The effect of overtime, job stress, and workload on turnover intention

Article history: Received: June 30, 2020 Received in revised format: June 3


Introduction
Human Resources (HR) is a one of the most critical assets of any company. With proper management of human resources, the company will continue to achieve good development as well. Many factors affect human resources in any particular firm. These factors differ from one company to another company. However, they can be predicted by general factors that have been put forward through various theories explained by human resource management experts based on experience and research studies. A company's performance is inseparable from the performance of each individual in the company itself, as stated by Junaidi et al. (2019). Efforts to improve individuals in the company are carried out in various ways. For example, through the provision of good compensation, training, a conducive work environment, carrying out effective education and training so it is expected that each employee can achieve satisfaction and be committed to the completion of the performance, which is the goal of the company. However, if this effort is not carried out by the company, it would have a negative effect on the development of each individual in the company. The effect influences employees' desire to move or leave the company (turnover intention). Simamora (2006) mentions that turnover intention is a movement by the membership limit of an organization. Work transfers, in this case, are avoidable voluntary turnover and unavoidable voluntary turnover. According to Tett and Mayer (1993) in Samad (2006), the intensity of turnover is the tendency or intention of employees to look for other job alternatives in different organizations consciously. According to the explanation, it could be concluded that the turnover intensity is the tendency or the existence of an individual plan to leave the company where he/she works at this time. Turnover intensity measures the size of the employee's wants to make a turnover from his/her company. The higher the desire of employees to leave the company, the higher the actual turnover that occurs in the company. The phenomenon that often occurs in a company is the high level of employee turnover, especially in private companies. The higher turnover means more frequent employee turnover. Of course, it will harm the company. Turnover has a negative impact when viewed from the perception of an employee at work, arguing that high stress will eliminate their creative ideas and reduce the enrichment of work empowerment (Javed et al., 2014). The company will incur more costs for recruiting new employees if the turnover rate is high (Kurniawaty et al., 2019;Nawaz & Pangil, 2016;Aburumman et al., 2020).
High levels of stress experienced by employees could have a negative impact on the company. It is where employees with high levels of stress cannot do their jobs properly, which in turn, will bring an adverse impact on the productivity of their workforce and affect the performance of the company itself.
This research was conducted based on the gap phenomenon regarding overtime at PT. Ungaran Sari Garment, Congol Unit, Semarang Regency. Based on the interviews, it was known that the company imposed overtime hours for employees if there were tasks that had not been completed, or other jobs required employees to complete them on the same day. Overtime hours were calculated after the end of regular working hours, after 14:30 WIB until finished. For that overtime, it was usually for every day. The workday count was Monday-Saturday in one week. The phenomenon of the end of the month required employees to work longer hours than regular working hours since the company imposed heavy work at the time. Also, it often happened when a company was undergoing shipments and high orders, and then employees had to be able to fulfil the targets and complete work in accordance with the deadlines made by the company. The applicable company rest hours were 30 minutes every day, with a schedule determined by the company. It changed every 21st of the month. It was because the company applied a rest hour system by taking turns so that production remained optimal.
Company data in terms of overtime, job stress, and workload showed relatively high results in employees of the sewing department of PT. Ungaran Sari Garment, Congol unit. Seeing the results of these variables, which were many causes of turnover intention, researchers were interested in examining more deeply about turnover intention in terms of overtime, job stress, and workload on the employees of PT. Ungaran Sari Garment in the sewing department. Departing from the problems above, this study aims to know the effect of overtime, job stress, and workload on turnover intention.

Overtime
According to Thomas (1992) in Hasanah (2016), overtime is work time that exceeds 40 hours per week, and lasts for at least three consecutive weeks. Hana et al. (2005) defines overtime as time that exceeds 8 hours in a day and 40 hours every week. Thomas and Raynar (1997) in Hasanah (2016) state that overtime results in various effects on the job performance. In addition to the fatigue factor of employees, overtime schedules also cause various aspects related to the work that must be added and accelerated procurement, such as materials, equipment, technical and administrative staff that meet the needs for work facilities. If in doing work late is due to poor management, then overtime expectations will not be efficient because management will find it increasingly difficult to provide support for the implementation of work (Tan, Sim, Goh, Leong, & Ting, 2020).

Job Stress
According to Badeni (2017), stress is a feeling of displeasure, anxiety and anxiety that can cause unwanted thoughts, emotions and physical states. Stress can also be defined as a response that easily adjusts to external circumstances that can result in, psychological, behavioral and physical deviations for members of the organization. Ahmad et al. (2012), Arijanto, Marlita, Suroso, & Purnomo (2020), Fukui, Rollins, & Salyers (2020), Nanda, Soelton, Luiza, & Saratian (2020), Yunharmini & Supartha (2020) proved that employees with job stress have high intentions to retire. Beehr (1978) in Suharnomo (2016) states stress as a situation that will force someone eliminate themselves due to psychological or physiological changes, where the person is forced to deviate from normal function. Some indicators that affect job stress are: a. Psychologically, prolonged stress will impact the person, b. Physical circumstances where employees will experience fatigue due to overworked, c. Behavioral circumstances in which employees will have a unpleasant circumstances during work hour.

Workload
According to Suwatno and Priansa (2011), workload is an activity that must be completed by a worker using technical analysis of positions or other management techniques within a certain period of time to obtain information about work efficiency and effectiveness. According to Kurnia (2010), workload is a process of analyzing the time spent by a person or group in completing work tasks. Workload indicators according to Suwatno and Priansa (2011) are work hours, educational background and type of work. According to Arshad et al. (2020) and Gibson (2009), workload can have a negative impact on employee health due to too much pressure so that many employees decide to change jobs.

Turnover Intention
Intention is a desire arising from an individual to have an activity, while turnover is the cessation of a worker from their companies. Therefore, it is a tendency to stop working from work voluntarily according to their work choice. There are different factors that influence turnover, such as labor market; workspace conditions, wages, work skills, supervision; intelligence, attitude, past, gender, interests, age and length of work as well as individual reactions to their jobs (Zeffane,1994dalam Yuniar, 2008. According to Dipboye (2018), several indicators of measurement of turnover intention are: a. Thinking of quitting: Employees think about leaving the job or staying in the work environment. This begins with the work dissatisfaction felt by the employee, then the employee thinks of leaving the workplace. b. Intention to search for alternatives: This reflects the individual's desire to find work for other organizations. If the employee has started thinking about quitting his job and try to find a new job. c. Intention to quit: This means an individual who intends to leave. The employee tends to retire if founds a better job.
According to the literature review, some of the hypotheses developed are as follows: H1: Overtime has a significant effect on turnover intention. H2: Job stress has a significant effect on turnover intention. H3: Workload has a significant effect on turnover intention. H4: Overtime, work stress and workload simultaneously have a significant effect on turnover intention.

Research Methods
This research belongs to quantitative model. In this study, the population was the employees of PT. Ungaran Sari Garment in the sewing department, consisting of 540 people. The sample was 85 employees of PT Ungaran Sari Garment. The sampling technique used was purposive sampling technique to the respondents. The purpose of using the purposive sampling technique was that not all samples had criteria that were consistent with what researchers have done. Therefore, this method has some criteria that need to be met by the samples involved. The criteria used as research samples are as follows: 1. Mileage to work. 2. Age of employee (20-25 years). 3. Working period (1-5 years). Data collection method were observation, documentation, and questionnaires. The questionnaire in this study employed closed questions and a Linkert scale, a measuring tools to measure the indicators about social phenomena. The dependent variable was turnover intention (Y), therefore, the independent variables were overtime (X1), job stress (X2), and workload (X3). The regression model in this study is ordinary least square regression model. The analysis estimates the population average or the values of the X variable according to the value of known independent variables using the parameter (Ghozali, 2005). Thus, multiple linear regression analysis tries to analyze turnover intention as the dependent variable in terms of overtime, job stress, and workload as independent variables. According to Ghozali (2011), to determine whether there is an impact for X to Y variables, testing the hypothesis proposed in this study was conducted. To determine the impact of the independent variables to the dependent variable, the F-test was performed, and partially, the t-test was carried out. Also, the terminated coefficient (R 2 ) was to determine how much the independent variables' ability to verify the dependent. T-test was a test used to determine the significant influence of independent variables (X1: overtime, X2: job stress, X3: workload) partially or individually on the dependent variable (Y: turnover intention). Testing each regression coefficient is significant if the absolute value of tcount > t-table or significant probability value is smaller than 0.05 (selected confidence level). Otherwise, it is insignificant when t-count < t-table or significant probability value is greater than 0, 05 (selected confidence level). F-test was employed to test whether the model used to verify the effect of independent variables (X1: overtime, X2: job stress, X3: workload) partially or individually on the dependent variable (Y: turnover intention) was accepted or not. Joint testing between the independent variables and the dependent variable was by looking at the significant level (F) at α = 5%. Testing each regression coefficient jointly is said to be significant if the absolute value of F-count > F-table or significant probability value is smaller than 0.05 (selected confidence level). Coefficient of determination (R 2 ) was employed to measure the overtime, job stress, and workload in explaining the variation of the dependent variable (turnover intention). The coefficient of determination is between zero (0) and (1). The higher R2 value means the independent variables' ability to explain the dependent variables provides the indicator that needed to predict the variables.

Results and Discussion
Multiple linear regression analysis was used to calculate the magnitude of the effect of overtime, job stress, and workload on turnover intention. Based on the problem formulation and the hypothesis stated earlier, the results were obtained, as in the following table below: From the regression equation, it could show that overtime, job stress, and workload had a positive effect on turnover intention. With the following information: It was obtained a constant value of 2.658, which means that if all the independent variables (overtime, job stress, and workload) did not change, the turnover intention was positive. The regression coefficient value of overtime variable toward turnover intention was 0.303 with a positive sign. It means that if one-unit turnover went up, turnover intention automatically also would rise, with the condition that other variables were considered fixed. The regression coefficient value of the job stress variable toward turnover intention was 0.532 with a positive sign. It indicates that when one-unit of 0.532 job stress went up, it automatically increased the turnover intention, with the condition that others were considered fixed. The regression coefficient value of the workload variable toward turnover intention was 0.485 with a positive sign. It signifies that if one-unit of 0.485 workloads raised, turnover intention automatically increased as well, only if other variables were considered fixed. The most dominant independent variable affecting turnover intention was job stress, with a regression coefficient of 0.532. The second one was a workload variable, with a value of 0.485, and the last was an overtime variable, with a value of 0.303. The results for overtime variables from the table above obtained t-count of 2.859 > t-table of 1.66342, with a significance value of 0.005 < 5%. It indicates that there was a positive and significant effect between employee overtime and turnover variables. Thus, the first hypothesis was accepted. With an increase in overtime, it would impact the increase in employee turnover of PT. Ungaran Sari Garment in the sewing department. Conversely, the lower the overtime, the lower the turnover of employees of PT. Ungaran Sari Garment in the sewing department. It is in line with previous research conducted by Hasanah & Suharnomo (2016), which stated that time demands of work had an indirect effect on employee turnover intention if mediated by the job stress of PT. Bank Tabungan Negara (Persero) Tbk. Harapan Indah Branch, Bekasi. Thomas and Raynar (1997) in Hasanah and Suharmono (2016), said that the overtime resulted in the emergence of various effects on the implementation of work. In addition to the fatigue factor of workers, overtime schedules also cause various aspects related to the implementation of work to be increased and accelerated procurement, such as materials, equipment, technical and administrative staff in charge of meeting the needs of work facilities. If working late due to poor management, the application of overtime will not be efficient, because management will be increasingly difficult to provide support for the implementation of work. Based on Table 2, it can be obtained that t-count for job stress was 6.935 > t-table (df = n-k = 85-2 = 83) of 1.66342, with a significance value of 0,000 less than 5%. It signifies that there was a positive and significant effect of job stress on employee turnover. According to Robbins (2003), if the individual experiences a high level of stress, he/she consequences the emergence of physiological, psychological, and behavioral symptoms. Further, Beehr (1978) in Hasanah & Suharnomo (2016) stated stress as a situation that will force someone to deviate from normal function due to physiological changes by themselves, such that the person is forced to deviate from normal function. These results are in line by Sa'adah, & Praseti (2017), they state that there was a positive effect between job stress and employee performance. Then, based on Table 2, it can be obtained that t-count of the workload was 6.335 > t-table (df = n-k = 85-2 = 83) of 1.66342, with a significance value of 0,000 less than 5%. It means that there was a positive and significant effect of workload on employee turnover intention. It is in line with the research results conducted by Putra and Prihatsanti (2016) that workload had a significant effect on turnover intention. According to Gibson (2009), the workload could also have a negative impact on employee health due to too much pressure, so that many employees decided to change jobs. Excessive overtime could also trigger employees to change jobs.  3 above shows that the F-count was 98.053, while the F-table was 2.48, with a significance level of 0.05 (the value of F-count > F-table. It can be concluded that the independent variables of overtime (X1), job stress (X2), and workload (X2) influenced the dependent variable of turnover intention jointly. Thomas (1992) in Hasanah & Suharnomo (2016) defines overtime as work time that exceeds 40 hours per week and lasts for at least three consecutive weeks. Thomas & Raynar (1997) in Hasanah & Suharnomo (2016) said that overtime resulted in the emergence of various effects on the implementation of work. In addition to the fatigue factor of workers, overtime schedules also cause various aspects related to the implementation of work to be increased and accelerated procurement, such as materials, equipment, technical and administrative staff in charge of meeting the needs of work facilities. According to Badeni (2017), stress is a feeling of displeasure, anxiety, and nervousness that can cause unwanted thoughts, emotions, and physical states. Stress can also be defined as a response that easily adjusts to external circumstances, which can result in physical storage, psychological and behavioral irregularities for members of the organization. Besides, according to Robbins (2003), when individuals experience high levels of stress, the consequence is the emergence of physiological, psychological, and behavioral symptoms from the individual. Meanwhile, according to Suwatno (2011), the workload is an activity which must be completed by a company and do it systematically with job analysis knowledge, workload analysis techniques, or other management arts within a specified period to obtain information about work efficiency and effectiveness of an organizational unit. Too much workload will undoubtedly cause turnover of employees of PT. Ungaran Sari Garment in the sewing department is getting higher. Based on the regression calculations, the value of Adjust R Square was 0.776. It indicates that the percentage of all independent variables' influence (overtime, job stress, and workload) could explain the dependent variable (turnover intention) of 77.6%. Meanwhile, 22.4% was explained by other variables not proposed in this study.

Conclusion
Based on the results of the research and discussion above, it can be concluded that partially, the overtime had a positive and significant effect on turnover intention. The value of the regression coefficient for the overtime toward turnover intention was 0.303, with a positive sign. It signifies that if one-unit of turnover went up, the turnover intention would automatically increase, only if other variables were considered fixed. Job stress had a positive and significant effect on turnover intention. The regression coefficient value of the job stress toward turnover intention was 0.532, with a positive sign. It means that if one-unit of 0.532 job stress increased, it automatically increased the turnover intention, with the condition if other variables were considered fixed. The workload had a positive and significant effect on turnover intention. The regression coefficient value of the workload toward turnover intention was 0.485, with a positive sign. It indicates that if one-unit of 0.485 workloads raised, it automatically increased turnover intention if other variables were considered fixed. Besides, the F-test was found that the variables of overtime, job stress, and workload jointly and simultaneously had a positive and significant effect on turnover intention. The advice that can be submitted in this study is that too much workload causes an increase in employee turnover of PT. Ungaran Sari Garment, in the sewing department. Therefore, the management of PT. Ungaran Sari Garments should add employees in the sewing department so that the overtime that occurs at this time can be reduced without exceeding the delivery time specified by the buyer. Furthermore, from the questionnaire that has been distributed, the job stress that dominates was a conflict between employees. It is therefore expected that management can mediate conflicts between employees, for example, with direct discussions with people involved in open discussions to get an agreement or can be done utilizing effective communication to prevent conflicts from reoccurring. Thus, it is expected that employees at PT Ungaran Sari Garment in the sewing department will be controlled, and there will be no turnover intention that will interfere with the company's performance. For further research, to improve this research, it can add research variables, such as, organizational commitment, job satisfaction, organizational culture, and other variables in examining the factors that influence turnover intention. Researchers can also use other methods in researching turnover intention to obtain more varied results.