveness, enthusiasm, energy, job satisfaction, vocation for professional activity, ethics etc.) or may be external in nature, depending on the objective pursued, education, sociological factors, organizational model and financial situation (for example, the desire to be promoted in a career, the need to support a family with several member etc.). In addition, behavioural typologies often identify types of addiction, positive and negative. Positive examples include achievement-oriented behaviours, analyzed by Scott (1997) and workaholics, whose behaviours were studied by Spence and Robbins (1992). Employment can be defined as "a positive, satisfying mood that is characterized by vigour, dedication and absorption" (Schaufeli, et al., 2002, p.73). The results of the research are in line with a taxonomy of work-related well-being, presented by Schaufeli (2013), who stated that employees trained at work, with a positive HWI index, experience pleasant emotions, while employees dependent on work, with a HWI negative index, may develop conflicting states in interpersonal relationships.
The HWI quantification methodology has been developed since the late 1980s, starting with the WorkBAT test (Workaholism Battery - Spence and Robbins, 1992), the work addiction risk test, WART (Work Addiction Risk Test - Robinson, 1999), DUWAS (Dutch Work Addiction Scale -Schaufeli, Shimazu and Taris, 2009) and BWAS (Bergen Work Addiction Scale Andreassen, et al., 2012). Job satisfaction is considered a psychological factor underlying productivity, being correlated with increased turnover, as indicated by many research studies in the field (Hom, et al., 1992;Tett and Meyer, 1993;Lee, et al., 1999). Amfiteatru Economic

Research Methodology
Statistical hypotheses were inferred from the literature. Thus, while Ross and Zander (1957) consider that personal needs have a positive impact on employee's satisfaction, Roelen, et al. (2008) demonstrate that the workload is a decisive factor in terms of job satisfaction. At the same time, Possenriede and Plantenga (2011) argue that employee's satisfaction is primarily a result of the time spent at work, and Thorsteinson (2003) considers that there is a close correlation between work attitude and job satisfaction. Last but not least, physical and mental state is one of the prerequisites for a good mood and job satisfaction (Koustelios and Bagiatis, 1997).
Thus, in order to achieve an economic modelling of the HWI impact on the employee's satisfaction, the following five statistical hypotheses were formulated: The objective of this paper is to validate the conceptual model by quantitative econometric analysis of the data collected through a questionnaire. The study also analyses the impact of the econometric model, presented in the next section, represented by personal needs, workload, time allocated to work, attitude towards work and physical and mental condition on performance and job satisfaction.
The questionnaire was addressed to managers and employees with higher education and at least one year of experience, within the specialized departments of multinational companies from Romania, that offer consulting and audit services. This sector was selected because it is the area in which the degree of investment in labour is one of the highest. The targeted companies are the biggest in the Romanian audit and consultancy and the respondents were selected by the method of systematic random stratification. Both managers and employees, as project coordinators, were selected in the sample in order to have a broader perspective on job satisfaction. Kadam and Bhalerao (2010) consider that, in the case of an unknown population, for a 95% confidence interval and a maximum error of ± 3%, the sample size must be at least 385 respondents. Thus, data were collected from 450 respondents, out of which 410 were valid, whereas 40 questionnaires were incomplete or incorrectly completed. The answers to the questions in the questionnaire were coded using a 7-point Likert scale, where the choices ranged from very little = 1 to very much = 7. The questionnaire was pre-tested to ensure that the format, language and order of the questions were appropriate.
The distribution of the respondents in the sample, depending on the position they hold within the selected company, could be seen in table no. 1. As can be seen from the table above, the sample consisted of 410 managers and employees from the analyzed enterprises, stratified by sex, income, age and level of education, who were interviewed directly, the questionnaire containing 30 questions.
For the data collection, the method of systematic stratified sampling was used, using a questionnaire with both closed and open questions. The data collection took place between February 1 and March 30, 2020. The main problem in data collection was the reluctance of respondents, at the level of execution, to complete questionnaires. In order to eliminate the concerns regarding the method of data collection, the information was anonymized and presented in an aggregated manner, and personal data were deleted. The main objectives pursued in conducting the study were: determining personal economic or professional needs and the physical and mental state of the interviewee.
To analyze the impact of heavy work investment on the employee's satisfaction, a quantitative analysis based on a survey was conducted using a questionnaire. The statistical hypotheses described above were tested using the Partial Least Square -Structural Equation Modelling (PLS-SEM) method, and the data were analyzed using the SmartPLS 3.3.2 statistical software (Ringle, 2020).

NP_1
The importance given to the professional career. NP_2 Existence of career development plans in the near future.

NP_3
Overtime is remunerated in proportion to the effort made. NP_4 A large amount of the budget is allocated to dependents. NP_5 Existence of bank loans contracted during the activity within the enterprise.

Workload (VM) VM_1
The solution of the received works exceeds the normal time of 8 hours daily. VM_2 The assigned tasks exceed in complexity and / or volume the average number of works received by the other departments in the company. VM_3 The large number of tasks received distracts attention VM_4 The complexity and / or the large number of works that come back makes us feel overwhelmed.

Working time (TM) TM_1
The work schedule is flexible. TM_2 Work tasks can be accomplished by working from home. TM_3 The number of hours worked daily exceeds 8 hours.

Attitude towards the work AM_1
Existence of dependence for the work performed. AM_2 The existence of states of stress or the feeling of pressures in solving your tasks. AM_3 The work done brings satisfaction. AM_4 The work done in your tasks involves passion.

Physical and mental condition SF_1
In the last years of activity there have been diseases caused by stress. SF_2 Existence of a high level of physical exhaustion SF_3 Mental pressure is felt every working day The independent variables described in Table no. 2 is the latent variables of the model, while the reflexive dependent variable is HWI. This is described in Table no. 3 and is determined by five proxy variables: "Degree of satisfaction with the workplace", "Degree of identification with the workplace" (Boles, et al., 2001), "Mental, emotional, workplace comfort", "Pride to be part of the organizational team" and "Degree of professional achievement" (Motowidlo and Kell, 2012). According to Weiner (1985), if the independent variables act negatively on the dependent variables that form HWI, we are in the situation of workaholism, if they act positively, we are in the situation of employees trained positively in work activities.
The dependent variable and the independent variables, described above, lead to the formation of the structural model (figure no. 1).

Figure no. 1. The structural model that incorporates the modelling of the impact of massive investments in labour satisfaction
The Model will be tested and then validated in the following section.

Results and discussions
Before the actual analysis of the model, the degree of significance of the variables of the conceptual model is checked. Thus, the validity, reliability and internal consistency of the collected data are analyzed. For this, the coefficients Cronbach's Alpha and Dillon-Golsteins' ρ are calculated and interpreted (Tenenhaus, et al., 2005).   Hair et al. (2013) show that there are no problems of collinearity between the independent variables in the model. It can also be seen that the values of the Cronbach's Alpha and Dillon-Golsteins' ρ coefficients for the 5 independent variables have values greater than 0.7. Thus, we can conclude that the independent variables in the model are significant.
The econometric study was performed using the PLS-SEM method. It consists of the analysis of two sub-models, namely the measurement model, the "outer model" and the structural model, the "inner model". The PLS-SEM method was chosen to perform the data analysis, being preferred to other similar methods, such as "Covariance Based -Structural Equation Modelling" (CB-SEM), because it is less sensitive to asymmetric distributions, small samples or the presence of multicollinearity, but at the same time, it is even more robust (Hair, et al., 2013).

The Measurement Model
The measurement model was evaluated using the convergent and discriminant validity method. Mathematically, the relationship between latent and reflexive variables can be determined using SEM (Jackson, 1969). where: δ is the latent dependent variable; θ is the latent independent variable; x and y are the observed variables; Ω x and Ω y are the matrix coefficients; ε x and ε y are the residual variables.

Convergent validity
Nacaskul (2017) states that convergent validity must be examined when a latent variable explains a statistically significant part of the variation of its constructs. In Figure no. 2, one may observe that the coefficients related to the variables from the measurement model are obtained with the help of SEM-PLS equations.

Figure no. 2. PLS-SEM 1 Measurement Model Source: Data analysis was performed by the authors using SmartPLS software 3.3.2.
Chin (2010) considers that constructs for which the values of the coefficients are less than 0.5 should be excluded from the measurement model. Consequently, we exclude the variables from the model: np_4, np_5, vm_3, vm_4 and am_2. After excluding these variables from the model, using the PLS-SEM equations and using the SmartPLS 3.3.2 software, we obtained the second measurement model. As it could be seen in the figure no. 3, the values of all the coefficients of the reflexive variables in the model are greater than 0.5, which means that all these coefficients are valid. Thus, the convergent validity of the model is confirmed.

Divergent Validity
Divergent validity can be tested using the Average Variance Extracted (AVE) indicator. Fornell and Larker (1981) state that if the square root of the values of the correlation coefficients between the latent variables is lower than the values of the AVE indicators related to the latent variables, then the discriminant validity is confirmed. In table no. 5 these values can be observed. As can be seen in the table above, the square roots of the correlation coefficients are smaller than the AVE indicator values for all latent variables, which confirms the discriminant validity of the measurement model.

The Structural Model
The structural model, inner model, was described in Figure no. 2. It shows us the connection between the exogenous variables in the model and the endogenous variable. The structural model (Claes and Cha, 1994) is described by the following equation: δ -the vector related to the endogenous latent variables; ϕ -the vector related to the exogenous latent variables; Σ -the vector related to the residual variable; Λ and Γ -path coefficient matrices.
The values of the coefficients of the structural model are estimated using structural equations. The value of R-square is 0.668, which means that 66.8% of the variability of dependent variables is explained by the model. Regarding the estimated values of the coefficients of the structural model, we note that all are positive, which means that all latent variables have a positive impact on HWI. Thus, out of the five indicators, the variable with the greatest impact on HWI is attitude towards work (factor 0.768), followed by working time (factor 0.647), personal needs (factor 0.594), workload (factor 0.331) and physical and mental condition (factor 0.123). The reflexive indicators of the dependent variable HWI have values higher than 0.5, which means that the variables (hwi_1, hwi_2, hwi_3, hwi_4 and hwi_5) are statistically significant. The high values of the reflexive variables, namely Degree of satisfaction at work, Degree of professional achievement show us that satisfaction and the feeling of professional achievement are well represented by the two indicators. In general, multiple factors improve model performance and increase reliability compared to single factors. In this model, the dependent variable is composed of five variables. As with formative variables, a reflective variable has a greater impact if it has high values of coefficients, reflecting stronger measurement paths. This is confirmed by the high values of the five coefficients of the dependent variables.
In addition, to test the six statistical hypotheses which were described in the previous section, a Bootstrap Test was performed on a sample of 300 respondents to generate the values of the t test (t-test) and the standard deviations (standard deviation) of the model parameters. According to statistical theory, bootstrapping allows the allocation of precision measurements to sample estimates. The results regarding the testing of the statistical hypotheses can be observed in table no. 6.  (Poikkeus, et al., 2020;Liss, et al., 2018), but invalidates the assumption that physical condition is an important factor of job satisfaction and work performances (Pi, et al., 2016). Thus, we can say that economic or professional needs, workload, time spent at work, attitude towards work and the physical and mental state of the employee are determining factors in terms of HWI.
These results, from the perspective of HWI, are in line with other previous studies on HWI, both in Romania and in other Central and Eastern European (CEE) countries. Thus, several studies in CEE countries indicate that employees (from these countries) tend to invest heavily in work. The large-scale incidence of high workload in CEE countries is supported by Taipale, et al. (2011) regarding the demand for jobs in eight European countries. The incidence was particularly high in Bulgaria and Hungary (as well as Sweden, Germany and the United Kingdom), and the high volume of work proved to be a strong predictor of job satisfaction in these countries (Schaufeli and Bakker 2004).
With respect to Romania, previous studies (Butucescu and Uscătescu, 2013) regarding the impact of investments in labour on employee satisfaction have indicated that the HWI index for Romanian employees is significantly influenced by social, economic status and demographic factors. According to the study, Romanian employees in the private sector have a higher degree of job satisfaction than those working in the public sector.

Conclusions
Starting with the economic modelling to determine the impact of the heavy work investment on the work satisfaction using structural equations SEM-PLS, this research paper identifies and evaluates the relationship between the determinants of the conceptual management model in analyzing the impact of the heavy work investment on satisfaction and employee's satisfaction. The results indicate that a significant proportion of the variability of the dependent variables is explained by the independent latent variables. Attitude towards work is thus one of the main criteria that influences the degree of employee satisfaction and their performance. Also, two other important factors that converge towards a high level of performance refer to the time allocated to the work and the workload. The results of the evaluation take into account other criteria, such as personal economic or professional needs and the physical and mental condition of employees.
Analyzing the results of the questionnaire in terms of modelling the variables, it can be noted that the evaluation of the impact of some determinants for the HWI indicator is essential in the management of management processes. The success of a process depends on the extent to which the organization, knowing the impact factors highlighted in section 3, acts to reduce or minimize the risks of physical and/or mental exhaustion of employees, with a large volume of work and excessive time allocated over the program. Early identification of these risk factors can help the organization plan and distribute specific activities in a balanced way. The intervention of the organization in order to correct some deviations can have the expected positive effects both in terms of performance obtained by the organization and in terms of human resources, engaged in activities.
In the presented model, the results of the study, analyzed using the modelling of structural equations, converge to the conclusions of the relevant studies in the literature, emphasizing that only in the case of passionate employees, the determinants of the heavy work investment can positively influence satisfaction and growth of the performance of organizations in Romania. The positive correlation could be attributed to the fact that the respondents were mostly leaders, which explains their commitment to the projects carried out, leading to outstanding professional results at the organizational level. Among the practical implications of the study results are the performances of the human factor, rewarded by the organization, which contributes to creating the employee profile and shows a positive attitude towards the work done.
Thus, from the analysis of the model, it is highlighted that we are in the situation of employees positively trained in the activities carried out (work engagement), which, also, could be considered a limitation of the model, as the questionnaire was addressed mainly to people with management positions that developed a positive attitude towards work and determined, implicitly, a positive HWI indicator, indicating on the high degree of satisfaction and, implicitly, on the performance.
Another limitation of this study is the relative reduced number of survey respondents. Other limitations of this research could be related to the subjective responses of the respondents and the number of constructs and reflexive variables. These limitations could be overcome in future research by increasing the sample size and balancing it between the level of execution and management, as well as by completing the number of constructs and reflexive variables, respectively by introducing open-ended questions in the questionnaire.
Further studies on the modelling of the impact of massive investment in labour on employee satisfaction and organizational performance should be extended to other professional domains and research may include some likely macroeconomic effects. Moreover, future research could analyze changes in respondents' work habits, in order to develop the current model. In addition, in terms of external factors, the model developed in this article could benefit from an extension by analyzing how HWI acts, in different organizational settings, based on variables, such as type of work, type of requirements and balance of resources, as well as work settings that reflect variations in work culture and management, such as the type of industry, public versus private organization, teamwork versus authoritarian management.