Work Stress of Online Motorcycle Taxi Drivers: The Role of Coworker Support, Autonomy and Affective Occupational Commitment

Purpose – Online motorcycle taxi driving is a field of work that has recently become a work option for some people. However, it could be this choice because of the compulsion that can cause stress. Drawing on social cognitive theory, this study investigates the antecedents of online motorcycle taxi drivers work stress, by considering coworker support, autonomy, and affective occupational commitment. Methodology – Data collection was carried out by surveying questionnaires, with 89 online motorcycle taxi drivers in Surabaya as respondents. Data analysis was done using Partial Least Square-Structural Equation Modeling (PLS-SEM) . Findings – Supportive of social cognitive theory, this study finds that coworker support and autonomy positively influence affective occupational commitment, while affective occupational commitment is likely to reduce work stress. Coworker support and autonomy have no significant direct effect on work stress. Affective occupational commitment mediates the relationship between coworker support and work stress. Originality – Previous studies on online motorcycle taxi drivers tend to identify organizational commitment and examine the effect of work stress on work attitudes. The present study considers affective occupational commitment as an antecedent of work stress, and a mediator of the influence of coworker support and autonomy, on online motorcycle taxi drivers' work stress in Surabaya .


Introduction
In recent years there has been a phenomenon of the emergence of various companies in the online transportation sector, such as Uber, which operates in 65 countries across the world. Grab, Didi, Olf, and Gojek operate in various Asian countries while Gett and Bolt are prevelant in Europe. Lyft is commonly used in America, Cobify in South America, and Jrney in South Africa (Okun, 2019). Consequently, a new job has been created, namely online motorcycle taxi drivers. In Indonesia, online motorcycle taxi company Grab and Gojek are the most popular choice because of the benefits they provide (Kompas.com, 2021). Online motorcycle taxi drivers can be presence of coworkers support emotionally with the attention and willingness to listen to colleagues (Chou & Robert, 2008). The relationship between them also shows high solidarity (Widiatmoko, 2021). Especially during the pandemic, caring for fellow drivers is important.
Autonomy is the extent to which individuals have the freedom and independence to manage their own work activities (Hackman & Oldham, 1974;Runhaar et al., 2013). In working as online motorcycle taxi drivers, individuals have more freedom to manage their work (Muthohar, 2020). Therefore, it is possible that coworker support and autonomy can reduce the work stress of online motorcycle taxi drivers.
Secondly, this study considers a personal factor, namely affective occupational commitment, as a mediator in the relationship between situational factors (coworker support and autonomy) and work stress. Affective occupational commitment is an individual's emotional attachment to his/her job (Allen & Meyer, 1990). Furthermore, occupational commitment is an attitude and psychological attachment between one's beliefs and acceptance of the values and choices of a job (Yuan et al., 2014). Working as online motorcycle taxi drivers is a form of nonstandard and self-employment, so it is important to analyze the role of this occupational commitment. Moreover, research results show that this type of worker is less committed, but other studies have found that they are also committed to their work (Omar, 2013). Although several studies have found a relationship between employee commitment and coworker support and autonomy, these studies focus on organizational commitment (e.g., Colla et al., 2019;Kmieciak, 2021). In addition, individuals with high occupational commitment will experience less stress (Lee et al., 2000). Therefore, online motorcycle taxi drivers who experience coworker support and autonomy may increase their occupational commitment and ultimately reduce work stress.
Although autonomy, coworker support, and occupational commitment are all factors which can reduce work stress for online motorcycle taxi drivers, there are not many studies that identify the relationship between them, especially within Indonesia. Testing of work stress antecedents for online motorcycle taxi drivers is still hardly done. Existing studies tend to identify work stress as an antecedent of other variables (e.g., Bayuaji, 2021;Pertiwi, 2020). Furthermore, studies on the commitment of online motorcycle taxi drivers tend to test it as an antecedent of performance (e.g., Rosita et al., 2018;Mangifera & Isa, 2017). These studies also identify commitment to the organization, not the job (i.e., occupational commitment). As far as we know, there are no studies that have investigated the relationship between coworker support, autonomy, affective occupational commitment, and work stress from online motorcycle taxi drivers in a singular research model.
Work stress is an individual's reaction to unwanted or uncomfortable workplace conditions (Montgomery et al., 1996). The reaction to stress can be referred to as strain . The reaction can be physical, such as headaches and cardiovascular illness, or psychological, like anxiety and insomnia Sebastian, 2013). Individual stress can be shaped by individual and situational factors (Gardner & Fletcher, 2009). Based on social cognitive theory, individual behavior is influenced by assessing the interaction of various factors in their social environment, such as task context and support from others (Bandura, 1991). Referring to this theory, in the cognitive process, stress reactions may arise if individuals assess and perceive an event or object as a threat (Sebastian, 2013). The emergence of work stress is the result of an individual's cognitive assessment or individual perception of the work situation (Smithikrai, 2014), such as autonomy (Dartey-Baah et al., 2020) and social support (Smith et al., 2017).
Social support can reduce the individual's perception of a significant threat from the source of stress, so that it can increase the belief that they can overcome the existing challenges (Smith et al., 2017). Social support can be a barrier that helps deal with difficult situations (Smithikrai, 2014) and is an important resource for releasing emotions (Gardner & Fletcher, 2009), and can reduce stress (Simons & Barone, 1994). Furthermore, social support can indirectly influence the belief in managing environmental demands (Bandura, 2002). Coworkers are parties who can provide social support in the form of emotions, by listening, showing concern (Beehr et al., 2000), being understanding, providing comfort (Gardner & Fletcher, 2009), and giving sympathy (Yousaf et al., 2020). Empirically, social support such as from coworkers is negatively related to work stress (e.g., Beehr et al., 2000;Yousaf et al., 2020). H1: Coworker support is negatively related to work stress Autonomy is the freedom individuals have in doing their work and the ability to control the time to work, how the work is done, and any decision making (Morgeson & Humphrey, 2008). Referring to social cognitive theory, the perception of control over work can lead to resources that increase confidence in doing work (Ford et al., 2020). Consistent with this theory, the higher the autonomy, the more flexibility and energy individuals have in doing their jobs, and helping them deal with the demands of daily tasks (Emre & De Spiegeleare, 2021). Furthermore, with autonomy individuals can be more relaxed and confident in their jobs (Zheng et al., 2020). Therefore, autonomy can act as a resource for individuals dealing with unpleasant work or situations. Autonomy in work is a form of job-relevant task characteristics as online motorcycle taxi drivers. People who spend much time driving on trips are prone to experiencing stress because they can face traffic congestion, delays, and the dangers of driving (Emre & De Spiegeleare, 2021). However, with autonomy, people can manage their schedules and work on their tasks to reduce stress (Emre & De Spiegeleare, 2021). Empirically, autonomy was negatively related to work stress (e.g., Lange & Kayser, 2022;Schiff & Leip, 2019). H2: Autonomy is negatively related to work stress In various studies, several terms such as professional commitment, occupational commitment, and career commitment are used to refer to work commitment (Lee et al., 2000). In the context of online motorcycle taxi drivers, it is important to identify their commitment to work. Lee et al. (2000) explain that the term 'occupation' has a more general meaning because it can include both professional and non-professional work, and refers to work where individuals earn income over a certain period of time. Commitment to work (or occupational commitment) is a psychological link between individuals and their work, based on affectionate reactions to the job (Lee et al., 2000). Individuals who have an affective commitment have a strong desire to stay in their occupation (Meyer et al., 1993). They have an emotional attachment to their work (Blau, 2009), identify with themselves, and have positive experiences with their work (Lee et al., 2000). Employee commitment can be built by several factors, such as social support and autonomy.
Social support can foster a positive attitude (Yousaf et al., 2020). As a form of social support, coworkers are those who experience the same work situation (Rousseau & Aubé, 2010). This makes colleagues parties who can better understand the problems faced by individuals. Support from colleagues who show concern, encouragement, and attention can form a pleasant work experience, enthusiasm, and affective commitment (Kmieciak, 2021). This pleasant experience can form an emotional attachment to the company. Furthermore, the support of their coworkers can play a role in their career development and commitment to their career (Lin, 2017). Thus, in the context of online motorcycle taxi drivers, it is also possible that the support of coworkers can create a sense of pleasure and comfort at work and subsequently make them attach themselves to their work. Empirically, Yu et al. (2021) found that social support has a positive effect on the commitment to work.
Autonomy gives individuals the power and control over their work and creates a sense of belonging (Colla et al., 2019;Lin & Ping, 2016). Jobs that provide autonomy may form individuals' judgment that the work situation is meaningful and enjoyable (Lin & Ping, 2016). Emotional attachment will emerge with autonomy in work (Colla et al., 2019). Emotional attachment is the basis of affective commitment (Lin & Ping, 2016). Thus, autonomy may increase affective occupational commitment. H3: Coworker support is positively related to affective occupational commitment H4: Autonomy is positively related to affective occupational commitment Referring to social cognitive theory, individuals' reactions can result from their attitudes or beliefs formed from their assessment of their social environment (Martin & Guerrero, 2020). Therefore, it can be argued that individuals' assessment of coworkers' support and autonomy can shape their commitment to work, and can further reduce their stress. Individuals' perceptions of job factors, such as social support and how the tasks are managed, may impact their commitment to work (Lin, 2017). Empirically, social support is positively correlated with a commitment to the profession (e.g., Yu et al., 2021;Lin, 2017). Several studies, e.g., Lee et al. (2000), Emre and De Spiegeleare (2021) and Jain and Duggal (2018) found a positive relationship between autonomy and employee commitment. However, except for Lee et al. (2000), these studies focus more on organizational commitment and not commitment to work. In addition, the study by Wilson et al. (2015) found that autonomy was not significantly related to emotional exhaustion and so, there may be an intermediate factor in the relationship between autonomy and work stress.
Employee commitment is negatively related to emotional exhaustion (Setti et al., 2016). Affective commitment to the organization is an important resource and a protector against stressful situations. It provides a sense of belonging so that it may withstand stress-generating factors (Setti et al., 2016). Individuals committed to the organization will be more accepting of stressful situations and therefore reduce their responses to the added stress (Park & Searcy, 2012). Empirically, affective commitment to the organization has a negative effect on burnout (Chambel & Carvalho, 2022) and negatively affects work stress (Abdelmoteleb, 2019). Setti et al. (2016) found that coworker support is positively related to affective commitment, and affective commitment is negatively related to emotional exhaustion. Furthermore, Park and Searcy (2012) found that affective commitment to the organization mediates the relationship between autonomy and mental well-being. Mental well-being is a construct that refers to the tendency of individuals who do not experience anxiety and stress. Although these studies consider individuals' affective commitment to their organization, Kobasa (1982) argues that persons' commitment to their job reduces their response to stress. In addition, occupational commitment is positively correlated with autonomy (Lee et al., 2000) and coworker support (Yu et al., 2021). Therefore, it can be argued that individuals who have an affective commitment to their work will have reduced work stress. Furthermore, individuals who perceive that their work provides autonomy and that their coworkers provide emotional support, will increase their commitment to work and also further reduce work stress. H5: Affective occupational commitment is negatively related to work stress H6:aThe relationship between coworker support and work stress is mediated by affective occupational commitment H7: The relationship between autonomy and work stress is mediated by affective occupational commitment Figure 1 below shows the relationship model investigated in this study. This model considers coworker support, autonomy, and affective occupational commitment as antecedents of work stress. The model also shows the mediating role of affective occupational commitment on the relationship between coworker support and work stress, and also between autonomy and work stress.

Research Methods
The research population is online motorcycle taxi drivers in Surabaya. This study uses convenience random sampling by distributing questionnaires. To determine the minimum sample size, this study used G*Power analysis (Hair et al., 2017). In determining the sample size, it is necessary to know the extent to which the effect of the independent variable has statistical power at a minimum of 80%, as required in social science research (Memon et al., 2020). This power test is needed to ensure that the data analysis results are free from errors, in accepting a hypothesis that should not be accepted, or rejecting a hypothesis that should be accepted (Widhiarso, 2012). In PLS-SEM, which uses OLS regression as the basis, it is recommended to determine the minimum sample size using the G*Power program (Hair et al., 2017). The G*Power program is commonly used in behavioral research (Faul et al., 2007). This program provides a power analysis in determining sample size by considering the number of predictors, power level, and effect size (Memon et al., 2020). Based on calculations using G*Power 3 analysis (Faul et al., 2007), the minimum sample size of this study is 68 data. To obtain respondents who meet the minimum number of samples, this study uses 'Google Forms' which were distributed to the respondent's friendship groups, and directly used a hard copy of the questionnaire. Data collection was carried out in two months.
Before collecting and analyzing data, especially testing construct validity, several things need to be done to strengthen the validity of a construct. Following the suggestions of Rahim et al. (2018) and Bhandari (2022), the current study identifies appropriate survey instruments that can be used to measure the constructs in this study. This study uses a survey instrument that has been used in previous studies. However, this study modifies the indicators used to suit the context of the study, i.e., online motorcycle taxi drivers. In addition, this study also provides operational definitions for respondents as a guide for responding to questionnaire indicators, and is designed in easy-to-understand sentences. After the questionnaire is ready, before being distributed to collect research data, a pilot test is conducted. The following is the survey instrument used in this study.
Coworker support is an individual's perception of the extent to which coworkers provide emotional support and show concern, measured by six indicators from Tews et al. (2013) (e.g., my fellow online motorcycle taxi drivers take time to listen to my concerns). Autonomy is an individual's perception of the extent to which his/her job provides freedom to work. This variable is measured by adopting four indicators of work autonomy from the study of Morgeson dan Humphrey (2006) (e.g., The job allows me to plan how I do my work). Affective occupational commitment is the extent to which an individual attaches his/her feelings to work. This variable adopts five indicators used in Blau's (2003) study (e.g., I am proud to be in this job). Work stress is an individual's experience of physical and psychological stress in their work situation, measured by seven indicators from Anderson et al.(2002) (e.g., As an online motorcycle taxi driver, I felt emotionally drained). Respondents will be asked to respond on a 5-point scale (point 1 = strongly disagree to point 5 = strongly agree).
Data analysis is carried out using Partial Least Square-Structural Equation Modeling (PLS-SEM). PLS-SEM is used in this study because it has advantages over covariance-based SEM. For example, PLS-SEM does not require a large sample size (Hair et al., 2017). The data collection for this study was carried out during the pandemic when the possibility of obtaining large amounts of data was limited. In addition, PLS-SEM can provide higher statistical power than covariancebased SEM (Hair et al., 2017). There are two stages of analysis in PLS-SEM. First, testing the measurement model, namely data quality which includes reliability and construct validity. The construct validity test includes convergent and discriminant validity tests. Second, testing the structural model to test the hypothesis.

Results and Discussions
The number of returned questionnaires that could be used for hypothesis testing was 89 data. This amount has met the minimum number of sample sizes. Table 1 shows the profile of respondents based on gender, age, marital status, education, and years of work. Respondents in this study had a profile primarily male (88%), married (75%), high school graduates (64%), with working years between 1-less than three years (56%), and aged more than 35 years (55%). Data analysis is carried out in two stages: testing the measurement and structural models. Measurement model testing was conducted to identify construct validity and reliability. Figure 2 shows the results of the measurement model test. The figure shows the four research constructs, their measuring indicators, and each indicator's loading value. Table 2 shows the results of the analysis of convergent validity and reliability.

Figure 2. The Measurement Model Test Result
To obtain valid and reliable indicators, several indicators must be dropped and not used in further testing (i.e., AC1, CS2, CS4, OT3, WS6, WS7). The outer loading for all indicators has shown a value above 0.7, and the average variance extracted (AVE) of all constructs has shown a value of more than 0.5 (Hair et al., 2017). Thus, all indicators and constructs in this study have met convergent validity. Reliability has been achieved based on the composite reliability (CR) value, which is more than 0.8 (Hair et al., 2017). The discriminant validity test was carried out using the correlation techniques heterotraitmonotrait (HTMT) ratio. Table 3 shows the results of the discriminant test. All constructs have a value of less than 0.85, indicating that one construct is different from another (Hair et al., 2017). The test results regarding the relationship between variables are shown in figure 3 and table 4. Structural model testing shows that coworker support and autonomy have no significant effect on work stress (ß = 0.157, ns.; = -0.093, ns.); thus, hypotheses 1 and 2 are not supported. Coworker support and autonomy have a significant positive effect on affective occupational commitment (ß = 0.311, p <0.01; = 0.243, p <0.05). This indicates that hypotheses 3 and 4 are supported. The results of this study support hypothesis 5, since affective occupational commitment has a significant negative effect on work stress (ß = -0.534, p <0.01).

Table 3. The Structural Model Test Result
The study finds that affective occupational commitment mediates the relationship between coworker support and work stress (ß = -0.166, p <0.05). Therefore hypothesis 6 is supported. However, hypothesis 7 is not supported because affective occupational commitment does not significantly mediate the relationship between autonomy and work stress (ß = -0.130, ns.). Furthermore, coworker support and autonomy explain 20.4% of the variance in affective occupational commitment. Meanwhile, coworker support, autonomy, and affective occupational commitment explain 27.8% of the variance in work stress. The findings of this study indicate that coworker support does not significantly affect work stress. The result does not support previous studies, such as the study of Beehr et al. (2000) and Yousaf et al. (2020), who found that coworker support had a negative effect on work stress. These results show that some respondents receive the support of their coworkers and this may reduce stress, but their stress is increased for other reasons. These results are consistent with Wallace (2013) that social support may not necessarily be a restraining effect on people's responses to stress. It depends on whether the support forms a bad or good perception of the situation (Wallace, 2013). According to the findings, although some individuals judged their work situation to be good as a result of their coworkers' support, some other individuals received peer support which increased their negative assessment of their work situation.
Contrary to expectations, this study finds that autonomy does not significantly affect work stress. It does not support previous studies, such as the Lange and Kayser (2022) and Schiff and Leip (2019) studies, which found that autonomy negatively affected work stress. These results indicate that some respondents perceive the existence of autonomy as reducing their stress, but for some it will increase their stress. Individuals who have autonomy will increasingly gain control over their work (Lin & Ping, 2016). For individuals who have confidence that they are able to exercise control over their work, their stress can be reduced (Schaubroeck & Fink, 1998). However, it is possible that there are individuals who are not ready and do not have sufficient competence to exercise control over their work (Schaubroeck & Fink, 1998). As a consequence, they become even more stressed (Schaubroeck & Fink, 1998).
This study finds that coworker support has a positive effect on affective occupational commitment. The higher the coworker support, the higher the commitment to work, and vice versa. Furthermore, the relationship between coworker support and work stress is mediated by affective occupational commitment. These results indicate that work stress can be reduced insufficiently if online motorcycle taxi drivers only have supportive coworkers as their resources. These resources need to be assessed positively and generate a commitment to their work. Referring to Rousseau and Aubé (2010), individuals who feel comfortable working in a company because of the support of their coworkers, will provide better commitment to their organization. Furthermore, peer support that allows individuals to better deal with their work situations will develop their commitment to their work (Lin, 2017). Thus, the support of colleagues needs to provide comfort and peace for individuals to deal with their work situations. It will further develop their commitment and reduce their stress. These findings support social cognitive theory, which states that an individuals' assessment of their work situation can shape their attitudes and their reaction.
The results of this study also show that affective occupational commitment has a negative effect on work stress. The higher the commitment, the lower the work stress, and vice versa. These results support the opinion of Schiff and Leip (2019) that work attitudes can affect stress. Furthermore, this study finds that autonomy positively affects affective occupational commitment. This finding is consistent with Lin and Ping (2016), that autonomy allows individuals to experience pleasant work situations and creates a sense of belonging at work. The higher the autonomy, the higher the commitment to work, and vice versa. However, affective occupational commitment does not mediate the relationship between autonomy and work stress.
According to the results, the autonomy received by individuals is considered an important resource to do their work and make their work enjoyable. However, it turns out that the commitment of the individuals does not significantly bring autonomy to reduce their work stress. These results indicate that some respondents perceive autonomy as a positive thing, thereby increasing their commitment to work and subsequently reducing stress. However, this commitment was not enough for some of the other respondents in reducing their stress. There are several explanations for this finding. First, even though they have received autonomy, some individuals are actually not ready and not competent enough to carry out that autonomy (Schaubroeck & Fink, 1998). Thus, even though autonomy gives them freedom in doing their work, it does not reduce their stress due to a lack of readiness. Secondly, Indonesian society tends to have high power distance cultural values . Since individuals with high power distance cultural values want their leaders' existence and clear direction, autonomy is not suitable for those with high power distance values (Daniels & Greguras, 2014). Thus, although autonomy may lead to a commitment to work, individuals still expect guidance from their supervisors. If the guidance is reduced, it is possible that the stress still increases.

Conclusions
The current study examines the effect of coworker support, autonomy, and affective occupational commitment on work stress from online motorcycle taxi drivers in Surabaya. In contrast to existing studies, which focus on organizational commitment related to the type of work of non-standard workers, the current study considers the role of affective occupational commitment. The results of this study support the social cognitive theory, which states that individual assessments of their work environment can shape their beliefs and reactions. This study indicates that coworker support and autonomy have a positive effect on affective occupational commitment, and affective occupational commitment has a negative effect on work stress. Unlike what was hypothesized, coworker support and autonomy did not directly affect work stress. However, coworker support indirectly affects work stress by mediating affective occupational commitment. In contrast, affective occupational commitment does not significantly mediate the relationship between autonomy and work stress.
The current study also has limitations. Namely, the research model only considers coworker support, autonomy, and affective occupational commitment as antecedents of work stress. Future researchers are suggested to consider other interesting factors as independent variables, such as job satisfaction, organizational support, and employee engagement. Further studies need to also consider the role of gender in the relationship between social support and work stress. Women need social support when facing events that are considered threatening (Gardner & Fletcher, 2009). Furthermore, social support may also be expected differently in different cultural contexts (Yousaf et al., 2020). Regarding the insignificant finding of the relationship between autonomy and work stress, it is necessary to investigate the role of cultures such as power distance and online readiness of motorcycle taxi drivers in carrying out their work. In order to understand the generalization of the research model, further studies may test it in different fields of work, especially fields with non-standard employment characteristics.
This study indicates that resources such as coworker support can reduce stress through affective occupational commitment. Therefore, it is important to increase the willingness of online motorcycle taxi drivers to provide mutual support to their colleagues. In order for peer support to lead to increased commitment, it must make individuals have a positive assessment of their working conditions. Online motorcycle taxi drivers need to strengthen friendships with colleagues who provide positive views and enthusiasm. In addition, this study found that autonomy can increase commitment to work, but even if it already has a commitment, autonomy does not significantly reduce stress. In connection with this result, online motorcycle taxi drivers need to be equipped with sufficient competence to have the confidence that they are capable of doing their job. Furthermore, they need to learn various work equipment and procedures in order to have the knowledge and skills that are qualified to carry out their job duties. Companies also need to provide sufficient information about knowledge requirements and work procedures. They also need to equip online motorcycle taxi drivers with the knowledge to provide the best service for consumers, and solve problems that tend to arise in serving them. In addition, companies need to provide activities that build togetherness among online motorcycle taxi drivers and provide them with a positive work experience. Thus, they will strengthen each other and provide enthusiasm and positive evaluation of their work.