JOB INSECURITY, STRESS, ABSENTEEISM AND PRESENTEEISM: THE ROLE OF FINANCIAL NEEDS.

PRESENTEEISM: THE ROLE OF FINANCIAL NEEDS.

Master thesis University of Groningen Faculty of Economics and Business MSc Human Resource Management

Groningen, June 2014 Zefi Apergi Student Number: S2559773 Pelsterstraat 16B1 9711 KL Groningen, Netherlands Tel: + 31 6 49 23 88 72 E-mail: z.a.apergi@student.rug.nl

2

ABSTRACT

Job insecurity has been recognized as a serious work stressor with negative health related outcomes. While stress unfolds to increase absenteeism because of illness, the adverse outcome, presenteeism, is also possible. A number of employees report sick when experiencing poor health, while others, attend work regardless. The current study investigated the role of financial needs in the relation between job insecurity, stress, absenteeism and presenteeism. In order to interpret this role, I used the “stress” and “disciplinary” theory as an explanatory mechanism on how working individuals may decide whether to call in sick. Using data from a survey among 173 Greek employees from various organizations, results showed that while job insecurity is a good predictor of stress, it decreases rather than increases absenteeism. However, the moderating effect of financial needs on the two opposing working behaviors was not supported.

3

Table of Contents

3.1 Procedures and Participants ... 20

4 1. INTRODUCTION

Job insecurity refers to the overall concern about the future existence of one’s employment (De Witte, 1999; Jacobson, 1991). As it is associated with feelings of uncontrollability and unpredictability about the working future, it has been conceptualized as a serious work stressor (Ashford, et al., 1989; De Witte, 1999; Sverke et al., 2002). Prolonged feelings of job insecurity are closely related to stress (Hartley et al., 1991) and can be detrimental for the physical and mental wellbeing. Evidence shows that several somatic and psychological problems are attributed to excessive stress levels (Viljoen et al., 2005).

Considering the above, workplace stress in the form of job insecurity implies declining physical and mental health rendering employees unable to work. In that sense, stress is a factor that can increase employees’ absenteeism (De Souza et al, 2006). The above is consistent with the “stress” theory which suggests that job insecurity is an indicator of decreased health and consequently increased absenteeism (Blekesaune, 2012).

While the health related outcomes of stress could lead to high absence rate, several factors may force employees to attend work despite feeling impaired health. Presenteeism is defined as showing up to work despite sickness (Gosselin & Lauzier, 2011). A possible explanation for this attitude is given by the “disciplinary” theory which suggests that in times of high unemployment working individuals may be reluctant to be absent from work because of the fear of involuntary job loss.

5 The contribution of the study is twofold. First, with the proposed conceptual framework, the paper contributes to the existing scientific knowledge by adding the “economic dependence” as a critical moderator in the relation between job insecurity, stress, absenteeism, and presenteeism. Second, the study may have some practical relevance for the organizations as it implies possible design interventions in order to prevent potential negative outcomes of both behaviors.

6 2. THEORETICAL BACKGROUND

2.1 Context of Greece

Greece was affected by the global economic crisis of 2008-2009 more than any other European country. After many consecutive years of growth, Greek economy entered into the most severe recession in its modern history with the financial crisis affecting all major sectors of the Greek economy (Zavras et al., 2012). From 2010, when the Greek government signed the Memorandum of Economic and Financial Policies with the international lenders, Greek citizens started facing serious socioeconomic turmoil (Economou et al., 2013). This is reflected in the labour market conditions that have significantly deteriorated, since opportunities in employment are decreasing and unemployment rates are soaring.

The current prolonged crisis with unemployment rising sharply in 2010-2011 (Eurostat, 2013) shows poor prospects of immediate recovery (Vandoros et al., 2013). Interestingly, the latest evidence of Eurostat shows that Greece has reached the highest unemployment rate of 27.3% in August 2013 among all the European countries. According to the European Foundation for the Improvement of Living and Working Conditions (2013), job insecurity has sharply increased in Greece. Specifically, the percentage of working people who thought that it is very likely or quite likely they would lose their job within the coming six months has risen from 8.2 in 2007 to 30.57 in 2012, consisting the sharpest increase among the European countries. Consequently, job insecurity is considered a major issue in Greek people’s life and well-being.

7

2.2 Job insecurity

There are several definitions of the concept of job insecurity in the literature. Greenhalgh and Rosenblatt (1984) defined job insecurity as the perceived powerlessness to maintain desired continuity in a threatened job situation. This approach has two basic dimensions; the severity of the threat and the degree of powerless and helplessness to deal with it. That means that employees feel job insecurity when they perceive the threat as severe and do not feel capable of dealing with it. Job insecurity reflects a fundamental and involuntary change in the continuity and security within the employing organization (Greenhalgh & Rosenblatt, 1984; Sverke & Hellgren, 2002). It can be regarded as the phase between employment and unemployment where employees cannot predict their working future (Hartley et al., 1991). The above definitions emphasize the threats of imminent job loss. Nevertheless, some commentators consider these definitions too narrow, because they fail to take into account other dimensions such as the deterioration of employment conditions, lack of salary development or career opportunities (Ashford et al., 1989; Greenhalgh & Rosenblatt, 1984; Roskies & Louis-Guerin, 1990). The current study considers job insecurity as the perceived likelihood of involuntary job loss (Sverke et al., 2002).

A distinction can be made between cognitive and affective job insecurity. Cognitive job insecurity refers to the perceived likelihood of losing one’s job (Borg & Elizur, 1992) and it stems from employees’ perceptions of possible job changes (Huang et al., 2012). Affective job insecurity refers to the fear of job loss as an emotional experience (Borg & Elizur, 1992). Alternatively, it can be described as being worried or emotionally concerned about potential changes in the organization (Huang et al., 2012). What is worth mentioning is that information sharing and communication at all levels of hierarchy may be crucial in reducing cognitive job insecurity (Huang et al., 2012). When an organization provides sufficient information and enhances communication about present or potential crises and organizational changes that will be encountered, cognitive job insecurity may be diminished. As employees would be well informed about organizational issues, information sharing and communication would also pose an indirect impact on affective job insecurity through employees’ cognitive evaluation of the situation (Huang et al., 2012).

8 uncertainties in the work environment even if they are exposed to the same objective situation (Greenhalgh & Rosenblatt, 1984; Hartley et al., 1991; Sverke & Hellgren, 2002). Due to the different evaluation and interpretation of a work situation and its threatening circumstances, individuals may experience different degrees of job insecurity. While job loss is an objective and distinct event, job insecurity is a purely perceptual phenomenon (Jacobson, 1991) stemming from subjective experiences (De Witte, 1999; De Witte, 2005; Sverke & Hellgren, 2002). Thus, the number of employees who experience job insecurity is considerably larger than the number of employees who actually undergo job loss (Jacobson, 1991).

9

2.3 Stress

The context of change and uncertainty can generate feelings of job insecurity. Prolonged feelings of job insecurity are closely related to stress (Hartley et al., 1991). The European Agency for Safety and Health Issues (2009) attributes several symptoms such as overall fatigue, headaches, muscular pain, irritability, and sleeping problems to work related stress. In addition, numerous studies have shown that the assessment that somebody will lose his/her job is a reason to provoke symptoms of anxiety and depression (Burchell, 2009).

Stress is fundamentally seen as a physical, embodied experience emerging from a set of interrelated circumstances and processes (Styhre et al., 2002) that can cause physical and mental health impairment (Godin et al., 2006). Lazarus (1966) defines stress as a condition that “occurs when an individual perceives that the demands of an external situation are beyond his or her perceived ability to cope with them”. Stated more simply, uncontrollability elicits feelings of stress. When individuals evaluate the stressor as a threat, their reactions are likely to be negative (Lazarus & Folkman, 1984).

It seems that occupational stress is an increasing phenomenon. A large-scale survey at Princeton (1997) showed that stress is experienced at all levels by the majority of employees as they feel “quite or extremely stressed” at work. Since work is viewed as a pivotal part of people’s lives and is related with economic, social, and personal satisfaction (Sverke et al., 2002), it is logical that job insecurity can increase stress levels because of the anxiety about the nature and existence of one’s job in the future. Excessive and continuous feelings of job insecurity can generate negative stress with immediate as well as long term outcomes (Lazarus & Folkman, 1984). As has already been mentioned this can be reflected in the individual’s physical and mental condition. Many somatic health problems can be related to stressors (Dirken, 1969). The duration and intensity levels of a stressor are seen as critical aspects of psychobiological stress (Dirken, 1969). Stress reflects not only the anxiety, discomfort and feelings of threat that appear as psychological reactions to stressors but also the somatic side effects (Dirken, 1969). In this study, the somatic side effects will be examined in terms of subjective health complaints.

10 pressure of “modern life” (Eriksen, & Ursin, 2002). Subjective health complaints are so common and frequent among individuals that can almost be characterized as “normal”. In their illustrative study for developing a scoring system for subjective health complaints, Eriksen, Ihlebaek and Ursin, (1999) showed that at least 75% of a normal population had at least one complaint during the examined period of 30 days. Interestingly, these complaints are not grounded in pathological findings or serious organic changes (Eriksen, & Ursin, 2002), making medical diagnosis difficult. It should be noted that, even though they are conditions with few or no pathological findings, they may reach levels that require medical assistance and sickness benefit (Eriksen, Ihlebaek & Ursin, 1999). As subjective health complaints can turn into intolerable conditions, they play a prominent role in short and long term sickness absence (Tellnes et al., 1989). Finally, according to Eriksen, Ihlebaek and Ursin, (1999) the concept includes the following five categories of health related problems; “musculoskeletal” (e.g., headache, neck pain, low/upper back pain), “pseudoneurological” (e.g., sleep problems, tiredness, sadness/depression), and “gastrointestinal” (e.g., stomach discomfort, dyspepsia) complaints, “allergy” (e.g., asthma, breathing difficulties, eczema), and “flu”.

The psychological dimension of stress can be expressed in terms of burnout which is a multidimensional, chronic and ongoing stress reaction that happens when individuals fail to cope with various stressful events (Rothmann et al., 2003). The longitudinal epidemiological study of Kant, Jansen, Van Amelsvoort, Mohren, and Swaen (2004) confirmed its chronic nature. In a sample of over 12.000 Dutch employees, it has been estimated that the average duration of severe burnout was about 2.5 years.

11 is a constant negative feeling towards work which can result in exhaustion, distress, low motivation, decreased effectiveness and poor attitudes. The present paper focuses on burnout as a form of work related stress in terms of emotional exhaustion. As will be presented in the following section, burnout and consequently emotional exhaustion is a severe aspect of stress as it is related to absenteeism, among other impending outcomes.

Overall, when individuals feel that their job is not secure, work related stress increases, which can cause poor mental and physical health. In this study both employees’ physical and mental condition are examined in terms of psychosomatic complaints and emotional exhaustion, respectively.

The above leads to the first hypothesis below:

H1: There is a positive relation between job insecurity and stress levels.

2.4 Absenteeism

As stated earlier, the experience of job insecurity can have detrimental effects on individual’s health through increased stress. Job insecurity is one of the most serious work related stressors employees have to face (Ashford, et al., 1989; De Witte, 1999; Sverke et al., 2002). Evidence shows that there is positive relation between job insecurity and self reported physical health complaints as well as objective health problems (Ferrie et al., 2001; Kinnunen & Natti, 1994). Downsizing and structural changes are highly linked to periods of economic downturns where feelings of job insecurity dominate. In this sense, they may account for employees’ poorer health at both the physical and psychological level rendering employees unable to work. Consistently, job insecurity appears to raise sickness absence at work. This positive relation has been shown in several studies (e.g. D’ Souza et al., 2006, Mohren et al., 2003, Probst, 2000).

12 absence imposes a burden for companies and states. For instance, Bozionelos and Sun (2013) stated that for a country similar to Spain in size, the total time lost because of sickness absenteeism is the same as if 155.000 workers were excluded from the labor force for one year. In addition, the burden imposed on a state can be illustrated by considering that for some European countries the cost related to sickness absenteeism is estimated about 5% of GDP (Bozionelos & Sun, 2013).

According to Alexanderson (1995), although sickness absenteeism can be viewed from different angles, it is primarily an indicator of impaired health. Sickness absence is one form of excused absenteeism meaning that it is medically certified (Bekker et al., 2009). It also provides an indication of the employee’s health and wellbeing (Ybema et al., 2010) and is related not only to immediate illness but also to conditions of prolonged stress in the working reality (Alexanderson, 1998). Absence from work is a choice, to some extent, since it is up to employees’ interpretation whether they are capable or incapable to be present at work (Bozionelos & Sun, 2013). According to the European Agency for Safety and Health at Work (2009), stress accounts for 50-60% of all lost working days. Thus, workplace stress accounts for a considerable portion of employee absenteeism.

13 as a coping mechanism in order to gain control over what might have become intolerable in terms of physical and emotional fatigue.

The above is consistent with stress theory, which examines the impact of external shocks such as layoffs, downsizing, unemployment, and reorganization on health and subsequently on sickness absence (Blekesaune, 2012). Data shows that downsizing is usually related to increases in absenteeism of remaining staff (Roed & Fevang, 2007), making clear that job insecurity is an indicator of decreased health and consequently, increased absence. In other words, a stressor such as job insecurity may increase work absences. For the so-called ‘‘stress’’ models (De Boer et al., 2002; Johns, 1997), absenteeism is viewed as the outcome of stressful work conditions. When employees are incapable of working because of illness or act in a proactive way trying to prevent the possible deterioration of their health, they may decide to be absent from work. Overall, stress theory suggests that increasing workplace stress will lead to increasing absenteeism.

With stress being the main reason for increased absenteeism the following hypothesis is formulated.

H2: Stress mediates the relationship between job insecurity and sickness absenteeism.

2.5 Presenteeism

14 To understand presenteeism, we first need to identify its origins. Several work related factors such as job stress and insecurity (Heponiemi et al., 2010) may evoke pressure to attend work despite feeling unhealthy. It seems that presenteeism may be a rising trend because of the greater reluctance of employees to be absent from work in times of economic uncertainty, downsizing, and staff shortages (Baker-McClearn et al., 2010). Among these reasons, is the fear of being replaced, which is a common feeling in periods of economic recessions (Grinyer & Singleton, 2000). A possible explanation is that absent employees may regard themselves as the first to be fired because employers would be aware of the costs derived from absent employees (Staufenbiel & Konig, 2010). Alternatively, the substantial costs produced by absenteeism may consist a risk for job termination and unemployment. This line of reasoning proposes that the risk of dismissal can be compensated by decreased absenteeism.

Consistent with the above is the “disciplinary effect”, which suggests that the higher the unemployment rate, the higher the period someone would remain unemployed without receiving income in case of dismissal (Shapiro & Stiglitz, 1984). For this reason, in times of high unemployment, some workers are “disciplined” from taking time off work due to illness and rest at home (Blekesaune, 2012). Insecure employment, as well as the poor prospects of finding an alternative occupation are crucial determinants when deciding to report sickness at work. Indeed, the “discipline mechanism” stems from workers’ attempt to limit the risk of potential layoffs (Livanos & Zangelidis, 2013). According to this theory, the choice of staying home using illness absence is viewed as “leisure” and therefore, employees decide deliberately between work and leisure (Shapiro & Stiglitz, 1984; Leigh, 1985). In sum, the disciplinary theory suggests that job insecurity leads to less absence but to more presenteeism. The longitudinal study of Livanos and Zangelidis (2013) showed that in countries with high unemployment, the sickness absence rate is low. In contrast, in times of employment security where employees feel no threat to lose their job, they tend to take more days of sick leave absence. Therefore, job security may have a negative effect on presenteeism and no impact on absenteeism (Gosselin et al., 2013).

15 postpone sickness absence from work, they put themselves at risk. This risk implies that by inhibiting recuperation, minor health issues may be transformed into more serious ones (Grinyer & Singleton, 2000). In this respect, the risk of longer term ill health may increase, with the individual being more prone to physical and mental declining health (Aronsson & Gustafsson, 2005; Demerouti et al, 2009). To illustrate this, Kivimaki et al. (2005) showed that the possibility of coronary heart disease was double in the group of unhealthy employees that had never taken sickness absence in comparison to the group of unhealthy employees with moderate use of sickness absence. In addition, the follow-up study of Bergstrom et al. (2009) demonstrated that sickness presenteeism seems to be an independent risk factor for future poor general health. Another remarkable finding is the relation between absenteeism and presenteeism, although not much research has been conducted. Specifically, Aronson et al. (2000), Hansen and Andersen (2008) and Janssen et al. (2013) found a significant positive relation between presenteeism and sickness absence with the last study confirming the association between presenteeism and future long spells of sickness absence.

Although at a first glance presenteeism could be seen as attractive for the organizations (Demerouti et al., 2009), it could, however, impose momentous costs on productivity (Goetzel et al., 2004). According to some authors, presenteeism can induce much more or more significant losses than absenteeism does (Stewart et al. 2003; Goetzel et al., 2004; Schultz & Edington, 2007). Roe (2003) supported that poor health status can lead to decreasing performance at both the individual and collective level. To illustrate this point, Roe (2003) argued that on the one hand, individual performance may diminish as “sick” employees would have to invest more time and effort to achieve the same output as the healthy colleagues. Workers would not be able to reach their full capacities and thus, productivity losses would be inevitable. On the other hand, since healthy employees get in contact and cooperate with sick colleagues, the transmission of an infectious illness is possible. In total, presenteeism can be detrimental for the performance of the entire organization.

16 absenteeism such as job insecurity (Aronsson et al., 2000), should also be associated with presenteeism. At this point, it should be stressed that the findings of MacGregor et al. (2008) suggested that a stressful life event, such as job insecurity, can be associated to the same extent with both absenteeism and presenteeism. Based on the above, the present study regards the feelings of stress as having a double effect on these two opposing working attitudes. Thus, absenteeism and presenteeism are viewed as two sides of the same coin that could equally be adopted by employees, in response to stress.

In conclusion, this section has argued that the stressful event of job insecurity may create fewer incentives for employees to report sick, in order not to jeopardize employment.

The above results in the third hypothesis below:

H3: Stress mediates the relationship between job insecurity and presenteeism.

2.6 Financial need

The evidence presented in the previous sections suggests that job insecurity can affect employees behaviors in two distinct and opposing ways. On the one hand, the health related outcomes of stress could lead to high absenteeism whereas, on the other hand, the fear of involuntary job loss in the context of general uncertainty could prevent employees from recovering at home when they are sick. In an attempt to reconcile the conflicting perspectives and findings, the present study proposes that the above dilemma in employees’ behavior may depend on economically based worries; the financial needs or economic dependence.

17 It is reasonable that the amount of money an individual needs could vary depending on personality and background traits and economic or family circumstances (Shaw & Gupta, 2001). George and Brief (1990) suggested that the family configuration affects the employees’ reactions to their job. For instance, men who are the main supporters of their household or who have many dependents may be more sensitive in their job’s economic attributes. In other words, people who have to support financially their family members or without any alternative source of income tend to be more financial needy than those who are singles or have more income sources (Shaw & Gupta, 2001). Another factor that could play a role in the degree of economic dependence on one’s job is the possibility of finding an alternative occupation. Specifically, employees who lack employability and whose main source of income is provided by the job under threat would be more dependent on that job than employees with high chances of finding another employment and alternative sources of income (Greenhalgh & Rosenblatt, 1984).

Economic dependence plays a decisive part in peoples’ life and job insecurity may be more than a threatening situation in respect to the economic aspect of a job (Richter et al., 2013). Nolan’s (2009) analysis demonstrates that the provision of money is crucial for some people’s identity within the family. Moreover, worries about the future existence of one’s employment can lead to feelings of low self esteem (Nolan, 2002). In this vein, we assume that employees who highly depend on their job with increasing financial needs will regard job insecurity as a more dangerous threat than the ones who depend less on their job.

Economic dependence is also related to the perception of choice. Doran and colleagues (1991) considered that low financial requirements are linked to more freedom of choice. Conversely, because of the greater financial obligations or the lower economic security, employees with high financial needs experience more limitations in their choices. Consequently, as financial requirements increase, employees would be more economically dependent on their job related income (Doran et al., 1991) and hence, they may adopt different working behaviors.

18 higher the prospect of job loss, the higher the consequent costs for individuals in case of layoff. Thus, workers may feel the responsibility to be present at work despite illness, in order to reduce the possibility of being dismissed. In particular, the high dependence on the work related income may foster the financial pressure to attend work rather report absent, in times of job insecurity. In other words, workers may substitute sickness presence for absence, depending on the centrality of their work related income.

In sum, in times of economic uncertainty, where strong feelings of job insecurity prevail, potential job loss is highly related to the economic aspect of the job as it promotes economic stability and the person’s identity (Richter et al., 2013). For this reason, the current study proposes that the magnitude of economic dependence may contribute to absenteeism or presenteeism behaviors. In particular, with higher levels of financial needs, the experience of job insecurity and associated stress will be less strongly related to absenteeism and more strongly to presenteeism: employees chose to go to work, despite feeling unwell, if financial needs are high.

The above arguments lead the following hypotheses:

H4a: The relation between job insecurity and absenteeism is moderated by employees’ financial needs.

H4b: Job insecurity is more/less negatively associated with absenteeism when financial needs are high/low.

H5a: The relation between job insecurity and presenteeism is moderated by employees’ financial needs.

H5b: Job insecurity is more/less positively associated with presenteeism when financial needs are high/low.

19 Considering the set of hypotheses in the theory section, I hypothesize that the indirect effect of job insecurity, as transmitted through experience of stress, on absenteeism and presenteeism, respectively, will be either mitigated or strengthened by the employee’s financial needs.

20 3. METHODOLOGY

3.1 Procedure and Participants

21 3.2 Measures

Job insecurity

To measure the independent variable, the Job Insecurity Questionnaire (JIQ) developed by De Witte (2000) was used. This measure consists of 11 items that reflect both the cognitive and affective dimension of job insecurity. A sample item of cognitive and affective job insecurity, respectively, is “I think that I will be able continue working here” (reverse coded) and “It makes me anxious that I might become unemployed”. Participants were asked to indicate their level of agreement according to a 5-point Likerd scale (1=Strongly disagree, 5=Strongly agree). A higher score indicates higher level of job insecurity. The Cronbach’s a coefficient was .93, indicating good internal consistency.

Stress

Stress was measured on two dimensions; psychosomatic complaints and emotional exhaustion. The somatic and psychological symptoms were assessed by the Subjective Health Complaint Inventory developed by Eriksen et al. (1999), which incorporates elements of both physical and mental health. The measure contains 29 different subjective health complaints. Participants were asked to specify, the extent to which they had been bothered by each symptom on a 4-point scale (1=not at all, 4=serious) in the last 30 days, regardless of their objective health condition. The items include five subscales; musculoskeletal pain (e.g., headache, neck pain, upper/low back pain, migraine), pseudo-neurology (e.g., extra heartbeats, heat flushes, sleeping problems, tiredness, dizziness, sadness) gastrointestinal problems (e.g., heartburn, stomach discomfort, ulcer/non ulcer dyspepsia), allergy (e.g., asthma, breathing difficulties, eczema, chest pain), and flu (e.g., cold/flu, coughing). They cover a wide range of the most common complaints of the working population and for this reason, the SHC inventory is considered as a simple and reliable instrument (Eriksen et al., 1999). Because for this measure there were many missing values, I included respondents who have answered at least 20 out of 29 items in the analysis. The coefficient alpha for this scale was .80. After averaging the items, the mean was 1.87 with a standard deviation of .53

22 from my work” and “Working all day is really a strain for me”. Respondents rated the frequency of each statement on a 7-point Likert scale ranging from “1=never” to “7=every day”. This scale had satisfactory internal consistency with alpha coefficient being .88 (M = 4.3 SD = 1.52).

A composite stress measure was computed by first standardizing the two scales of psychosomatic complaints and emotional exhaustion and then averaging the score on the two measures. The correlation of the two measures was significant (r=.57, p<.01).

Absenteeism

Sickness absenteeism was assessed in terms of duration and frequency. Two items taken from the study of Pousette and Hanse (2002) were used. For the first item, participants reported the total number of days they had been absent from work (duration), during the past 12 months. The number of days was categorized into five categories (zero, 1-3, 3-9, 9-29 and more than 29 days), (M=1.74, SD=.86). The second question reflects the absence frequency, during the past 12 months and has five alternative answers (zero, 1-2, 3-4, 5- 6 and more than 6 times) (M=1.65, SD=.79) These two items had a bivariate correlation of r=.79, p<.01 and internal consistency .89. The composite variable of absenteeism was computed after calculating the Z scores of each item.

Presenteeism

23 Financial Needs

In order to evaluate respondents’ degree of economic dependence, I used the additive index adapted from George and Brief (1990) and Doran and colleagues (1991). It consists of questions about the marital status, spouse’s employment status, number of children aged 22 and under, alternative income sources and the person’s perception of his/her ability to find an alternative job similar to the current. Single and married respondents with a full time working spouse are viewed as having lower financial requirements and for this reason they are assigned a score of 0. In contrast, respondents with a part time or non working spouse are assigned a score of 1. Conceptually, the number of children increases the financial burden. For this reason, respondents without children are coded with 0, while each child below 22 years is coded with 1. Moreover, additional income sources (second jobs, rental income) render employees less financially dependent on their job (Ehrenberg & Smith, 1988; Hill, 1987). Consequently, respondents whose household income derived from job is less than 50% is coded with 1. In case this income ranges from 51 to 75% or it exceeds 75%, it is valued with 2 and 3, respectively. Finally, the ability to find alternative employment with similar financial attributes evokes less financial pressure to remain on the job (Bartel, 1979). Employability was measured by one item developed by Veiga (1981). Specifically, no chance or a slight chance, 50-50, and higher than 50-50 chance of finding another employer were valued 1, 0 and -1, respectively. The four items described above were used in a recent study of Brett, Cron and Slocum (1995).

Control Variables

Several relevant demographics and job characteristics that co-vary with job insecurity namely, gender, age and type of employment contract (De Witte, 2005; Naswall & De Witte, 2003) were added as control variables.

24 a 5 point scale ranging from 1 “never” to 5 “always”. Higher scores indicate higher level of autonomy. The alpha coefficient of the scale was .84.

Regarding the second variable, certain job demands, such as work overload, are recognized as job stressors and are related to impaired health in the working environment (Schaufeli et al., 2009). In addition, work overload has been recognized as a crucial determinant of presenteeism (Demerouti et al., 2009) and therefore, it was included as control variable. The workload scale was developed by Furda (1995), using a 5-point Likert type response anchored by “1=never” and “5=always”. Sample items include “Do you have to work very fast?” and “Do you usually have enough time to finish your work?” (recoded). Higher scores imply increased workload. The scale had good internal consistency with Cronbach’s a being .78.

3.3 Data analysis

Statistical analysis was performed using SPSS 20.0 statistical software for Windows. In order to evaluate the study hypotheses, I conducted a simple mediation and a moderated mediation analysis as recommended by Preacher, Rucker, and Hayes (2007) using the Macros provided by Hayes1 (2013). The extent to which the relationship between job insecurity and absenteeism/presenteeism is mediated by feelings of stress and the extent to which the relationship between stress and absenteeism/presenteeism are moderated by employees’ financial needs were tested. First, the indirect effect of the independent variable (job insecurity) on the two dependent variables (absenteeism and presenteeism), through the mediator (stress), was tested by running the simple mediation analysis (Model 4) of Hayes (2013). Then, the analysis utilized the conditional indirect effect illustrated by Model 14 of the same author, where the relationship between the mediator and the dependent variable is moderated by the moderator. Therefore, job insecurity was entered as independent variable, absenteeism and presenteeism as dependent variables, stress as mediator, and financial needs as moderator (V). A series of control variables namely, gender, age, employment contract, autonomy and workload were included as covariates. The two models were ran separately for each dependent variable (absenteeism and presenteeism). The confidence interval was set to 95%.

25 4. RESULTS

4.1 Descriptive Statistics

Correlation analysis was performed to examine the associations among major variables. Table 1 presents the bivariate correlations among study variables. As predicted, job insecurity shows a positive correlation with stress (r = .43, p < .01). Moreover, job insecurity exhibits a negative association with absenteeism (r = -.18, p

< .05) and a positive correlation with presenteeism, although not significant (r = .14).

Stress seems to be positively related to presenteeism (r = .23, p < .01). However, no significant correlation was observed between stress and absenteeism (r = -.03). Interestingly, the two working behaviors of absenteeism and presenteeism were negatively, but not significantly correlated (r = -.12). Furthermore, financial needs were found to relate positively to absenteeism (r = .06) and negatively to presenteeism (r = -.07). Nevertheless, none of these correlations is significant.

It should be noted that significant correlations between the main study variables and the control variables (gender, age, employment contract, autonomy, workload) were found. First, stress and absenteeism are positively related to gender (r = .26, p < .01;

r = .21, p < .01). Second, age seems to be associated with only financial needs (r = .49, p < .01). Regarding the examined job characteristics, the type of employment

contract is positively correlated with job insecurity (r = .27, p < .01) and negatively with absenteeism (r = -.23, p < .01) and financial needs (r = -.27, p < .01). Workload shows a positive relation with stress (r = .47, p < .01), while it is negatively related to absenteeism (r = -.23, p < .01). Finally, results show a negative association between autonomy and job insecurity levels (r = -.31, p < .01).

4.2 Hypotheses testing

The simple mediation and moderated mediation results are summarized in Table 2. Hypotheses 1, 4 and 5 were tested by the moderated mediation model while Hypotheses 2 and 3, by the simple mediation model.

Hypothesis 1 predicts that job insecurity is positively related to stress. The results of the regression analysis show a strong positive relation between the two variables (b =

.74, p < .001). Zero is not included in the confidence interval; therefore, the direct

26 results show that work related stress is being affected by gender (b = .96, p < .001). In particular, it seems that females are experiencing more stress than males. Finally, workload poses a strong positive effect on employees’ feelings of stress (b = 1.11, p

< .001).

Regarding the outcomes for the dependent variables, results indicate a significant negative effect of job insecurity on absenteeism (b = -.76, p < .01). In contrast, this effect although positive, is not significant for presenteeism (b = .07, p = .59). As suggested in theory, it seems that stress affects positively both absenteeism and presenteeism. However, none of these effects is significant (b = .24, p = .44; b = .14,

p = .36). Among the control variables, gender is positively associated with

absenteeism (b = 1.10, p < .01), with females being more prone to this working attitude. Finally, a negative significant effect is noticed between absenteeism and workload (b = -.89, p < .01). No significant relations were found between presenteeism and any of the control variables.

Hypotheses 2 and 3 predict that stress mediates the relation between job insecurity, absenteeism and presenteeism. Results of the simple mediation analysis show that the indirect effect of job insecurity on both absenteeism and presenteeism is positive, although not significant (b = .13; b = .06). The non significance of the effect can be inferred from confidence interval which includes zero. Thus, no evidence is found for the mediational impact of stress in the relationship between job insecurity and absenteeism/ presenteeism; therefore, hypotheses 2 and 3 are not confirmed.

Hypotheses 4a and 5a predict that the relation between stress, absenteeism and presenteeism is moderated by financial needs. The conditional indirect effects were calculated using the bootstrapping procedure suggested by Preacher et al. (2007). The results (see lower part of table 2) show the relation between job insecurity, absenteeism and presenteeism (through stress) at three levels of financial needs; the mean, one standard deviation above the mean and one standard deviation below the mean. As can be seen, none of the relations is significant; hence, the moderating effect cannot be supported. Overall, no support was found for Hypotheses 4a and 5a and consequently for Hypotheses 4b and 5b.

27

TABLE 1

Means, Standard Deviations, and Pearson Correlation

28

TABLE 2

Results of Regression Analysis

Notes. CI = Confidence interval; LL = lower limit; UL = upper limit. SD = standard deviation; M = mean

Unstandardised regression coefficients are reported; standard errors in parentheses. Bootstrap sample size = 1,000. *p< .05; **p< .01; ***p< .001

DV DV

Predictor Stress B (SE) Absenteeism B (SE) Presenteeism B (SE)

Control variables Gender .96 (.25)*** 1.10 (.37)** -.11 (.19) Age -.00 (.01) .01 (.02) .01 (.01) Employment contract -.24 (.35) .45 (.50) .06 (.25) Autonomy .22 (.19) -.19 (.26) -.05 (.13) Workload 1.11 (.19)*** -.89 (.31)** .26 (.16) IV Job insecurity .74 (.17)*** -.76 (.26)** .07 (.13) M Stress - .24 (.32) .15 (.16) Financial needs - .19 (.15) -.10 (.07)

Stress x Fin. needs - -.02 (.08) -.02 (.03)

R2 .42*** .20** .09

Simple Mediation Bootstrap indirect effects on abs./pres.

(through stress) _{B (SE) LL UL B (SE) LL UL}

IV Job insecurity - .13 (.10) -.03 .34 .05 (.06) -.08 .17 Conditional indirect effects on abs./pres. at three levels of financial needs B (SE) 95% CI B (SE) 95% CI LL UL LL UL

Low fin. needs

-1SD (2.03) - .14 (.16) -.15 .48 .07 (.08) -.07 .25

Average fin. needs

M (3.49) - .11 (.10) -.06 .31 .05 (.06) -.07 .17

High fin. needs

29 5. DISCUSSION

This study investigated the role of financial needs in the relation between job insecurity, stress, absenteeism and presenteeism. By introducing the “stress” and “disciplinary” theory, the study provided new insights into the dynamics of stress, and specifically the way it could affect absenteeism and presenteeism. I developed and tested a model that explored the mechanisms that linked the antecedents of job insecurity associated with stress, absenteeism and presenteeism as two opposing working attitudes, and the moderating effect of financial needs in the relation between stress, absenteeism and presenteeism. I expected that stress would act as mediator in the relation between job insecurity, absenteeism and presenteeism and that the level of financial needs would either strengthen or attenuate the relation between job insecurity associated with stress, absenteeism and presenteeism. Data were collected from an online survey among 173 Greek employees who worked in various organizations.

5.1 Findings

The contribution of the study points to the role of financial needs as a critical moderator in the relation between job insecurity, stress, absenteeism and presenteeism. Several key findings were reached. First, it should be mentioned that although a suppressor effect was noticed in the relation between stress and presenteeism, all other effects were consistent with the correlations, indicating that suppression was not a serious threat to the findings.

The first finding suggests that the perceptions of job insecurity reflect a stressful work experience. More specifically, results show that women experience more feelings of stress in comparison to men. A possible explanation is that stress results from conflicting demands or expectations associated with different roles (role conflict). Women’s work together with increased household and childbearing responsibilities could create a role conflict which in turn could increase stress. Another explanation could be that women cope differently with stressful conditions than men do.

30 distinction was observed; although in total job insecurity reduced absenteeism, it seems that women are more prone to absenteeism than men. The above difference can be attributed to physical differences. A large body of literature regards women as the “sicker” sex (Bekker et al., 2009). Furthermore, according to substitution theory, I expected that low and high absenteeism will result in high and low presenteeism, respectively. Specifically, I expected that the negative effect of job insecurity on absenteeism will result in a positive effect on presenteeism. Nevertheless, no such relation was found.

Regarding the control variables, the third key finding is that workload is positively related to stress and negatively related to absenteeism. However, no significant relation between workload and presenteeism was observed. In fact, the findings suggest that the amount of responsibilities a worker has to deal with increases stress while reduces absenteeism. Although, I expected that temporary employees would experience more stress in terms of job insecurity because of higher unemployment risk (Virtanen et al., 2006), results show that employment contract does not influence this relation. Another prediction was that people in temporary positions would be less prone to absenteeism because temporary jobs may not entitle employees to paid sick leave (D’ Souza et al., 2006). Nevertheless, no significant correlation was found between employment contract, absenteeism and presenteeism. A potential explanation for the above two findings is that for non-permanent workers, job insecurity is an integral part of their working life and thus, they become more “inured” and less vulnerable to job insecurity (Virtanen et al., 2003). Moreover, while high degree of automy in one’s job was expected to alleviate workplace stress and consequently lead to less absenteeism, no significant relation was observed between the variables. A possible explanation is that the sample’s level of autonomy was relatively high and for this reason, employees may have taken this job characteristic for granted. Finally, the experience of stress and the decision to report sick is not affected by workers’ age. The mediation hypotheses 2 and 3 were not confirmed. Consequently, workplace stress associated with job insecurity cannot be seen as a factor that leads to absenteeism and presenteeism behaviors. An alternative interpretation is that work overload could play that role as significant relations with both stress and absenteeism were noticed.

31 that the feelings of stress created by the fear of dismissal could “discipline” some workers from being absent while experiencing declined health. In that sense, high financial needs could lead to less absenteeism but to more presenteeism. Results show that this reasoning is not confirmed. No evidence for the moderating impact of financial needs in the aforementioned relations was found. Thus, the decision to report sick does not depend on the economic dependence on one’s job.

Overall, the analysis suggests that employees who experience job insecurity report less sickness absence. While job insecurity seems to be a good predictor of stress together with work overload, the decision to be absent or present at work when sick is not attributed to stress and the “disciplinary effect”. In addition, financial needs showed no effect on the above behaviors. According to the results, work overload affects negatively absenteeism and thus, it can be seen as a potential causal factor influencing that behavior. The present research suggests that reduced absenteeism in times of high job insecurity could be attributed to increased workload.

5.2 Theoritical implications

The findings described above set the ground for several theoretical implications. First, the study strengthens the evidence that job insecurity is an important stressor (Ashford, et al., 1989; De Witte, 1999; Sverke et al., 2002). Second, previous researches that have focused on the impact of job insecurity on absenteeism showed conflicting findings (e.g D’ Souza et al., 2006; Livanos & Zangelidis, 2013). The present paper showed that high job insecurity is related to reduced absenteeism and is consistent with the study of Livanos and Zangelidis (2013). Third, the finding that women are more prone to absenteeism is supported by several researches (e.g. Patton & Johns, 2007; Bekker et al., 2009; Livanos & Zangelidis, 2013).

32 cannot be seen as the explanatory mechanisms for employees’ decision to repost sick. The above dilemma in employees’ decision may be attributed to other factors such as work overload. Results showed that high workloads are positively related to stress and negatively related to absenteeism. For this reason, this study reinforces the idea job insecurity is related to excessive workload and more working hours (De Cuyper et al., 2008) and that in this case, employees take fewer days of sickness absence. The above findings are contrary to the study of D’Souza et al. (2006) which observed a positive relation between heavy workloads, perceived insecurity and sickness absence. Indeed, job demands could not only imply pressure to work harder, but also pressure to attend work. Thus, less absence may be seen as a sign of high commitment (Demerouti et al., 2009).

In this paper, absenteeism and presenteeism were viewed as two sides of the same coin. Based on the “substitution hypothesis” (Caverley et al., 2007) factors that preclude being absent when ill should prompt presenteeism. In other words, if absence is inadvisable during periods of high job insecurity, the substitution hypothesis would predict presenteeism instead. The results of the current study showed a strong negative effect between job insecurity and absenteeism. The expected positive effect between job insecurity and presenteeism was not found and hence, employees do not substitute presenteeism for absenteeism. In conclusion, the substitution hypothesis for the factor of job insecurity and associated stress cannot be supported. The above finding should provide food for further investigation.

5.3 Practical implications

This study provides several practical implications for managers, HR practitioners and policymakers at a national level. The present findings suggest that job insecurity and excessive workloads increase stress. Thus, organizations should invest in the well being of their personnel. Wellness programs and professional assistance such as consultation and coaching seem to have promising effects in coping with workplace stress. However, being aware of the high cost of such programs in the context of the financial crisis, the present study focuses on several other organizational aspects aiming to reduce employees’ perceptions of job insecurity and stress levels.

33 decision making as two organizational practices that can reduce unpredictability and uncontrollability in the context of general uncertainty. It is true that communication during uncertain times is slow and inaccurate. Being well informed about organizational issues as well as participating in decision making could make employees feel more valued in the organization, enforce their perception of control and thus, reduce job insecurity. Therefore, by communicating several organizational issues such as the hiring and dismissal policies, employees’ sense of controllability and predictability in times of job insecurity could be enhanced (Tilakdharee, 2010). The study showed that work overload is an important work stressor. Overload could imply longer working hours for employees. To reduce the strain created by excessive workloads, management teams could consider “flexitime” schemes which are based on flexible working time arrangements (Harrison et al., 2000). Although these policies have several benefits for both the employees and employers, a lack of efficient monitoring could lead to system abuse. In addition, overload can be efficiently dealt with, with time and priority management strategies. For this reason, e-learning courses that enable employees to make efficient planning of the work, set goals and priorities and develop time management competencies should be available.

Moreover, based on the results of the study, job insecurity and overload diminish absenteeism. While reduced absenteeism could be seen as a positive organizational behavior, the degree to which it restricts proper recovery should be identified by managers and supervisors. Therefore, organizations should reconsider their approaches regarding regular work attendance. Employees who do not take the appropriate sick leave when needed, may perform less well and place additional burdens upon their co-workers (Deery et al., 2014). Overall, reduced absenteeism could be seen as positive only if it is not at the expense of proper recuperation.

34 designed in order to protect employees’ general health from the negative effects of job insecurity.

5.4 Limitations

Several limitations should be taken into account when the present results are interpreted. A first limitation of this study is its cross-sectional design which means that the data were collected at one single point in time. Therefore, causal interferences are impossible to make. It is important to note that the characteristics of the study population are different from the general population. For example, the vast majority of respondents were occupied in permanent positions, the average age was relatively low and the educational status high; it could be the case for example, that younger employees were more willing to reduce their absence rate because of greater career-related concerns (Gosselin et al., 2011). It is important to mention that the current study took place in the surrounding area of Greece, where there are high levels of uncertainty and job insecurity due to crisis.

Moreover, data were collected through self-reports, which are likely to create problems of common-method bias (Podsakoff, et al., 2012). Especially absenteeism and presenteeism which were measured with two and one items respectively could be affected by recall bias. The development of sophisticated, more detailed and, if possible, more objective measures of presenteeism is strongly encouraged (Demerouti et al., 2009). Economic dependence was assessed in terms of four factors. Nevertheless, it could be the case that other aspects may consist a financial burden for employees. For example, the provision of elderly care and the number of unemployed children, regardless of their age, could have played a role when evaluating the degree of economic dependence on one’s job . It should be recognized that the possibility to generalize the present findings is limited and that complementary studies of other population segments are necessary to confirm the observed results.

35 5.5 Future directions

This study also offers some suggestions for future research. First, the relation between job insecurity and absenteeism could be further investigated by distinguishing between the cognitive and affective dimension. In addition, to elaborate further this relation, additional control variables such as the management level, organizational tenure, type of industry, could be taken into consideration.

In this study the attention is drawn to the economically based concerns of employees; the financial needs. Since the moderating role of economic dependence was not confirmed, the inclusion of additional variables as potential moderators may be fruitful to better understand the factors that could lead to absenteeism and presenteeism. Pay attitudes and in particular the perception of “pay fairness”, (Shaw & Gupta, 2001) could be used as a potential moderator in the examined model. Furthermore, Johns (2011) argued that the work context (e.g task significance, task interdependence), personal characteristics (e.g. consciousness, internal locus of control, neuroticism) and work experience (equity, family to work conflict, work to family conflict) influence absenteeism and presenteeism and therefore, may offer further explanation in the examined model.

Once the better comprehension of these behaviors is achieved, future research should also focus on the potential outcomes of such behaviors. It is well documented that high absenteeism increases business costs (sick pay, staff replacement) and leads to lower productivity (D’ Souza, 2006). However, interpreting reduced absenteeism as an indicator of better health may be misleading. The question that may arise is whether the reduced absenteeism associated with high unemployment and increased workload does not inhibit the recovery process. In sum, periods of high job insecurity are related to increased workload. Thus, employees are forced to attend work and reduce their absence rate. The outcomes of the above behavior on employees’ well being should provide food for additional research.

5.6 Conclusion

37 6. REFERENCES

Alexanderson, K. 1995. Sickness absence in a Swedish county with reference to

gender, occupation, pregnancy and parenthood. Linkoping: Faculty of Health

Sciences, Linkoping University.

Alexanderson, K. 1998. Sickness absence: a review of performed studies with focused on levels of exposures and theories utilized. Scandinavian Journal of Public Health, 26: 241-249.

Allan, C., Brosnan, P., Horwitz, F., & Walsh, P. 2001. From standard to non-standard employment: labour force change in Australia, New Zealand and South Africa.

International Journal of Manpower, 22(8): 748–763.

Anagnostopoulos, F., & Niakas, N. 2010. Job Burnout, Health-Related Quality of Life, and Sickness Absence in Greek Health Professionals, European Psychologist; 15(2): 132–141.

Aronsson, G., & Gustafsson, K. 2005. Sickness presenteeism: Prevalence, attendance-pressure factors, and an outline of a model for research. Journal of Occupational and

Environmental Medicine, 47: 958–966.

Aronsson, G., Gustafsson, K., & Dallner, M. 2000. Sick but yet at work. An empirical study of sickness presentee­ism. Epidemiol Community Health, 54: 502−509.

Ashford, S. J., Lee, C., & Bobko, P. 1989. Content, causes and consequences of job insecurity: A theory-based measure and substantive test. Academy of Management

Journal, 32: 803–829.

Baker-McClearn, D., Greasley, K., Dale, J., & Griffith, F. 2010. Absence management and presenteeism: The pressures on employees to attend work and the impact of attendance on performance. Human Resource Management Journal, 20: 311–328.

Bartel, A. P. 1979. The migration decision: What role does job mobility play?

American Eco-nomic Review, 69: 775-786.

Bekker, M., Rutte, C., & Van Rijswijkc, K. 2009. Sickness absence: A gender-focused review. Psychology, Health & Medicine, 14(4): 405–418.

Bergstrom, G., Bodin, L., Haghberg, J., Lindh, T., Aronsson, G., & Josephson, M. 2009. Does sickness presentee­ism have an impact on future general health? Int Arch

Occup Environ Health, 82: 1179−1190.

Blekesaune, M. 2012. Job insecurity and sickness absence: Correlations between attrition and absence in 36 occupational groups. Scandinavian Journal of Public

Health, 40: 668–673.

Borg, I. & Elizur, D. 1992. Job insecurity: correlates, moderators and measurement,

International Journal of Manpower, 13: 13–26.

38 Brett, J. F., Cron, W. L., & Slocum, J. W., Jr. 1995. Economic dependency on work: A moderator of the relationship between organizational commitment and performance, Academy of Management Journal, 38: 261–271

Brief, A. P., & Aldag, R. J. 1989. The economic functions of work. In K. M. Rcwland & G. R. Ferris (Eds.), Research in personnel and human resources management, 7: l-23. Greenwich, CT: JAI Press.

Brief, A. P., & Atieh, J. M. 1987. Studying job stress: Are we making mountains out of molehills?. Journal of Occupational Behavior, 8: 115-126.

Burchell, B. 2009. Flexicurity as a moderator of the relationship between job insecurity and psychological well-being, Cambridge Journal of Regions, Economy

and Society, 2: 365–378.

Caverley, N., Cunningham, J. B., & MacGregor, J. N. 2007. Sickness presenteeism, sickness absenteeism, and health following restructuring in a public service organization. Journal of Management Studies, 44: 304–319.

Chirumbolo, A., & Areni, A. 2005. The influence of job insecurity on job performance and absenteeism: the moderating effect of work attitudes. South African

Journal of Industrial Psychology, 31(4): 65–71.

D’ Souza, R., Strazdins, L., Broom, D., Rodgers, B., & Berry, H. 2006. Work demands, job insecurity and sickness absence from work. How productive is the new, flexible labour force? Australian and New Zealand Journal of Public Health, 30(3): 205-212.

De Boer, E. M., Bakker, A. B., Syroit, J. E., & Schaufeli, W. B. 2002. Unfairness at work as a predictor of absenteeism. Journal of Organizational Behavior, 23: 181– 197.

De Cuyper, N., Bernhard-Ottel, C., Berntson, F., Witte, H. D., & Alarco, B. 2008. Employability and employees’ wellbeing: Mediation by job insecurity, Applied

Psychology: An International Review, 57: 488-509.

De Witte, H. 1999. Job insecurity and psychological well-being: Review of the literature and exploration of some unresolved issues. European journal of work and

organizational psychology, 8 (2), 155-177.

De Witte, H. 2000. Labour ethics and job insecurity: measurement and consequences for well-being, satisfaction and labour input, in J.H Buitendach, S. Rothmann & H. Questionnaire in South Africa. South African Journal of Industrial Psychology, 31(4): 7–16.

De Witte, H. 2005. Job insecurity: Review of the international literature on definitions, prevalence, antecedents and consequences. South African Journal of

Industrial Psychology, 31(4): 1-6.

Deery, S., Walsh, J., & Zatzick, C. 2014. A moderated mediation analysis of job demands, presenteeism, and absenteeism. Journal of Occupational and

39 Demerouti, E., Le Blanc, P. M., Bakker, A. B., Schaufeli, W. B., & Hox, J. 2009. Present but sick: A three-wave study of job demands, presenteeism and burnout.

Career Development International, 14: 50–68.

Dew, K., Keefe, V., & Small, K. 2005. Choosing to work when sick: Workplace presenteeism. Social Science & Medicine, 60: 2273–2282.

Dirken, J. M. 1969. Arbeid en stress. Groningen: Wolters Noordhof.

Doran, L., Stone, V., Brief, A. P., & George, J. M. 1991. Behavioral intentions as predictors of job attitudes: The role of economic choice. Journal of Applied

Psychology, 76: 40-46.

Economou, M., Medianos, M., Peppou, L., Patelakis, A., Stefanis, C. 2013. Major depression in the Era of economic crisis: A replication of across-sectional study across Greece. Journal of Affective Disorders, 145: 308–314.

Edwards, J. R., & Lambert, L. S. 2007. Methods for integrating moderation and mediation: A general analytical framework using moderated path analysis.

Psychological Methods, 12(1): 1-22.

Ehrenberg, R. G., & Smith, R. S. 1988. Modern labor economics (3rd ed.). Glenview, IL: Scott, Foresman.

Enzmann, D. 1994. Work- and client-related stressors as predictors of burnout:

Results from a three-wave longitudinal study. 23rd International Congress of

Applied Psychology. Madrid, Spain.

Eriksen, H. R. & Ursin, H. 2002. Sensitization and subjective health complaints.

Scandinavian Journal of Psychology, 43: 189–196.

Eriksen, H. R., Ihlebaek, C., & Ursin, H. 1999. A scoring system for subjective health complaints (SHC). Scandinavian journal of social medicine, 1: 63-72.

European Agency for Safety and Health at Work. 2009. European Risk Observatory Report.(https://osha.europa.eu/en/publications/reports/TE-81-08-478-EN

C_OSH_in_figures_stress_at_work)"

European Commission. (2007) Towards Common Principles of Flexicurity: More and Better Jobs through Flexibility and Security.

European Commission, 2009. Economic crisis in Europe: causes, consequences and responses, European Communities, Luxemburg.

European Foundation for the Improvement of Living and Working Conditions. 2013. Impact of the crisis on working conditions in Europe. (https://hiva.kuleuven.be/resources/pdf/publicaties/R1534_tn1212025s.pdf)"

Eurostat. 2014. Unemployment Statistics.

(http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=une_rt_a&lang=en)