Esprit Rock

5236845239 Employee autonomy

5236845239
Employee autonomy, stress and job satisfaction among workers in Europe
Abstract
This paper examines the effect of employee autonomy on job satisfaction based on the (most recent) 6th European Working Conditions Survey, which is obtained by conducting in-depth interviews among 44000 workers originating from 35 European countries. Based on binary logistic regressions, it was concluded that overall employee autonomy increases job satisfaction, which is likely to result in increased performance, low turnover rates and satisfied customers. However, the relationship between autonomy and job satisfaction is curvilinear, as autonomy which exceeds a certain level, causes job satisfaction to decrease which may be explained by excessive stress, which is proven to negatively influence job satisfaction. Additionally, this paper provides the first evidence of the importance of individual aspects of autonomy, such as an employee’s ability to influence important decisions, being involved in improving the organisation and being allowed to choose the speed in which tasks are completed. All these aspects were found to have a significant positive impact on job satisfaction. On the other hand, it was found that employees who are allowed to choose their own methods of work, experience a decrease in job satisfaction. These results have implications on both managers as well as staff members looking to maximize job satisfaction among employees.
KEY WORDS – job satisfaction ? autonomy ? employee autonomy ? stress ? employee happiness
-28575117665
Bachelor thesis – International Business Dani Vissinga – S2951096
Supervisor – Dr. Milena Nikolova
University of Groningen
Faculty of Economics and Business
INTRODUCTION
No doubt exists about the importance of job satisfaction. The quote ‘choose a job that you like, and you will never have to work a day in your life’ is often attributed to the Chinese philosopher Confucius, who was born in 551 BC. However, no substantial support exists claiming that Confucius did indeed make this statement. The source of the quote might be ambiguous, but it proves that being satisfied with your job has been important for a long time. Happiness studies are booming, and one of its main aspects is job satisfaction. Creating a satisfied and loyal customer base, which generates an ongoing profit, may seem like the major objective of a company. However, prior research has shown that a main source of having a satisfied clientele is having happy employees (Hyken, 2017). Employee dissatisfaction on the other hand, is directly related to a high turnover rate (Gregory, 2011). Numerous sources of dissatisfaction have been found, such as lack of recognition, high levels of stress, insufficient communication and limited opportunities for growth and development (Gregory, 2011).
It is suggested to enhance employee satisfaction by giving rewards and recognition to those employees who perform exceptionally, which motivates them to continue their efforts as well as setting an example for other employees, inspiring them to provide excellent service to customers. Another way to motivate employees is by giving them a certain degree of autonomy (Gagné ; Deci, 2005), which will result in higher levels of motivation. Having autonomy entails many aspects, such as the degree of dependence on your boss or superior, the amount of influence you have on the work you do, your assigned tasks and when you do it, your relationship with your boss, and so forth. Hence, autonomy can be considered as an important factor potentially influencing overall job satisfaction.
It can be stated that job satisfaction is essential to ensure the success of any business.
Therefore, this research will provide managers with insights on the topic of job satisfaction in relation with autonomy and all the aspects it entails. Happiness studies have gained increased importance, and therefore the determinants to keep incumbents happy are relevant for management in order to increase productivity and to lower turnover. Prior research has shown that motivation on the work floor is highly important and autonomy can play a major role in this (Chang, Chen ; Chi, 2016). Thus, this research will examine the question whether autonomy has a significant impact on job satisfaction. Additionally, in order to attempt to fill the literature gap, individual aspects of autonomy will be analyzed, in order to obtain knowledge about the most relevant determinants of job satisfaction. Furthermore, many researchers have considered stress as a major source of job satisfaction, however, merely small-scale studies have tested this relationship. Therefore, in order to answer these research questions, Eurofound’s 6th European Working Conditions Survey (EWCS) will be utilized (Eurofound, 2015). It paints a clear and representative picture of working conditions, occupations, age groups and sectors among nearly 44000 workers originating from 35 European countries. Both overall autonomy, 8 independent measures of autonomy as well as stress are tested on job satisfaction to explore their relationship with job satisfaction, based on a recent and representative sample of employees within Europe. This paper reinforces the prior conclusions that employee autonomy has a significant effect on the level of job satisfaction. It was found that an employee who has autonomy is significantly more likely to be satisfied with their job compared to employees with low levels or no autonomy. Furthermore, it is found that being able to choose your speed of work, being allowed to influence important decisions and improvements on the work floor and having a say in your colleagues is positively related to job satisfaction. However, the relationship between autonomy and job satisfaction was shown to be curvilinear, indicating that at one point increasing autonomy does no longer increase job satisfaction. Nevertheless, stress, which is related to high levels of autonomy, negatively influences job satisfaction.
LITERATURE REVIEW
Job satisfaction
The main objective of Human Resource Management is to ensure that the organization is able to achieve success through people, which requires employees to deliver high-quality performance (Armstrong, 2006). In order to achieve this, worker happiness is crucial (Gregory, 2011). Despite the extensive prior research on job satisfaction, scholars have not universally agreed upon a definition. A simple, yet clear explanation states that the term ‘job satisfaction’ refers to the attitudes and feelings people have about their work (Armstrong, 2006). Hoppock (1935) says these feelings one has about their job is formed by a combination of psychological, physiological and environmental circumstances. According to Kalleberg (1977) job satisfaction is a result of a sense of recognition by the employer of the employee’s needs. Therefore, employees should seek to find those factors that increase employee happiness. Job satisfaction also includes the extent to which an employee is satisfied with the rewards obtained from the job, especially in terms of intrinsic motivation (Statt, 2004). The definition that is used most widely among researchers is provided by Locke (1976), who stated that job satisfaction is the pleasurable or positive emotional state resulting from the appraisal of one’s job or job experience.

Thus, job satisfaction can be considered as the feelings and emotions one gets from their job, which is determined by the circumstances at work, such as the received rewards and recognition.
Extensive research on the topic of job satisfaction has been conducted, and a very well-known and influential model in research on the relationship between work and health is the job demand and control (JDC) model (Karasek, 1979), widely known as the job strain model. It identifies two crucial job aspects, namely job demands and job control. Job demands are measured by the amount of strain, which are the requirements set at work, which includes aspects such as time pressure, effort and difficulty. Employees who report to experience high levels of job demands, that create work-life balance conflicts, tend to exhibit low levels of job satisfaction (Beauregard & Henry, 2009). The level of job control is determined by the decision latitude, which concerns the freedom and control the incumbent has over his or her job. This model shows that the strain itself does not lead to high psychological stress, the combination of strain and decision latitude determines the level of stress experienced. Later, a social dimension was added to the model, creating the JDCS, job- demand-control-support model (Johnson & Hall, 1988). According to this theory, jobs which are characterized by high demands, low control and low support are considered to be the most harmful work situations. Social support moderates the negative impact of potential high strain. The credibility of the model was retested by van der Doef and Maes (1999), who reviewed and analyzed 63 samples published in the timeframe of 1979-1997 that included substantial amounts of references to the job demand and control model. Their hypothesis was confirmed: a high level of strain combined with lo decision authority leads to stress and dissatisfaction.
Motivation
Motivation has been a persistent topic within the field of psychology and is becoming increasingly important in the field of business due to the fact that motivation simply produces in a sense of revenue. Motivation means being moved to perform an action, whereas being unmotivated is characterized by not having an inspiration or incentive to act (Ryan & Deci, 2000). Generally, individuals can be intrinsically motivated or extrinsically motivated. The most basic distinction between intrinsic motivation and extrinsic motivation is that the former refers to doing something because it is interesting or enjoyable and the latter refers to doing something because it will lead to a separable outcome (Ryan & Deci, 2000). Thus, intrinsic motivation is guided by the fact that doing it is personally rewarding and extrinsic motivation is guided by obtaining a reward. Often, people tend to be motivated by rewards, both monetary as well as non-financial rewards, such as status or appreciation, which are called extrinsic rewards. However, one can be intrinsically motivated as well, by e.g. acting out of interest or curiosity. Those intrinsically motivated acts are not specifically extrinsically rewarded, but they can generate pleasurable feelings, such as satisfaction and creativity. High levels of motivation have proven to increase production, which is of utmost importance for managers (Ryan & Deci, 2000). These findings led to the establishment of the self-determination theory, in which the combination between extrinsic forces and intrinsic motives are embedded (Ryan & Deci, 1985).
According to the self-determination theory, people’s growth tendencies and psychological needs are the fundamental needs for motivation that foster positive processes. Thus, motivation is another crucial aspect within the field of job satisfaction. Ryan ; Deci (1985) used an empirical process to identify three needs that facilitate the optimal functioning of human beings: competence (Harter, 1978; White, 1963), relatedness (Baumeister ; Leary, 1995; Reis, 1994) and autonomy (deCharms, 1968; Deci, 1975). Competence is the feeling of being capable while seeking to control the outcome in order to experience mastery (White, 1959). Relatedness stands for the universal need to feel affiliated and connected to others (Baumeister ; Leary, 1995). In the research of the self-determination theory, autonomy was defined as the universal urge to be in charge of one’s own choices and to have a feeling of being in control.
If these factors are being satisfied, not only optimal functioning of human beings is enhanced, it also fosters social development and personal well-being, which in turn may influence job satisfaction. Thus, if the social context in which individuals are embedded are responsive to these three psychological needs, an environment where development and personal well-being is created.
The importance properly balancing competence, embeddedness and autonomy was reinforced by Danner & Lonky (1981) who did an experiment with 90 children, between the age of 4 and 10. When allowing the children to choose among learning centers that differed in level and required understanding, the vast majority picked those learning centers that were just beyond their initial ability levels. The second experiment, the same children received rewards, praise or no rewards for either working at, above or below their predicted level of understanding. It was proven that extrinsic rewards had little to no effect on intrinsic motivation, and children who were initially highly motivated, experienced a decrease in their level of motivation when receiving extrinsic rewards. This study proved that even at a young age, challenging activities are generally intrinsically motivated, as it provides one with intrinsic rewards, namely the feeling of being competent and having autonomy.
Autonomy
As proven in Ryan & Deci’s (1985) self-determination theory, having a certain degree of autonomy is crucial to fully exploit the opportunity for personal development and to create optimal well-being. It was even found that external factors, such as deadlines, decrease intrinsic motivation due to the fact that it restricts one’s autonomy (Amabile, DeJong & Lepper, 1976).
According to Verme (2009), who utilized the results of the World Values Survey, in which 267870 observations from 84 countries were analyzed, the degree of control and freedom is the most significant predictor of satisfaction. Therefore, in order for an employee to feel like they are fully utilizing their capabilities and to stimulate motivation on the work floor, a certain degree of autonomy is expected to be necessary. Ugboro & Obeng (2000) reinforced the importance of autonomy, by empirically proving that employee empowerment and autonomy is positively related to job satisfaction, which results in customer satisfaction. This leads to the first hypothesis:
Hypothesis 1: High overall employee autonomy increases overall job satisfaction
However, it was found that autonomy can also be overdone. High levels of autonomy, which often means having an excessive amount of control, showcase the downside of autonomy. Thompson and Prottas (2005), used the National Study of the Changing Workforce survey from 2002, to examine the relationship between job satisfaction, personal well-being and job autonomy. The findings provide strong evidence for the importance of autonomy on the work floor. High levels of job autonomy were highly correlated with both job satisfaction as well as overall life satisfaction. However, they found that high job autonomy is not only associated with satisfaction, but also with stress. When the degree of autonomy substantially increased, satisfaction can be overtaken by stress. If the desired level of autonomy is exceeded, employees might be afraid to be fully associated with a task, resulting in stress and dissatisfaction. Also, if an employee is not properly equipped, both in skills as well as knowledge, autonomy can lead to tension and underperformance (Simmering, N.D.).
Thus, this leads to the second hypothesis:
Hypothesis 2: Autonomy has a curvilinear relationship with job satisfaction: At a certain point increasing autonomy will not further increase job satisfaction
The relationship between stress and autonomy has been shown in previous research, however, the question whether stress significantly impacts overall job satisfaction remains unanswered. Brewer (2003) attempted to answer this question by analyzing a random sample of 133 technical and industrial teachers, where significant relationships between job satisfaction and stress related factor were revealed. Another research among 21307 public school teachers in South-Africa showed that stress is indeed negatively related to job satisfaction, and may even cause severe illnesses such as hypertension, heart disease and stomach ulcers (Peltzer, Shisana, Zuma, van Wyk & Zungu-Dirrwayi, 2009). The negative relationship was once again reinforced by Bemana, Moradi, Ghasemi, Taghavi & Ghayoor (2013) who examined stress and job satisfaction among public municipality employees from the Shiraz area in Iran. Clearly, research on the relationship between stress and job satisfaction has been done before, however, generalizability is one major limitations of all these previous studies. Therefore, this research will examine the relationship between stress and job satisfaction for a representative sample among workers in Europe, leading to the third hypothesis:
Hypothesis 3: Experiencing stress regularly decreases job satisfaction
According to Smith (N.D.), autonomy entails many aspects in which an employee can control their work situation. It can involve having a say in the projects that are pursued, the clients, colleagues and the methods in which the work is done. It also includes having responsibility about important decisions made within the organization (Baldoni, 2013). Autonomy means self-governing (Collier, 2002), having decision latitude on the work floor and being able to influence and make decisions regarding work-related tasks (Gagné & Deci, 2005). Many aspects of autonomy have been highlighted in previous studies, but the individual effect on job satisfaction of each aspect have not been examined. Therefore, this paper attempts to fill the literature gap by examining individual aspects of autonomy, and their respective effect on job satisfaction.
The three aforementioned concepts are strongly linked. Motivation is needed in order for an incumbent to experience a feeling of satisfaction with their job. However, in order to be motivated, possessing a degree of autonomy is crucial. Therefore, there is a triangular relationship between autonomy, motivation and job satisfaction.

Within the European Working Conditions Survey, different indicators of autonomy were highlighted (Eurofound, 2015). Because overall autonomy is expected to be positively related to job satisfaction, it is expected that each individual aspect of autonomy is positively related with job satisfaction, resulting in the following hypotheses:
Hypothesis 4: If an employee has influence on the order in which their tasks are done, they are more satisfied with their jobHypothesis 5: If an employee has influence over the methods they use to complete their tasks, they are more satisfied with their job
Hypothesis 6: If an employee is in charge of the speed in which they do their work, they are more satisfied with their job
Hypothesis 7: If an employee has decision making authority, they are more satisfied with their job Hypothesis 8: If an employee is involved in improving the organization, they are more satisfied with their job
Hypothesis 9: Being allowed to apply your own ideas at work increases job satisfaction
Hypothesis 10: Being able to determine your own colleagues, increases job satisfactionHypothesis 11: Taking a break when desired, without having to ask for permission, increases job satisfaction
METHODS
Data and sample
The purpose of this paper is to analyze the effect of numerous aspects of employee autonomy on job satisfaction. Prior research on the topic of job satisfaction and its determinants is often outdated, which violates its representativity in today’s society. Therefore, in order for this research to be reliable and to showcase the importance of autonomy and its underlying relationship with job satisfaction in today’s society, Eurofound’s 6th European Working Conditions Survey (EWCS), which was conducted in 2015, is utilized. The EWCS paints a picture of Europe’s working conditions across countries, occupations, sectors and age groups by conducting interviews among 43850 workers across all the European Union countries, as well as Albania, Montenegro, Norway, Serbia, Switzerland, Turkey and the former Yugoslav Republic of Macedonia (Eurofound, 2015). From the 35 countries included in the research, each country has a target sample size of 1000 respondents. However, in order to evenly reflect the size of the country and its workforce, the sample size of Polish, Spanish, Italian, French, German, British and Turkish respondents was increased according to the employed population. The EWCS is conducted every 5 years, in order to assess the working conditions of men and women across Europe by conducting in-depth interviews between February and September of the corresponding year. The sample includes all residents of the aforementioned countries above the age of 15, and above the age of 16 in Bulgaria, Norway, Spain and the United Kingdom. Only those people who worked for at least an hour in the week preceding the interview were considered to be employed, thus included in the sample.
Measures
Dependent variable
The dependent variable in this research is job satisfaction. The overall job satisfaction was examined in the EWCS, measured on a 4-point Likert scale, ranging from, are you: very satisfied, satisfied, not very satisfied or not at all satisfied. A new variable called ‘jobsatisfactiondummy’ was created and recoded into a dichotomous scale, where 0 represents not very satisfied and not at all satisfied, and 1 represents very satisfied and satisfied. The descriptive statistics shown in table 1 below indicate a mean of 0.841, from which can be concluded that 84,1% of the 40752 respondents are either satisfied or very satisfied with their current working conditions.
Independent variables
The main independent variable in this research is autonomy. Based on all the aspects that employee autonomy entails, it is extremely difficult to measure autonomy by one variable, so in order to triangulate data on autonomy, several questions from the EWCS were examined, each representing another aspect of autonomy. First, the extent to which an employee is allowed to choose or change parts of their job is examined. This includes the autonomy an employee has concerning the order of tasks, the methods of work and the speed of work. Second, it is explored whether the employees in the sample are able to influence important decisions on the work floor, whether they are involved in improving the work processes, whether they are allowed to apply their own ideas into their work, whether a break can be taken when desired and whether they have a say in the selection of colleagues they work with. All these aspects are tested on job satisfaction, to examine the individual effect. These variables are dummy variables, with value 0 indicating no autonomy and value 1 indicates autonomy. From the descriptive statistics shown in table 1 below, based on the means, it can be concluded that only being able to choose your colleagues (‘choicecol’) can be done by less than 50% of the respondents, namely 44%. All the other aspects of autonomy are possessed by more than half of the workers in the sample. The exact numbers can be found in table 1 below.

Afterwards, a new variable to measure overall autonomy was created by summing all the individual variables and recoding them in one categorical variable, with a range from 0 to 8, called ‘sumautonomy’. The mean of 5,328 indicates that on average, workers possess 5 aspects of autonomy within their job. Therefore, above average autonomy includes those workers who have at least 6 of the indicated aspects of autonomy.
In order to test whether job satisfaction and autonomy have a curvilinear relationship (Hypothesis 2), a squared version of the sum variable called ‘sumautonomy’ was created, resulting in a new variable called ‘sumautonomysquared’. However, in order to reduce the correlation between the sum variable and its quadratic version and to strengthen the reliability of the analysis, the sum variable is centered by subtracting the mean from each value. This variable is called ‘sumautonomymean’ and another squared version called ‘sumautonomymeansquared’ is created.
Additionally, stress will serve as the second independent variable in order to test its relationship with job satisfaction. The EWCS respondents were asked whether they experienced stress during their work, where the answer options were measured on a 4-point Likert scale, ranging from always, most of the time, sometimes, rarely to never. In order to test the hypothesis whether experiencing regular stress influences job satisfaction, a variable called ‘regularstress’ was created, where always and most of the time were coded as 1, and the other answer options were coded as 0.
Table 1: Descriptive statistics
Obs. Minimum Maximum Mean Std. Deviation Jobsatisfactiondummy 40752 0,00 1,00 0,841 0,366 Ordertasks40691 0,00 1,00 0,670 0,470 Methodtasks40700 0,00 1,00 0,690 0,463 Speedtasks40606 0,00 1,00 0,720 0,447 Involved 38294 0,00 1,00 0,700 0,459 Choicecol 36417 0,00 1,00 0,440 0,497 Takebreak40561 0,00 1,00 0,680 0,467 Ownideasyn40285 0,00 1,00 0,790 0,406 Influence 39918 0,00 1,00 0,730 0,443 Sumautonomy 34723 0,00 8,00 5,328 2,354 Regularstress40559 0,00 1,00 0,268 0,443 All the questions were measured on a dichotomous scale, where 0 indicates no (no sign of autonomy) and 1 indicates yes (there is autonomy). To test for multicollinearity the correlation coefficients between variables were examined. It was shown that no correlation above the commonly used threshold of 0,7 was found, so there is no indication of multicollinearity. The results can be found in table 2 below.

Table 2: Correlation matrix
(1) (2) (3) (4) (5) (6) (7) (8)
(1)Ordertasks1 (2)Methodtasks0,576 1 (3)Speedtasks0,498 0,521 1 (4)Involved 0,271 0,269 0,208 1 (5)Choicecol 0,218 0,211 0,168 0,357 1 (6)Takebreak0,297 0,228 0,246 0,214 0,258 1 (7)Ownideas0,339 0,370 0,283 0,438 0,304 0,258 1 (8)Influence 0,319 0,323 0,258 0,467 0,369 0,268 0,487 1
Control variables
First of all, age is used as a control variable. According to the aforementioned self-determination theory (Ryan and Deci, 1985; 2000), autonomy is a psychological need for all human beings and over time people gain the ability to make self-appropriate choices. Therefore, it is likely that the desire for autonomy comes at a later age, which was empirically tested and confirmed (Sheldona, Houser-Markoa & Kasser, 2006). The corresponding study compared a sample of middle-aged people to a sample of college student, and confirmed that people develop a greater degree of autonomy as they age. Therefore, age may have a substantial effect on the perception of autonomy, hence it will be used as a control variable. In the EWCS sample, respondents are between the age of 15 and 87, with an average age of 43.
The second control variable is gender. Adler (1993) empirically found that on average, men have more job autonomy than women. This was partly reinforced by Ecker (2005) who analyzed job autonomy and its relation to gender in Sweden and the US. It was found that overall Swedish men experience more job autonomy than women from the same country. Thus, job autonomy may significantly differ among genders. Gender is measured on a dichotomous scale where men are coded as 1 and women are coded as 0. The mean of 0,51 indicates that the sample contains slightly more men than women.

The degree of perceived autonomy is likely to vary among occupations, so the final control variable is occupation. In the European Working Conditions Survey the respondents are divided among 10 broad occupational categories coded from a range from 0 to 9, each representing a category. As there is no meaning behind the assigned number of each occupation, they were recoded to 10 separate dummy variables. Each dummy represents an occupation, where 1 means the corresponding respondent falls within that occupational category, and 0 means they fall within another category. As can be seen in table 3 below, service and sales workers account for the largest share of occupations within the sample, with 21,5%, followed by professionals (18,4%), craft and related trades workers (12%) and technicians and associate professionals (11,6%) consecutively.
Table 3: Descriptive statistics and frequencies of control variables
Obs. Minimum Maximum Mean Percentage Cumulative percentage Age 40757 15 87 43,01 – – Male 40885 0 1 0,51 – – Armedforces154 0,00 1,00 – 0,4 0,4 Managers 2665 0,00 1,00 – 6,5 6,9 Professionals 7511 0,00 1,00 – 18,4 25,3 Technicians 4720 0,00 1,00 – 11,6 36,9 Clerks 3640 0,00 1,00 – 8,9 45,8 Serviceandsales8786 0,00 1,00 – 21,5 67,3 Skilledagricultural1636 0,00 1,00 – 4,0 71,3 Crafts 4898 0,00 1,00 – 12,0 83,3 Operators 2877 0,00 1,00 – 7,1 90,4 Elementaryoccupation3921 0,00 1,00 – 9,6 100 In order to analyze the effect of individuals aspects of autonomy on overall job satisfaction, binary logistic regressions will be executed. This method has been chosen based on the dependent variable, job satisfaction, which is a dichotomous variable, taking on the value of either 1 or 0.
A detailed description of the variables used in this paper is provided in appendix 1.

ANALYSIS AND RESULTS
The results highlight the importance of autonomy. All the odds ratios and model characteristics can be seen in table 4 below, the results of the control variable ‘occupation’ can be found in appendix 3.
When performing a binary logistic regression on job satisfaction and the combined measure of autonomy, there is an odds ratio of 2,396 which is statistically significant at the p<0,01 level (Model 1). This indicates that if an employee has at least 6 aspects of autonomy, so there is an above average level of autonomy, that person is 2,396 times more likely to be satisfied with his or her job, when controlling for age, gender and occupation. Therefore, Hypothesis 1 can be confirmed.
When testing Hypothesis 2, a regression analysis of the sum variable of autonomy and its squared version shows coefficients of 0,049 and -0,002 respectively, both significant at the p<0,01 level. When regressing the quadratic versions, ‘sumautonomymean’ has a coefficient of 0,029, significant at the p<0,01 level, and ‘sumautonomymeansquared’ has a coefficient of
-0,001, also significant at the p<0,01 level. All the output, including control variables can be found in appendix 2. This indicates that there is diminishing marginal utility when increasing autonomy. This means that at first, increasing autonomy positively affects job satisfaction, however, after a certain point increasing autonomy does not yield extra job satisfaction. This means that hypothesis 2 is confirmed.

Through the use of logistic regression, the association between stress has on job satisfaction is tested (Model 4). When regressing stress and job satisfaction, there is an odds ratio of 0,321, significant at the p<0,01 level. This implies that if an incumbent regularly experiences stress, they are 67,9% less likely to be satisfied with their job. This means that the third hypothesis can indeed be confirmed.

Furthermore, it was also shown that most individual aspects of autonomy (Model 2), do in fact significantly influence job satisfaction when controlling for age, gender and occupation (Model 3). When testing the effect of having influence on the order of tasks, the odds ratio was shown to be 1,106, significant at the p<0,05 level. This indicates that if an employee has influence over the order in which their tasks are performed, they are 1,106 times more likely to be satisfied with their job. However, when controlling for age, gender and occupation, the odds ratio decreases to 1,069 and is no longer significant, even on the p<0,10 level. Therefore, it can be concluded that having influence of the order of tasks has no significant relationship with job satisfaction. Thus, hypothesis 4 is not confirmed.

When testing the fifth hypothesis, whether being able to influence the methods of work increases job satisfaction, when all other variables remained constant, there was an odds ratio of 0,896. It is significant at the p<0,01 level, which means that being able to influence your methods of work, decreases the chance of being job satisfied. Therefore, the fifth hypothesis is not confirmed. Being in charge of your speed of work is, however, a significant predictor of job satisfaction. With an odds ratio of 1,170, significant at p<0,01, when controlling for age, gender and occupation, it can be said that someone who is able to control the speed they work in, is 1,170 times more likely to be satisfied with their job. Therefore, hypothesis 6 is confirmed.

Additionally, having influence over decisions that are important at work, is expected to positively influence job satisfaction. When performing a logistic regression analysis while keeping the controls constant, influence has an odds ratio of 1,521, again significant at p<0,01 level. This indicates that someone who has influence over important decisions at work, is 1,521 times more likely to be satisfied with the job, thus confirming hypothesis 7.

Next it was tested whether an employee being involved in improving the organisation increases job satisfaction while controlling for age, gender and occupation. The odds ratio of 1,484 indicates that an employee who is being involved in improving the organisation, is 1,484 times more likely to be satisfied compared to someone who is not involved, . The odds ratio is significant at p<0,01 level, indicating that hypothesis 8 can be confirmed.
Being able to apply your own ideas at work, also has a significant positive effect on job satisfaction. The odds ratio of 1,418, significant at p<0,01 level shows that also the eighth hypothesis can be confirmed; someone who is allowed to apply their own ideas at work, is 1,418 times more likely to be satisfied with his or her job, if age, gender and occupation are controlled for. Thus, hypothesis 9 is confirmed.

Not only being in charge of how and what you do shows to have a significant impact on job satisfaction, also being in charge of who you work with is of great importance according to the analysis. By testing the effect of being able to determine your work colleagues on job satisfaction, it became apparent that the odds ratio, which is significant at p<0,01 level, is 1,219, when the control variables are included. This indicates that someone who is able to determine their work colleagues, is 1,219 more likely to be satisfied with their job, compared to someone who has no influence over their colleagues. This means that hypothesis 10 can be confirmed.
Also the final hypothesis, hypothesis 11, can be confirmed. When being controlled for age, gender and occupation, the odds ratio is 1,409, significant at p<0,01 level. Therefore it can be concluded that someone who is able to take a break when desired is 1,409 times more likely to be satisfied with their job than someone who should ask for permission before taking a break.
When performing a logistic regression, there is no value properly indicating the portion of explained variance by the model (Linssen & Sieben, 2009). Therefore, often a ‘Pseudo R-Square’ is used, such as the Nagelkerke R-Square. When testing for all the individual aspects of autonomy, the model has a Nagelkerke R-Square of 0,084. This indicates that when determining job satisfaction, 8,4% of the variance is caused by the model. Thus, the 8 utilized aspects of autonomy jointly cause 8,4% of the change in the level of job satisfaction. When including the control variables, age, gender and occupation, the Nagelkerke R-Square slightly increases to 0,096, indicating a 9,6% explained variance by this model.

Table 4: Regression results on job satisfaction (jobsatisfactiondummy), odds ratios
Variables Model 1 Model 2
Model 3
Model 4 Highautonomy2,396*** (0,031) – – – – –
Ordertasks- – 1,106** (0,041) 1,069 (0,041) – –
Methodtasks- – 0,905** (0,042) 0,896*** (0,043) – –
Speedtasks- – 1,166*** (0,039) 1,170*** (0,040) – –
Involved – – 1,578*** (0,037) 1,484*** (0,038) – –
Choicecol – – 1,159*** (0,037) 1,219*** (0,037) – –
Takebreak- – 1,392*** (0,033) 1,409*** (0,034) – –
Ownideasyn- – 1,446*** (0,040) 1,418*** (0,041) – –
Influence – – 1,542*** (0,039) 1,521*** (0,040) – –
Regularstress- – – – – – 0,321*** (0,029)
Age 1,001 (0,001) – – 1 (0,001) 1,001 (0,001)
Male 0,996 (0,033) – – 0,949 (0,034) 1,060* (0,031)
Constant 2,413*** 1,677*** 1,422*** 3,924*** Nagelkerke R-Square 0,066 0,084 0,096 0,097 Observations 34458 34637 34458 40229 Chi-Square 1332,22*** 1719,73*** 1958,62*** 2357,48*** Degrees of freedom 12 8 19 12 Standard errors in parentheses reported next to odds ratios *** p<0.01, ** p<0.05, * p<0.1
DISCUSSION AND LIMITATIONS
In this paper, different aspects of autonomy were both individually as well as jointly tested on job satisfaction, in order to examine the effect of (aspects of) autonomy on job satisfaction. Having a certain degree of autonomy was expected to be crucial for job satisfaction, to enhance mostly intrinsic motivation, which in turn leads to greater productivity and lower employee turnover.
It was indeed proven that if an employee has a high degree of autonomy, they are more likely to be satisfied with their job, so Ryan & Deci’s (1985) well-known self-determination theory, which states that autonomy is a necessity to optimize happiness, is confirmed.

However, it was found that at one point, increasing the degree of one’s autonomy does no longer increase job satisfaction. This reinforces Karasek’s (1979) job demand and control model, as job demands exceed the optimal level, and thus results in dissatisfaction. One major factor associated with creating job dissatisfaction is stress, which is proven in this paper. The relationship between autonomy and stress has been examined before, where Thompson and Prottas (2005) found that excessive job autonomy is linked to stress, resulting in decreased job satisfaction. Peltzer et al (2009) and Bemana et al (2013) found a negative relationship between stress and job satisfaction, however, both studies’ main limitation is their representativity. Therefore, this relationship was re-examined in this paper through the use of the EWCS which is representative for workers among Europe. The prior finding were confirmed. Stress is indeed significantly correlated with a decrease in job satisfaction.
Furthermore, 8 individual effects of different aspects of autonomy were tested. Based on this analysis, it was shown that being able to choose your methods of work negatively influences job satisfaction. Determining your own order of tasks was shown to have no significant effect on job satisfaction. Being able to choose your speed of work on the other hand, is positively related to job satisfaction. Also being able to influence important decisions, being involved in improving the organisation, being able to take a break when desired, having the opportunity to apply your own ideas and having a say in the colleagues one works with, are positively related to job satisfaction. This paper shows that being involved in improving the work organization and work processes and having influence over important decisions at work can be considered as those aspects that have the largest positive impact on job satisfaction. This implies that both involving employees in the improvement of the organization as well as giving them influence in the decision making processes are the aspects of autonomy that have the largest impact on increasing job satisfaction, which should recognized by those managers seeking to maximize job satisfaction.
The analysis of the different aspects of autonomy has proven that each individual aspect might only have a minor influence on job satisfaction. However, if many aspects are combined, the overall effect of autonomy accounts for approximately 10% of the change in job satisfaction. Therefore, when examining or trying to increase job satisfaction, autonomy should be considered of to be of utmost importance.
However, the limitations of the study have to be taken into consideration. First of all, autonomy is a latent variable. This indicates the fact that it cannot easily be measured, as it entails many aspects. Based on existing literature and the European Working Conditions Survey, 8 crucial aspects were selected and included in the analysis. However, other aspects which were not in the EWCS may be an aspect of autonomy. Therefore, future research on the subject of autonomy may offer clarity about potential other aspects that autonomy entails.

Secondly, the EWCS’s results can be considered as a detailed and representative dataset. However, the data is solely based on workers in Europe, which is a relatively small fragment of the 196 countries worldwide. Thus, results may vary among other continents not included in the analysis. Of course, Europe cannot be considered as a homogenous continent, however, it is likely that values differ across the world, which jeopardizes the generalizability of these results for the world’s entire population. Therefore, opportunity for future research is offered, to extent this research among countries all over the globe.
Last, the triangular relationship between stress, autonomy and job satisfaction can be further explored in the future. This paper shows significant relationships between job satisfaction and stress, and job satisfaction and autonomy. Future research may further explore the causality between stress and autonomy.

CONCLUSION
Overall, this paper reinforces prior conclusions that autonomy significantly impacts job satisfaction. Additionally, this paper shows that not only autonomy as a whole, but also its separate aspects are correlated with job satisfaction. If an employee has autonomy on several aspects, an increased level of overall job satisfaction is expected. As described in this paper and proved by the analysis, autonomy is of great importance and therefore this paper can serve as a base on how to increase job satisfaction among workers. Individuals aspects of autonomy can have a significant impact on employees, so by implementing relatively small policy changes, happiness is likely to increase substantially.
REFERENCES
Adler, M.A. 1993. Gender Differences in Job Autonomy: The Consequences of Occupational Segregation and Authority Position. The Sociological Quarterly, 34(3), 449-465.
Amabile, T. M., DeJong, W., Lepper, M. 1976. Effects of externally imposed deadlines on intrinsic motivation. Journal of Personality and Social Psychology, 34, 92–98.
Armstrong, M. 2006. A handbook of human resource management practice (10th ed.). London: Kogan Page Publishing.

Baldoni, J. 2013. Employee Engagement Does More than Boost Productivity. Harvard Business Review. https://hbr.org/2013/07/employee-engagement-does-more, April 3rd, 2018.
Baumeister, R., & Leary, M. R. 1995. The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497-529.

Beauregard, T. A., & Henry, L. C. 2009. Making the link between work-life balance practices and organizational performance. Human Resource Management Review, 19(1), 9-22.

Bemana, S., Moradi, H., Ghasemi, M., Taghavi, S.M. & Ghayoor, A.H. 2013. The Relationship among Job Stress and Job Satisfaction in Municipality Personnel in Iran. World Applied Sciences Journal, 22(2), 233-238.

Brewer, E. 2003. The relationship between job stress and job satisfaction among industrial and technical teacher educators. Journal of Career and Technical Education, 20(1).

Chang, Y. K., Chen, S., Tu, K.W. & Chi, L.K. 2016. Effect of Autonomy Support on Self-Determined Motivation is Elementary Physical Education. Journal of Sports Science and Medicine, 15, 460-466.

Collier, J. 2002. What is autonomy? Philosophical Explorations. https://pdfs.semanticscholar.org/15ed/ff35e5fb5cde6960e1d51c85a815355d6648.pdf, April 7th, 2018.
deCharms, R. 1968. Personal causation. New York: Academic Press.

Danner, F. W., & Lonky, E. 1981. A cognitive-developmental approach to the effects of rewards on intrinsic motivation. Child Development, 52, 1043-1052.

Deci, E. L. 1975. Intrinsic motivation. New York: Plenum.
Ecker, K. 2005. Gender differences in job autonomy in Sweden and the United States. https://www.diva-portal.org/smash/get/diva2:854398/FULLTEXT01.pdf, May 3rd, 2018.

Eurofound. 2015. 6th European Working Conditions Survey. https://www.eurofound.europa.eu/surveys/european-working-conditions-surveys/sixth-european-working-conditions-survey-2015, March 18th, 2018.
Granovetter, M. S. 1965. Getting a job: A study of contracts and careers. Chicago: University of Chicago Press.

Gagné, M., & Deci, E.L. 2005. Self-determination theory and work motivation. Journal of Organizational Behavior, 26, 331–362.

Gregory, K. 2011. The Importance of Employee Satisfaction. Neumann Business Review. http://doi.org/10.18411/a-2017-023Harter, S. 1978. Effectance motivation reconsidered: Toward a developmental model. Human Development, 1, 661-669.

Hoppock, R. 1935. Job satisfaction. New york: Harper and Brothers.

Hyken, S. 2017. How Happy Employees Make Happy Customers. Forbes. Forbes Magazine. https://www.forbes.com/sites/shephyken/2017/05/27/how-happy-employees-make-happy-customers/#4c97a4ed5c35, April 4th, 2018.

Johnson, J. V., & Hall, E. M. 1988. Job strain, workplace social support, and cardiovascular disease: a cross-sectional study of a random sample of the Swedish working population. American Journal of Public Health, 78, 1336-1342.

Kalleberg, A. L. 1977. Work values and job rewards: A theory of job satisfaction. American Sociological Review, 42, 124-143.

Karasek, R.A. 1979. Job demands, job decision latitude and mental strain: implications for job redesign. Administrative Science Quarterly, 24, 285-308.

Linssen, L. & Sieben, I. 2009. Logistische regressie analyse: Een handleiding. www.ru. nl/publish/pages/525898/logistischeregressie.pdf, April 8th, 2018.
Locke, E.A. 1976. The nature and causes of job satisfaction (In M.D. Dunnette ed.). Handbook of industrial and organizational psychology: 1297-1349. Chicago: Rand McNally.
Peltzer, K., Shisana, O., Zuma, K., van Wyk, B., & Zungu-Dirrwayi, N. 2009. Job stress, job satisfaction and stress-related illnesses among South African educators. Stress and Health, 25, 247–257.

Reis, H. T. 1994. Domains of experience: Investigating relationship processes from three perspectives (In R. Erber & R. Gilmour ed.) Theoretical frameworks for personal relationships: 87-110. Hillsdale, NJ: Erlbaum.
Ryan, R. M. & Deci, E. L. 1985. Intrinsic motivation and self-determination in human behavior. New York: Plenum.

Ryan, R.M. & Deci, E.L. 2000. Intrinsic and Extrinsic Motivations: Classic Definitions andNew Directions. Contemporary Educational Psychology, 25, 54–67.

Ryan, R.M. & Deci, E.L. 2000. Self-Determination Theory and the Facilitation of Intrinsic Motivation,Social Development and Well-Being. American psychologist, 55(1), 68-78.

Sheldona, K.M., Houser-Markoa, L., Kasser, T. 2006. Does autonomy increase with age? Comparing
the goal motivations of college students and their parents. Journal of Research in Personality, 40, 168-178.

Simmiering, M.J. Autonomy. Reference for business, http://www.referenceforbusiness.com/management/A-Bud/Autonomy.html, April 14th, 2018
Smith, E. n.d. What Is Employee Autonomy? Chron.com. http://smallbusiness.chron.com/employee-autonomy-20930.html, March 26, 2018
Statt, D. 2004. The routledge dictionary of business management (3rd ed.). Detroit: Routledge Publishing.

Thompson, C.A. & Prottas, D.J. 2005. Relationships Among Organizational Family Support, JobAutonomy, Perceived Control, and Employee Well-Being. Journal of Occupational Health Psychology, 10(4), 100-118.
Ugboro, I.O. & Obeng, K. 2000. Top management leadership, employee empowerment, job satisfaction and customer satisfaction in TQM organizations: An empirical study. Journal of quality management, 5(2), 247-272.

Van der Doef, M. & Maes, S. 1999. The Job Demand-Control(-Support) Model and psychological well-being : a review of 20 years of empirical research. Work and stress, 13(2), 87-114.

Verme, P. 2009. Happiness, Freedom and Control. Centre for Research on the Public Sector, 141, 1-25.

White, R. W. 1959. Motivation reconsidered: The concept of competence. Psychological Review, 66, 297–333.
White, R. W. 1963. Ego and reality in psychoanalytic theory. New York: International Universities Press.
APPENDIX
Appendix 1: Explanation of variables
Variable name Corresponding EWCS question Explanation and measurement
Jobsatisfactiondummy On the whole, are you very satisfied, satisfied, not very satisfied or not at all satisfied with working conditions in your main paid job? The 4 answers possibilities were recoded into a dummy where ‘not very satisfied’ and ‘not at all satisfied’ are coded as 0 and ‘very satisfied’ and ‘satisfied’ are coded as 1.
OrdertasksAre you able to choose or change your order of tasks? Measured on a dichotomous scale where 0 is no and 1 is yes.
MethodtasksAre you able to choose or change your methods of work? SpeedtasksAre you able to choose or change your speed or rate of work? Involved You are involved in improving
the work organisation or work
processes of your department or organisation Choicecol You have a say in the choice of
your work colleagues TakebreakYou can take a break when you
Wish OwnideasynYou are able to apply your own
ideas in your work Influence You can influence decisions that
are important for your work Sumautonomy – A sum variable of all the separate aspects of autonomy. The higher the number, the higher the degree of autonomy an employee has.

Highautonomy- When an employee says to have at least 6 out of 8 aspects of autonomy, they have high autonomy on the workfloor. Thus, lower than 6 aspects is coded as 0 and 6 or higher aspects is coded as 1.

ArmedforcesWhat is the title of your main paid job? Armed forces occupations
Managers Managers
Professionals Professionals
Technicians Technicians and associate professionals
Clerks Clerical support workers
ServiceandsalesServices and sales workers
SkilledagriculturalSkilled agricultural, forestry and fishery workers
Crafts Craft and related trades workers
Operators Plant and machine operators and assemblers
ElementaryoccupationElementary occupations
Male What is your gender? Male is coded as 1 and female is coded as 0
Age Starting with yourself, how old are you? Measured on a scale, ranging from 15 to 87
Appendix 2: Regression results on job satisfaction (jobsatisfactiondummy), coefficients
Variables Model 1 Model 2 Sumautonomy 0,045*** (0,003) – – Sumautonomysquared0,001*** (0,000) – – Sumautonomymean- – 0,029*** (0,001) Sumautonomymeansquared- – -0,001*** (0,000) Male -0,004 (0,004) -0,004 (0,004) Age 0,000 (0,000) 0,000 (0,000) Armedforces0,092*** (0,030) 0,092*** (0,030) Managers 0,067*** (0,010) 0,067*** (0,009) Professionals 0,072*** (0,008) 0,072*** (0,007) Technicians 0,066*** (0,008) 0,066*** (0,008) Clerks 0,076*** (0,009) 0,076*** (0,008) Serviceandsales0,047*** (0,008) 0,047*** (0,007) Skilledagri-0,088*** (0,013) -0,088*** (0,013) Crafts 0,034*** (0,009) 0,034*** (0,008) Operators 0,039*** (0,010) 0,039*** (0,009) Constant 0,617 0,844 R-Square 0,053 0,053 Degrees of freedom 12 12

Standard errors in parentheses reported next to coefficients
*** p<0.01, ** p<0.05, * p<0.1
Appendix 3: Regression results on job satisfaction (jobsatisfactiondummy), occupation control variables
Variables Model 1 Model 3 Model 4 Armedforces1,992*** (0,265) 1,787**(0,269) 2,550***(0,259) Managers 2,101*** 1,608*** 3,745*** (0,089) (0,091) (0,079) Professionals 1,953*** 1,613*** 3,190*** (0,060) (0,062) (0,055) Technicians 1,780*** 1,472*** 2,689*** (0,065) (0,066) (0,060) Clerks 1,842*** 1,630*** 2,383*** (0,068) (0,069) (0,064) Serviceandsales1,417*** 1,270*** 1,689*** (0,054) (0,055) (0,049) Skilledagri0,611*** 0,504*** 0,723*** (0,088) (0,090) (0,069) Crafts 1,264*** 1,148** 1,479*** (0,061) (0,063) (0,056) Operators 1,189*** 1,201*** 1,264*** (0,066) (0,068) (0,062) Standard errors in parentheses reported underneath coefficients
*** p<0.01, ** p<0.05, * p<0.1