STRESS AT WORK-
A PREDOMINANT PHENOMENON OF THE
MODERN WESTERN CULTURES?
June, 2011
Stephanie Schworm
Student number: 1890891
Pelsterstraat 16B2, 9711 KL Groningen, the Netherlands +31628560425
Stephanie.Schworm@hotmail.com
University of Groningen Faculty of Economics and Business Msc. Human Resource Management
ABSTRACT
INTRODUCTION
“Men for the sake of getting a living forget to live.” Margaret Fuller
Work related stress is a common phenomenon in societies nowadays. It has become a global concern for economies, organisations and individuals (Atkinson, 2000, as cited by Newton and Jimmieson, 2009). Taking the example of Germany, it came forward in research that more than 80% of the German population is suffering under stress, most of which is caused at the workplace (Statistisches Bundesamt (IT1)). Occupational stress has substantial financial consequences or implications for both, economies and organisations, such as lost productivity, stress-related law suits and healthcare expenses (Sulsky & Smith 2005), as well as increased levels of absenteeism, employee turnover and decreased levels of performance (Joiner, 2001). The USA has lost nearly $300 billion a year on stress (American Institute of Stress) and research suggests that stress may cost up to 10% of a country‘s gross domestic product (Midgley 1997 as cited by Newton & Jimmieson, 2009). A United Nations report stated that job stress became the ‗20th
Century Disease‘ and the World Health Organisation labelled it a ‗world -wide epidemic‘ (Kanjia & Chopra, 2009). Work related stress, in terms of the inability to deal with pressure at work, has as most common individual level consequences a low level of job satisfaction and psychological wellbeing (Ganster & Schaubroek, 2004), and ultimately stress can impact the physical health of an individual (Newton & Jimmieson, 2009). Seen the negative consequences of stress on the financial performance of economies, organisations and the well-being of individual‘s, it is worth investigating what other possible causes for work stress there are in addition to the already researched antecedents as for instance high workloads, interpersonal conflicts or major changes in the working conditions (Ganster & Schaubroek, 2004).
self-worth and well-being is as well influenced by national level cultural values because these values serve as criteria for this experience (Erez & Early, 1993). Hence, work environments differ across cultures (Erez, 2010). National culture is still highly relevant for predicting individual‘s behaviours (Ng, Sorensen & Yim, 2009), and shapes the behaviour of its members by providing a dominant logic of action, which is a repertoire of habits, skills and styles (Swindler, 1986 as cited by Stephan and Uhlaner, 2010). Societal members utilise national culture as a frame of reference to understand organisations, the environment and their relationships with each other (Geletkanycz, 1997). Seen that culture influences individual behaviours, habits, skills, styles, understandings as well as the work environment and the fact that stress is an individual experience and stands in close connection to the work environment, the question is raised whether the level of stress varies per cultural setting. Here, the central research question is derived:
Is work related stress influenced by national culture, i.e. can culture be a predictor for stress?
In this paper, it is proposed that work centrality, which is the degree of importance an individual is assigning his or her job at any given point in time, is a key construct to shed light into the question introduced above. Even though much existing literature focuses on the desirability of work centrality as it is related to positive organisational outcomes (Sharabi & Harpaz, 2010), work stress as a negative consequence is meant to be investigated here. High levels of work meaning are often accompanied by personal costs in terms of higher work stress (Kuchinke, Cornachione, Young & Kang, 2010). Therefore, it is believed here that the more work centrality an individual experiences, the more likely it is that feelings of stress occur. At first sight, it seems paradox and contradicting that high levels of work centrality are connected to better organisational outcomes (Sharabi & Harpaz, 2010) and at the same time are likely to cause higher levels of stress (Kuchinke et al., 2010), which have a negative effect on, among others, performance (Ganster & Schaubroek, 2004). It is believed that commitment plays a crucial role here, as higher levels of work centrality are accompanied by higher levels of commitment (Sharabi & Harpaz, 2010), which in turn makes employees direct their efforts towards important tasks when under stress, therewith positively affecting performance (Hunter & Thatcher, 2007). Moreover, commitment protects an individual from stress and its consequences (Leong, Furnham & Cooper, 1996). The following sub-question is derived from these propositions:
It is furthermore expected that values in the work environment regarding work are congruent with broader cultural values and affect employees‘ understanding of work (Newman & Nollen, 1996). Thus, also the centrality of work is believed to be anchored in broader cultural values. This proposition leads to the next sub-question:
b) Can work centrality explain the possible relationship between national culture and stress?
This paper contributes to the existing body of research in twofold. First, a broad body of research focuses on antecedents and consequences of job stress. Nevertheless, the role of work centrality as an antecedent of stress is not widely studied, as the benefits of work centrality were emphasised more in past research (Sharabi & Harpaz, 2010).
Second, the role of culture in this relationship adds an additional dimension to the research and offers the opportunity to analyse whether the level of work centrality and its effects on stress changes from country to country. Most cross-national research on work centrality compares individual level predictors of work centrality in two to five countries (Parbotheea & Cullen, 2003), and the conceptual development of theories of meaning of work highly focused on individualistic value orientations, creating a severe need for a development and empirical testing in collectivistic cultures (Ardichvili & Kuchinke, 2009). This study is providing an empirical investigation whether cultural values, additional to individualism and collectivism, generally can play the roles of predictors of work centrality.
The practical relevance of this research is especially to be found for multinational corporations, which often find the application of one universal set of management principles to employees of different nations and cultures inappropriate (Ng et al., 2009) as research has shown that congruence between management practices and characteristics of national culture foster better a performance (Newman & Nollen, 1996). It came forward in research that effective management of occupational stress has become increasingly important to human resource practitioners, as the promotion of physical and psychological well-being as well as positive job related-attitudes and performance of employees has been gaining relevance in the field (e.g., Teo and Waters 2002; Hang-yue, Foley and Loi 2005; Quick, Macik- Frey and Cooper 2007). Effective management practices to tackle the problem of stress might differ per cultural setting, and multinational corporations might need to adapt their management practices accordingly.
elaboration on the relationships between work centrality, stress, organisational commitment and national culture. The ISSP data set ‗Work Orientation III‘ is used to test these relationships empirically, on the outcomes of which it will be reflected in the discussion section of this paper.
THEORETICAL FRAMEWORK 1. Definition of the constructs
Work centrality
Work Centrality is one dimension of the concept ‗Meaning of Work‘, which is used to measure the meaning of work with regard to the desired working conditions, work outcomes, work role identifications, social norms about working and work centrality (Kuchinke, Cornachione, Young & Kang, 2009). The concept work centrality refers to the degree of importance that an individual gives his or her job at any given time (MOW 1987, as cited by Sharabi & Harpaz, 2010). It is the belief of an individual concerning the extent to which work plays a crucial role in his or her life (Paulley, Alliger & Stone-Romero, 1994). Individuals who strongly identify with their work role place large value to their job and see it as a central part of their lives (Hirschfeld & Feild, 2000). Work centrality is composed of two approaches, namely i) the relative work centrality that compares the importance of work to the relative importance of other major life areas such as family or leisure, and ii) the absolute work centrality that evaluates the importance of work to the individual without using other comparative standards (Sharabi & Harpaz, 2010).
Stress
decision making latitude forms the potential strain that causes stress. The decision-making latitude of the individual controls to which extent the individual is constrained in its response to the work demands and thus regulates the extent to which the individual experiences stress (Karasek, 1979). The term ‗stress‘ is used hereafter to substitute terms as ‗work related stress or ‗occupational stress‘ since their meanings are in line with each other.
Commitment
Commitment has been defined in several different ways, whereby the underlying meaning is shared in all those ways: Organisational commitment is considered to be a bond or linking of the individual to the organisation (Leong et al. 1996). For instance, Klein Molloy & Cooper (year unknown) refer to commitment in work settings as a perceived bond of a psychological nature that employees have with their job or often a social entity (as cited by Johnson, Chang & Yang, 2010). Commitment is ―the relative strength of an individual‘s identification with and involvement in a particular organisation‖ (Porter, Steers, Mowday & Boulian, 1974, p. 604). It reflects three factors, namely (a) the identification of an employee with the organisation‘s goals and values, (b) their willingness to make a great effort on behalf of the organisation and (c) their intention and desire to stay with the organisation (Porter et al., 1974).
Culture
Culture is defined as ―the collective programming of the mind which distinguishes the members of one human group from another‖ (Hofstede, 1980, p. 25). Culture is a phenomenon at the group, institutional or societal level, but it still is highly relevant for predicting individuals‘ behaviours (Ng et al., 2009). Cohen (2010) claims that cultures share (in imperfect ways) values, beliefs, practices, norms and definitions and that it is dependent on the extent to which a group shares most or all of these elements, that group can said to be a culture. In the same vein, House, Hanges, Javidan, Dorfman and Gupta (2004:15) define culture as ―shared motives, values, beliefs, identities and interpretations or meanings of significant events that result from common experiences of members of collectives that are transmitted over generations‖. Hence, culture does not need to be limited to national borders. However, most of the research does differentiate cultures by means of countries, which is also going to be the case in this research.
dimensions and his study (1980, 2001) became a major source of reference of cultural differences around the world, his study is not left without points of criticism, of which two will be outlined in the following. First, the study was conducted over forty years ago, and seen the demographic, economic, political and cultural changes it remains questionable whether the study is still relevant. The phenomenon of cultural convergence has had profound consequences on societies including the development towards lesser cultural differences (Shenkar, 2001). Hofstede reviewed the study in 2001; nevertheless, the data gathered for the first publication of his research remained the same. Second, the study may suffer from sampling problems as i) the sample has only been taken of IBM employees, ii) at the time of the survey mainly male employees were employed, iii) most of the employees were white collar workers, iiii) all employees are working for an American company, experiencing a common corporate culture (Orr & Hauser, 2008) and iii) data has not been collected in socialist countries (Ng, Lee & Soutar, 2007).
Schwartz‘s cultural framework differs from Hofstede‘s. However, Schwartz argues that Hofstede‘s dimensions are included in his value dimensions, which provide significant advantages over the ones of Hofstede. Two of these advantages tackle the above mentioned points of criticism of Hofstede. First, the value dimensions have been tested with more recent data (collected between 1988 and 1992). Second, a matched sample of students and teachers has been used for the analysis, which was obtained from more diverse regions, including socialist countries (Ng et al., 2007).
The GLOBE study by House et al. (2004) is a study of 62 countries with regard to culture, leadership and organisations and was designed to replicate and expand Hofstede‘s work. Building on his dimensions, the GLOBE study encompasses nine dimensions which measure both, actual society practices (―As is‖) and values (―Should be‖) (Venaik & Brewer, 2010). These dimensions are listed and briefly defined below.
“Uncertainty avoidance is the extent to which members of an organisation or society strive to avoid uncertainty by relying on social norms, rituals, and bureaucratic practices. People in high uncertainty avoidance cultures actively seek to decrease the probability of unpredictable future events that could adversely affect the operation of an organisation or society and remedy the success of such adverse effects‖ (House, et al., 2004:11-12).
“In-group collectivism reflects the degree to which individuals express pride, loyalty and cohesiveness in their organisations or families.‖ (House et al., 2004:12)
“Institutional collectivism is the degree to which organisational and societal institutional practices encourage and reward collective distribution of resources and collective action (House et al., 2004: 12)
“Assertiveness orientation is the degree to which individuals in organisations or societies are assertive, confrontational, and aggressive in social relationships‖ (House et al., 2004: 12). “Future orientation is the degree to which individuals in organisations or societies engage in future-oriented behaviours such as planning, investing in the future, and delaying individual or collective gratification.‖ (House et al., 2004:12)
―Humane orientation is the degree to which individuals in organisations or societies encourage and reward individuals for being fair, altruistic, friendly, generous, caring, and kind to others.‖ (House et al., 2004:13)
“Performance orientation is the extent to which an organisation or society encourages and rewards group members for performance improvement and excellence.‖ (House et al., 2004:13)
“Gender egalitarianism is the degree to which a society or organisation minimises gender role differences while promoting gender equality.‖ (House et al., 2004:12)
The fact that those dimensions measure values and society practices is a distinguishing feature, which differentiates this study from Hofstede and Schwartz. It is moreover based on the most recent sample, which has been gathered across three different industries, namely financial services, food processing and telecommunications. The study is not as criticised as much as Hofstede‘s, because there are generally less controversies (Venaik & Brewer, 2010). The nine dimensions identified in the study all show satisfactory agreement indices and internal consistencies and are confirmed by multilevel confirmatory factor analyses (House et al., 2004). According to Stephan and Uhlaner (2010), the nine dimensions are an appropriate measure to analyse national culture, but high intercorrelations among the nine dimensions create a problem of multicollinearity when they are used in the same model. As a solution, Stephan and Uhlaner (2010) proposed to create second- order factors of these dimensions, which provide a more complete representation of culture as they are allowed to be combined in the same analysis. Using seven of the original nine dimensions1, two second-order factors were identified, which are i) Performance Based Culture and ii) Socially Supportive Culture. The reliabilities for these two variables are satisfactory, and moreover provide a trustworthy
1
and convenient analysis of culture (Stephan & Uhlaner, 2010), which is why for the purpose of this paper these two second-order factors are used to measure culture. The factors are based on the society practice, and not values. This provides a more accurate measure of culture because it reflects the actual behaviours and actions of individuals within their cultures rather than an impressed value framework of a culture. Below, both second-order factors are briefly described.
Performance Based Culture (PBC)
A PBC is characterised as a culture that rewards individual accomplishments over collective memberships in e.g. a family, and that views future-oriented planning as a key manner to achieve high performance. The variable consists of high positive loadings of future orientation, uncertainty avoidance and performance orientation and negative loadings of in-group collectivism and power distance. Humane orientation shows low positive loadings, whereas assertiveness shows higher positive loadings. Even though it seems bizarre that high uncertainty avoidance belongs to a performance based culture, this is consistent with the content of a PBC, as predictability is necessary to engage in future- oriented behaviours such as planning. Countries scoring high on the PBC belong to the Anglo, Germanic Europe and Nordic Europe clusters2. Countries in the Eastern European, Latin European and Latin American cluster score the lowest, whereas societies in the Confucian and Southern Asian clusters are showing medium scores (Stephan & Uhlaner, 2010).
Socially Supportive Culture (SCC)
The SSC reflects a culture in which there is a positive social climate where people support each other. The high loadings of two of the seven dimensions are in line with this, namely the high positive loading of humane orientation and the high negative loading of assertiveness. This means, that a SSC based society is generally very sensitive, friendly and generous with each other and tolerate mistakes, but are less dominant, assertive and tough (House et al., 2004). A SSC can also be a measurement of social capital as a descriptive norm which encourages supportiveness and helpfulness repeatedly. Countries scoring high on SSC are belonging to the Southern and Confucian Asian as well as to Anglo and Nordic Europe. The lowest scores are exhibited by societies belonging to the Germanic, Eastern and Latin Europe cluster, whereas countries belonging to the Latin American cluster exhibit medium scores (Stephan & Uhlaner, 2010).
2
The seven GLOBE dimensions of which these two second-order factors are constituted are uncertainty avoidance, power distance, in-group collectivism, assertiveness orientation, future orientation, humane orientation and performance orientation.
The practice values of in-group collectivism, power distance, and assertiveness are significantly correlating with Hofstede‘s individualism, power distance and masculinity (Ng et al., 2009). Therefore, findings on these dimensions of Hofstede are going to be alienated in the subsequent literature review to the dimensions of GLOBE.
2. Conceptual background and hypotheses
The effects of Work centrality on stress
The effects or consequences of work centrality on individuals in organisations are a double edged sword, i.e. there were positive as well as negative effects found (Sharabi & Harpaz, 2010). One of these negative effects is work stress (Kuchinke et al., 2010). The meaning of work, and work centrality being part of that, is accompanied by individuals investing a great amount of time and effort in their work. Individuals who value work are likely to have less time and energy to get involved in the non-work related part of their lives. Moreover, high work meaning often leads to increased engagement with work and added demands, which may increase the experienced stress (Kuchinke et al., 2010). This is in line with research which has shown a positive relation between work centrality and working hours, whereby working hours, in turn, are positively related to stress. Individuals to whom work plays a central role may be more likely to work longer hours and to continue work even whilst unwell (Ala Mursula, Vahtera, Kevin & Pentti, 2002). Employees who strongly identify with their work are willing to invest more hours into their occupation, as they are more likely to go the extra mile to accomplish their work tasks thoroughly and on time since this fulfils their role expectations. Stress, in turn, may be caused by the depletion of an individual‘s energy through long working hours (Ng & Feldman, 2008).
H 1: Work centrality has a positive relationship with work stress.
Commitment as a moderator of the relationship between work centrality and stress
Schmidt and Lee (2008) see commitment as an overarching construct that includes, among others, work centrality and conceptualise it as one‘s general perspective on work and life, which remains free of deep influences by specific organisational and occupational experiences. In this vein, commitment thus has a strong link to work centrality and, following the aforementioned line of reasoning, both variables show a positive interaction and are even considered to be strongly interwoven. People who are highly committed to their work not only identify with their work role (Hirschfeld & Field, 2000), but they invest mental and effective energy. The work centred individual will act in a way that fosters positive evaluations to enhance one‘s self esteem. Good wages and performance serve to this effect and organisational commitment strengthens the connection to this role (Mannheim, Baruch & Tal, 1997). Organisational commitment was found to buffer the effects of stress in two empirical studies (Begley & Czajka, 1993, Siu, 2002). Possibly, the enhanced self-esteem and a good performance contribute to this buffering effect, together with the note that commitment provides employees with a sense of stability, security and belonging, making them less vulnerable to any kind of work stressors. In the same vein, it is suggested that individuals with a lower degree of commitment are less able to resist the effects of stressful events in their environment (Schmidt, 2007). This leads to hypothesis 2.
H2: Commitment moderates the relationship between work centrality and stress. When levels of commitment are high, the relationship will be weakened, whereas low levels of commitment will strengthen the relationship.
The effects of Culture on Work Centrality
framework for working in Korea, emphasising mental discipline and harmonious behaviours (Kuchinke, 2009). With the industrialisation, work more and more lost its religious aura, changing for the preference of leisure and enjoyment (Kim & lee, 1978, as cited by Kuchinke, 2009). Therefore, it cannot be denied that culture does have an effect on the prevalent degree of work centrality in a country.
A change in economic conditions may have some impact on the rank of order of work goals (England, 1991). Russia for instance experienced a significant increase in work centrality during the transition period in the 1990s, which were characterised by economic hardship and social instability. In general, work centrality in Russia decreased in economically stable and prospering situations and increased in times of hardship (Ardichvili & Kuchinke, 2009) - this pattern was also found in Germany (Borchert & Landherr, year unknown, as cited by Ardichvili & Kuchinke, 2009). Concluding, work centrality is highly dependent and fluctuates according to culture, the socio-economic environment as well as economic stability.
Nevertheless, work in any society is central to human existence, as it provides the necessities of life as well as sources of identity, opportunities for achievement, the standing within the society and community and social gratitude (Ardichvili & Kuchine, 2009). Attitudes and values towards work are one of the most central a person may hold (Harpaz & Fu, 2002), and the notion that work plays a central role for the majority of individuals of industrialised countries has been confirmed by several scholars (Brief & Nord, 1990, as cited by Harpaz & Fu, 2002; England & Misumi, 1986; Mannheim, 1993). Even though there is no empirical evidence stating that this notion holds as well for less developed countries, the assumption is made that this is the case due to the existential need work satisfies and due to the fostered wellbeing through employment.
stronger work centrality can be observed in non Protestant countries. Examples of this are Confucianism and Buddhism, which are predominant in Asian countries. Confucian values entail self- control, duty and conformity which strongly relate to the values of a PWE. Buddha‘s work ethic encourages workers to put forward their best efforts and condemns laziness. In this vein, it seems that Weber‘s beliefs of asceticism, hard work, success and anti-leisure (Ghorpade et al., 2006) are important tenets of other philosophies too. There even seems to be a shift in the historically protestant countries away from this commitment to hard work and asceticism towards more valuing leisure. According to the scheme of Inglehardt & Baker (2000) historically protestant countries are representing countries which are highly performance based (see appendix 2). Empirical evidence exists that this shift away from a strong employment commitment has happened, as historically protestant countries show less strong commitment to employment than historically catholic and communist countries (Warr, 2008), whereby historically catholic countries include mainly countries which are medium socially supportive. An explanation for this might be that historically protestant countries emphasise development and self expression highly, whereas historically catholic countries place a little more emphasis on obtaining the basic conditions of life (Warr, 2008). Also Southern Asian countries and countries anchored in Confucianism, which are highly socially supportive, centre more around the latter (Inglehardt & Baker, 2000). It can thus be argued, that levels of work centrality are higher in SSCs than in PBCs.
Supporting this is the supposition that work ethics, in terms of a commitment to hard work, in developing countries is stronger than in the developed world (Niles, 1994). It is therefore assumed that generally, individuals living in developing countries hold more positive norms and attitudes towards work, as it provides for their daily living in an environment that is often not characterised by social welfare systems that provide for the unemployed. Since, generally, societies that hold positive norms and attitudes towards work experience work as very central and highly cherish it (Harpaz & Fu, 2002), individuals in developing countries might generally experience higher levels of work centrality. Developing countries are represented in greater number in SSC cultures, therewith suggesting that the level of work centrality in these cultures is slightly higher than in PBCs.
H3a: In a PBC, the level of work centrality is high, indicating a positive relationship.
H3b: In a SSC, the level of work centrality is higher than in a PBC, indicating a stronger positive relationship.
The effects of culture on Stress
To begin with, stress and its connection to culture can be examined from an evolutionary perspective. While survival is the primary goal of all species, including humans, stress is the feeling that occurs when an event stands in the way of survival. In ancient times, survival was determined by hunting, collecting and mating, and during the evolution survival became a matter of generating income, achieving position, status and job security. These are properties that ensure progeny and their protection, so stress is to be found in circumstances that hinder the attainment and maintenance of these goals and therewith hinder the continuity of mankind. Even though goals and the connected differing stressful events have changed, this common goal builds the connection across cultures and time. Cultural frames or references create a context in which stress and coping responses are defined and delineated and determine the individual experience of physical and mental functioning (Hobfoll, 1998). Therefore, stress can be experienced to differing degrees, depending on the prevalent culture.
The precise effects of national culture on stress have not been researched yet as such, and the body of literature serving as a basis to make assumptions about the relationship is scarce. Nevertheless, evidence exists that prompts the assumption that, unlike common sense may suggest as stress is a concept of greatest interest to Western society (Hobfoll, 1998), PBCs are negatively related to stress while SSCs are positively related to stress. To recall, the PBC has high positive loadings of future orientation, uncertainty avoidance and performance orientation and negative loadings of in-group collectivism and power distance. An SSC shows a high positive loading of humane orientation and a high negative loading of assertiveness, with significantly lower loadings of uncertainty avoidance, future orientation and performance orientation than a PBC. Moreover, the loadings of in-group collectivism and power distance are higher than in a PBC (Stephan & Uhlaner, 2010). In the following, the body of evidence indicating the proposed relationships is going to be presented.
Performance orientation and in-group collectivism
be demonstrated in an individualistic country as the USA, whereas this orientation would be minimally demonstrated in a collectivistic country as China. The empirical testing revealed the opposite: Individuals in China showed a higher performance orientation than in the USA. This evidence is by no means purposed to challenge the high loadings of performance orientation in a PBC, but is meant to underline that also in non-individualistic countries, people may feel pressured and the need to perform well to fulfil the group expectations and contribute to its success (Ng et al., 2009). Therefore, individuals in such cultures might feel more stressed than individuals from cultures that are low in in-group collectivism. Contributing to this fact is that in low in-group collectivism countries, one acts according to one‘s own discretion and preferences (Triandis, 2004), and decides or chooses him/herself how much time and energy needs to be devoted to the job and how much additional effort one is willing to make to ensure the work is done well (Ng et al., 2009). Concluding, it is suggested, that work environments in countries which are high in in-group collectivism stress is more likely to be triggered than in countries which are low in in-group collectivism.
Power distance and in-group collectivism
Power distance
This leads to the third piece of evidence, which refers back to the concept of stress as conceptualised by Karasek (1979), where stress is seen as the result of an interaction between the job environment and the decision making latitude. In high power distance cultures, managers often do not provide job enrichment and empowerment to their employees (Aycan, Kanungo, Mendonca,Yu, Deller, Stahl & Kurshid, 2000), which can be translated into a lower decision making latitude. Framing this into Karasek‘s job strain model, individuals in high power distance cultures are more likely to feel stress when confronted with high job demands, as this combination fosters psychological and physiological strain (Fox, Dwyer, & Ganster, 1993). Empowerment is more common and effective in low power distance countries (Medonca & Kanungo, 1994), thus in the same vein it can be argued that individuals, when faced with high job demands, are able to react to this demands with a higher decision making latitude and thus experience less strain. Concluding, this suggests that individuals living in societies where high power distance is prevalent also in the work environment are more likely to experience stress than individuals living in societies where power distance is low.
Future orientation and uncertainty avoidance
Another set of assumptions can be made with regard to the fairness of the two sets of cultures. It is argued, that a PBC might be a fairer culture, as rules are predominant in this culture and their application is done in a fair manner, which stems from the low power distance in combination with the high uncertainty avoidance. Even though it could be assumed that in high uncertainty avoidance cultures stress is generally handled better as it can be dependent more on superiors than in low uncertainty avoidance cultures, research has shown that this is not the case (Peterson et al., 1995). Additionally, a PBC might also be a more transparent culture to work in due to the high reliance on systematic planning, as the high loading in future orientation suggests. Future orientation is moreover positively related to wellbeing (Shobe & Adams, 2001), whereby wellbeing is an indicator of the absence of strain (Warr, 2008). Consequently, a low future orientation is more likely to increase the experienced stress levels.
Assertiveness
Dastmalchian, 2009), which is in line with the high loadings of humane orientation and low loadings of assertiveness. However, the low loadings of assertiveness might indicate higher levels of stress. Individuals in countries with high levels of assertiveness score high on intrinsic motivation, as they are achievement oriented and success driven (Ng et al., 2009). Intrinsic motivation, in turn, is associated with lower levels of stress (Baker, 2004). Therefore, individuals who show lesser degrees of assertiveness are more likely to be less intrinsically motivated, and in turn more stressed. This is underlined by the fact that, when anger is suppressed to keep group harmony, anxiety or depression might be caused (Liu, Spector & Shi, 2007), whereby depression has a positive relationship with stress (Clay, Anderson & Dixon, 1993).
TABLE 1
Overview of the body of evidence regarding culture and stress
Loading/ Dimension Stress
High Future orientation = Lower stress level High Assertiveness = Lower stress level High Uncertainty avoidance ≠ Higher stress level Low in-group collectivism = Lower stress level Low Power distance = Lower stress level
This body of evidence, which is summarised in table 1, gives rise to the conclusion that stress is not predominant in PBCs, but is in SSCs.
H 4a: A PBC is negatively related to stress H 4b: A SSC is positively related to stress
The hypothesised relationships can be seen below in figure 1.
FIGURE 1 Conceptual Model H3a H3b H1 H4a H4b
Work Centraliy Work stress
METHOD Sample
For the purpose of this study, the data from the 2005 ISSP survey ‗Work Orientation III‘ is going to be analysed. The ISSP is a continuing programme of cross-national collaboration to collect important survey data for social science research on an annual basis, whereby each year a different topic is covered. For each survey, the data is collected by independent institutions in several countries. The dataset used here was collected in thirty-one countries, and in total 43440 individuals answered the questions of the survey. The data providing the measurement for culture by means of the PBC and SSC factors was taken from Stephan & Uhlaner (2010), which compromises data on sixty-two countries. This measurement is based on the response bias corrected practice value data collected by House et al. (2004) in the GLOBE study.
These two datasets were merged according to countries for the purpose of this study, whereby the country level data was disaggregated to the individual level. In the dataset by Stephan & Uhlaner (2010), South Africa and Switzerland were represented by means of a black and white, and a French and German speaking sample, respectively, whereas in the ISSP these countries were included as a single entity. Therefore, before merging the dataset, the weighted average of both samples in South Africa and Switzerland were taken (Statistics South Africa, 2010; Federal Statistical Office Switzerland, 2002). The twenty-four countries which both datasets have in common and are thus subject of investigation in this study are: Australia, Former East and West Germany, Great Britain, USA, Hungary, Iceland, Slovenia, Russia, New Zealand, Canada, Philippines, Israel, Japan, Spain, France, Portugal, Denmark, Switzerland, Finland, Mexico, Taiwan, South Africa and South Korea. The size of the sample compromises 18304 individuals, of which were 47.3% female and 52,7% male, forming a mean age of 41 years. The sample was reduced to employed respondents only, therewith excluding unemployed and self employed respondents from the dataset.
Measurement Instruments
To enable a statistical analysis for the hypotheses presented, the variables ‗stress‘, ‗work centrality‘ and ‗commitment‘ were developed from several items obtained from the questionnaire of the ISSP.
scale reaching from ‗Always‘ to ‗Never‘. The scale was reliable; the alpha coefficient for this variable was 0,63.
Work centrality. The scale of Work centrality was developed in line with the definition of the variable in literature, namely that it is a combination of a measure for absolute and relative work centrality. Representing absolute work centrality, the items ‗V9: A job is just a way of earning money‘ and ‗V10: I would enjoy having a paid job even if I did not need the money‘ were used. The respondents could choose from ‗Strongly agree‘ to Strongly disagree‘ on a 5 point likert scale to react on these statements. To constitute the measure for relative work centrality, the following items were combined: ‗V4: I prefer to spend less/more time in a paid job‘, ‗V5: I prefer to spend less/more time in doing household work‘, V6: I prefer to spend less/more time with the family‘, ‗V7: I prefer to spend less/more time with friends‘, and ‗V8: I prefer to spend less/more time in leisure activities‘. Respondents reacted to these statements on a 5 point likert scale, reaching from ‗Much more time‘ to ‗Much less time‘. In order to make the scale reliable, several items were recoded, namely: V4, V5, V6, V7, V8 and V9. The reliability of the scale was satisfying with an alpha coefficient of 0,63.
Commitment. The scale was as well developed in line with the given definition of commitment, and was therefore constituted of the following three items: ‗V52: I am willing to work harder to help the firm succeed‘, V53: I am proud to be working for this firm‘ and ‗V54: I would turn down another job and more pay to stay with the organisation‘. Respondents were provided a 5 point likert scale reaching from ‗Strongly agree‘ to ‗Strongly disagree‘. The scale was reliable with an alpha coefficient of 0,67.
Bokemeier & Shephard, 1983). Following, the type of occupation is likely to influence the feeling of stress. In order to control for occupation, the isco 88 code used in the dataset was transformed into 11 classes as developed by Golthorpe using dummy variables. This transformation was done by using the Treiman‘s code as provided by Harry Ganzeboom (IT2). Subsequently, these classes were collapsed into a five class system, consisting of White Collar Workers (WSW), the Petit Bourgeoisie (PB), Farm workers (FW), Skilled workers (SW) and Non-skilled workers (NSW). Due to the fact that all respondents are in paid labour, and thus no respondent of the sample belonged to the self employed Petit Bourgeoisie, this variable was left out as a control variable. The class of White Collar Workers was left out of the analysis as the reference category.
Statistical analysis
For a thorough analysis of the data, two statistical tests were run. First, a multiple regression analysis was conducted, compromising several models as hypothesised with stress as a dependent variable. Doing so enabled a closer examination of changes in the independent variables (PBC and SSC), the mediator (work centrality) or the moderator (commitment) that influence the dependent variable (stress). In order to reassure the outcomes, the independent variables, the mediator and the moderator were standardised. The standardised control variables are added to the regression analysis in order to ensure the outcomes are not biased.
RESULTS
The analysis was started by means of testing the correlations with listwise deletion between the standardised variables. Work centrality and stress positively correlate (r= .19; p <.001), whereas work centrality did not significantly correlate with commitment (r= -.01, p = n.s.). Commitment correlates significantly with stress (p= -.10, p <.001). PBC and stress do not correlate significantly (r=.00, p= n.s.), whereas SSC does positively correlate with stress (r= .08, p <.001). Both PBC (r= -.08, p <.001) and SSC (r= -.04, p <.001) show significant correlations with work centrality, even though the correlation coefficients are rather small. Table 2 shows a complete overview of all correlations.
As the first step for the multilevel analysis, an ANOVA test with country as a factor was conducted in order to test whether stress, work centrality and commitment levels significantly differ across countries. The test showed highly significant F statistics (p values < .001 for all F ratios), therewith confirming highly significantly differing levels of these variables across countries. This is a prerequisite for accurately interpreting the outcomes of the multilevel analysis. The results of this test can be seen in table 3.
TABLE 3
Results of the ANOVA test
N= 17854 for Stress
TABLE 2
Variable means, standard deviations, and correlations
N= 16517
***
Correlation is significant at p < .001 (2-tailed),
**
Correlation is significant at p<.05 (2-tailed),
*
Correlation is significant at p <.10 (2-tailed)
Variable Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 1 Sex -1.40E-1 1.00 1.00 2 Age -2.53E-1 .71 - .04*** 1.00 3 Highest degree ... .91 .04*** - .11*** 1.00 4 Weekly hours ... .98 - .26*** - .01* - .00 1.00 5 FW3 .01 .10 - .07*** .04*** - .08*** .01 1.00 6 SW4 .28 .45 - .25*** .01 - .32*** .08*** - .07*** 1.00 7 NSW5 .03 .17 - .07*** .06*** - .13*** - .13*** - .02** - .11*** 1.00 8 PBC ... 1.00 .02** .09*** .05*** - .11*** - .07*** - .13*** - .03*** 1.00 9 SSC ... 1.01 .01* .07*** .12*** - .01* .01*** - .03*** .05*** .13*** 1.00
10 Work Centrality -6.49E-2 .92 - .06*** .07*** - .06** - .06*** - .03*** - .07*** .02*** - .08*** - .04*** 1.00
11 Commitment -9.37E-3 .99 .05*** - .10*** - .02*** - .08*** - .03*** .08*** - .02*** - .08*** - .09*** - .01 1.00
Testing the hypotheses
In order to test the hypotheses, first a series of regression analyses were performed. Commitment is established as a moderator when 1) work centrality is significant in predicting stress; 2) commitment is significantly related to stress and 3) the interaction term of work centrality x commitment is significant in predicting stress. Work centrality will be established as a mediator when 1) the regression coefficients of PBC and/or SSC are significant in predicting work centrality; 2) the regression coefficients of PBC and/or SSC are significant in predicting stress; 3) the regression coefficient of PBC and/or SSC is no longer significant in predicting stress after adding work centrality in the equation; and 4) the regression coefficient for work centrality is significant in predicting stress after controlling for PBC and/or SSC (Baron & Kenny, 1986). The results of the analyses are presented in Table 4.
Conducting a multilevel analysis as the second step of analysing the data, however, sheds a clearer light into the results. It can be determined whether levels of work centrality and stress indeed differ per country, and whether this is explained by culture. The significance levels of the variables in the analysis as well as the -2 log likelihood provide the necessary criteria. Ultimately, moderation and mediation can only be established if the effects remain significant in the multilevel analysis. The results of this analysis are presented in table 5.
TABLE 4
Results of the Regression Analysis
N= 16700 for predicting Work centrality; Unstandardised coefficients; standard errors are in parentheses N= 16199 for predicting Stress
***
Correlation is significant at p < .001 (2-tailed)
**
Correlation is significant at p<.05 (2-tailed)
*
Correlation is significant at p < .10 (2-tailed)
Predicting Work centrality
Predicting Stress
Variable Model 1 Model 2 Model 1 Model 2 Model 3
Step 1
Constant 2,06*** (.01) 2,07*** (.01) 2,75*** (.01) 2,75*** (.01) 2,75*** (.01) Sex - .23*** (.00) - .02 *** (.00) - .05*** (.01) - .04*** (.01) - .04*** (.01) Age .04*** (.00) .04*** (.00) .06*** (.01) .03*** (.01) .04*** (.01) Highest education level - .01** (.00) - .01** (.00) - .01 (.01) - .00 (.01) - .01 (.01) Weekly hours - .03*** (.00) - .03*** (.00) - .18*** (.01) - .17*** (.01) - .17*** (.01) FW3 - .12*** (.03) - .14*** (.03) - .30*** (.07) - .34*** (.07) - .34*** (.07) SW4 .04*** (.01) .03*** (.01) .04** (.02) .04** (.02) .04** (.02) NSW5 .05** (.02) .04** (.02) .09** (.04) .09 (.05) .05 (.04) Step 2 PBC - .03*** (.00) - .02** (.01) - .02** (.01) SSC - .01*** (.00) .05*** (.01) .05*** (.01) Work Centrality .16*** (.01) .16*** (.01) Commitment - .10*** (.01) - .09*** (.01) Step 3 Work Centrality .16*** (.01)
Work Centrality* Commitment .02** (.01)
F change 46.61*** 70.12*** 101.12*** 197.59*** 4.90**
R2 change .02*** .01*** .04*** .04*** .00**
TABLE 5
Results of the multilevel analysis
Predicting
Work Centrality Predicting Stress
Model 1 Model 1 Model 2
Intercept 2.61*** (.24) 2.32*** (.30) 2.31*** (.30)
sex - .05*** (.00) - .07*** (.01) - .07*** (.01)
age .00*** (.00) .00*** (.00) .00*** (.00)
Highest education level - .01** (.00) - .01 (.01) - .01 (.01) Working hours - .00*** (.00) - .01*** (.00) - .01*** (.00) FW3 - .03 (.03) - .33*** (.07) - .33*** (.07) SW4 .02** (.01) .02 (.02) .02 (.02) NSW5 .05** (.01) .03 (.04) .03 (.04) PBC - .06 (.04) - .04 (.05) - .04 (.05) SSC - .05 (.05) .15** (.06) .15** (.06) Commitment - - .10*** (.01) - .10*** (.01) Work Centrality - .38*** (.02) .38*** (.02) Interaction - - .01* (.01) Number of cases 17042 16517 16517 -2 loglikelihood 13852.20 40517.94 40514.32 Degrees of freedom 12 14 15 Covariance/ Wald statistic Residual .13/ 92,25*** .68/ 90.81*** .68/ 90.81*** Country .01/ 3.37** .01/ 3.15** .01/ 3.15** Standard errors are in parentheses
***
Correlation is significant at p < .001 (2-tailed)
**
Correlation is significant at p <.05 (2-tailed)
*
Correlation is significant at p < .10 (2-tailed)
Hypothesis 1 stated that there is a positive relationship between work centrality and stress.
Multilevel analysis: Also here, the model has a good fit (-2 ll= 40514.32). The relationship between work centrality and stress remains highly significant in the multilevel analysis (e= .38, p < .001), whereby a very strong increase in the strengths of the effect can be observed. This demonstrates that levels of stress vary in countries, which is explained by the levels of work centrality the individuals in these countries are experiencing.
Conclusion: H1 is supported.
Hypothesis 2 predicted that commitment would moderate the relationship between Work centrality and stress.
Multiple regression: Adding commitment in step two when predicting stress proved to be highly significant in Model 2 (∆R² = .04, ∆F = .190.53, p < .001), and commitment is significantly related to stress (b= -.09, p < .001). Moreover, adding the interaction term in step three (work centrality x commitment) is significant in Model 3 as well (∆R² = .00, ∆F = 6.15, p <.05), showing a significant regression coefficient (b= .02, t= 2.48, p <.05). This satisfies the second and third condition for establishing a moderation effect. Therefore, hypothesis 2 is supported.
between work centrality and stress, therewith lowering stress levels in general. Low levels of commitment may strengthen the relationship, therewith increasing stress levels in general. Conclusion: H2 is partially supported.
FIGURE 2
The relationship between Work centrality and Stress under different levels of Commitment
Hypothesis 3a stated that PBC and work centrality are positively related, and hypothesis 3b proposed a stronger positive relationship between SSC and work centrality than the relationship hypothesised in 3a.
Multiple regression: In predicting work centrality, the first step (Model 1) produced highly significant results (∆R² = .02, ∆F = 36.50 , p < .001). Adding PBC and SSC in the next step (Model 2) again show highly significant results (∆R² = .0, ∆F = 20.89, p<.001). Moreover, the regression coefficients are highly significant and indicate an effect of PBC (b= -.03, t= -11.43, p< .001) and SSC (b= -.01, t= -4.57, p< .001) on work centrality. Nevertheless, the unstandardised regression coefficients suggest negative relationships for both, H3a and H3b, therewith not supporting these hypotheses.
Multilevel analysis: The results can also not be supported by the multilevel analysis. The results for PBC (e= -.06, p= n.s.) and SSC (e= 0.5, p= n.s.) both remain insignificant, which means that the country variance in work centrality cannot be explained by culture.
Conclusion: Hypotheses 3a and 3b are not supported
Hypothesis 4a stated that PBC and stress are negatively related, and hypothesis 4b proposed a positive relationship between SSC and stress.
Multiple regression: Predicting stress, adding the control variables in step 1, PBC and SSC in step 2 and work centrality in step 3 showed significant results in Model 1 (∆R² = .04, ∆F = 80.27, p <.001). Model 2 (∆R² = .04, ∆F = .190.53, p <.001) and Model 3 (∆R² = .00, ∆F = 6.15, p < .05). In Model 2 PBC and SSC showed highly significant regression coefficients (PBC: b= -.02, t= -2.96, p <.001; SSC: b= .05, t= 7,58, p <.001). This also counts for Model 3 (PBC: b= -.02, t= -2.84 , p< .001; SSC: b= .05, t= 7,57 ,p< .001), proving a negative and positive relationship with stress, respectively. This supports hypotheses 4a and 4b.
Multi level analysis: Applying the multilevel analysis to the data yielded different results. In predicting stress, the model showed a good fit (-2 ll= 40514.32). However, PBC did not show significant results (e= - .04, p= n.s.), whereas SSC did show significant results whereby the size of the effect tripled (e= .15, p < .05), therewith rejecting hypothesis 4a and supporting hypothesis 4b. It can thus be said that the between country variance in stress cannot be explained by PBCs, but only by SSCs.
Conclusion: H4a is not supported; H4b is supported
DISCUSSION
Findings and theoretical implications
This study is meant as a first exploration into the role of culture as a predictor for occupational stress. Cultural research focusing on the individualism- collectivism continuum with regard to job strain does exist, (e.g. Liu et al., 2007), but a comprehensive study as this one where culture is tested as a predictor of stress is not represented in the body of research. Moreover, cross-cultural research with regard to stress or work values is mainly conducted in a handful of countries (e.g. Liu et al., 2007; Furnham, Pertrides, Tsaousis, Pappas & Garrod, 2005). This study complements this body of research through its complexity with regard to culture and the number of countries.
In the beginning of this paper, three research questions were raised, which are going to be discussed in the following. In line with the structure of the theoretical framework, the sub questions are discussed first, followed by the main research question.
Sub question a: How does work centrality influence stress and which role does commitment play in this relationship?
First, it was hypothesised that work centrality has a positive relationship with stress. It was confirmed that levels of work centrality and stress differ across countries and that higher levels of work centrality generally lead to higher levels of stress at the workplace. This study delivers evidence that supports the notion that work centrality is not only something desirable, but also has negative consequences. Most of the existing research focuses on the attractiveness of work centrality in individuals (Sharabi & Harpaz, 2010), whereas this study reconfirms the unique finding of Kuchinke et al. (2010) that work centrality has stress as a negative outcome.
degrees of investment, but because of the affective bond those individuals have with the organisation, a certain stability, security and sense of belonging that makes them more resistant to any kind of stressors is stimulated (Kobasa, 1982). This study provides some empirical evidence in support for seeing commitment as a buffer of stress rather than a stress enhancing factor.
Sub question b: Can work centrality explain the possible relationship between national culture and stress?
A prerequisite to establish work centrality as a mediator and therewith explain the possible relationship between culture and stress is that culture actually has an influence on the levels of work centrality. This prerequisite is not fulfilled, as both PBCs and SSCs do not have any relationship with work centrality.
One possible reason that these relationships do not hold might be that some scholars argue that the extent to which work is central in one‘s life is a result of certain personality traits rather than of cultural influence. Robust evidence was found that certain personality traits predict job attitudes, choices, values and behaviours. Ultimately, they do have an influence on work values, i.e. people‘s orientation to employment in general (Furnham et al., 2005). Those personal values might overlap with the national belief system (Sidani & Jamali, 2010) and thus culture.
A second explanation for the lacking relationship might be the possible influence of organisational culture. Research brought forward that organisational culture has a significant effect on commitment (van Vianen, 2000). Lok et al. (2005) argue that one unitary organisational culture is widely challenged, and that it is rather a question of perceived subcultures that influence the levels of commitment. In particular supporting subcultures stimulate positive attitudes towards organisational membership, therewith fostering commitment. As explained before, some scholars even see commitment as an overarching construct, which includes, among others, work centrality (Schmidt & Lee, 2008). Seen that the two constructs are highly interwoven, it can be assumed that organisational rather than national culture has an influence on work centrality.
country where working people are generally older and possess higher education levels, the degree of work centrality might differ from a country where the opposite attributes are prevalent.
Central research question: Is work related stress influenced by national culture, i.e. can culture be a predictor for stress?
To answer this question, culture was conceptualised by means of PBCs and SSCs, and their relationship with stress studied. It came forward that SSCs are positively related to stress, whereby PBCs were not significant in predicting stress, despite the evidence of differing stress levels across countries. Due to the SSCs potential to explain these differing stress levels, it can be concluded that culture is a predictor for work related stress, even though this could only be confirmed partially.
What might shed light into why PBCs are insignificant in predicting stress are the perceived work values and levels of wellbeing across cultures. As argued before, the ethics of hard work and asceticism made room for a rising importance of leisure in historically protestant countries, which represent countries in which PBCs are predominant. In turn, engaging in leisure activities is enhancing people‘s feeling of wellbeing and health (Zganec, Merkas & Sverko, 2011) and generally, these countries also have a higher emphasis on personal wellbeing and self development than historically catholic and communist countries, which emphasise acquiring the basic conditions of life more (Warr, 2008). Historically catholic countries represent mainly countries which are medium socially supportive. As Confucian countries, as argued by Niles (1999), have a stronger work ethic than protestant countries, individuals in SSCs might place less value on leisure activities. This gives rise to the assumption, that in PBCs, ‗work to live‘ is the prevalent philosophy, opposing to SSCs who may live according to the ‗live to work‘ principle to a greater extent. Individualism, paired with low power distance, is prevalent in PBCs and a strong predictor of personal wellbeing (Spector et al. 2001, Diener, Diener & Diener, 1995), due to the freedom of choosing one‘s own life course and pursuing one‘s own personal goals. Country clusters representing strong PBCs are all industrially developed nations that report greater levels of wellbeing than less developed nations (Diener et al., 1995). Empirical evidence supports this note with the finding that Anglo and Western European nations had better psychological health and higher job satisfaction than nations in Asia, South America or the Middle East (McCormick & Cooper, 1988, as cited by Spector et al., 2001).
organisations nowadays to invest in work life balance and stress reduction initiatives. Those became very common and central in these cultures. This is for example realised through flexible work schedules, which greatly enhance the level of work life balance an individual is experiencing (Hayman, 2009). A good work life balance, in turn, stimulates a greater job satisfaction (Kanwar, Singh & Kodwani, 2009) and boosts the general wellbeing of an employee. As wellbeing is an indicator of the absence of strain or stress (Spector et al., 2001), this raises reason to believe that the development outlined above could explain the absence of a relationship between PBCs and stress.
Practical implications
The most obvious practical implication of this study is that fostering commitment is highly beneficial for organisations, not only because it is crucial in the attainment of desirable organisational outcomes as decreasing turnover and absenteeism (Griffeth, Hom & Gaertner, 2000) and fosters positive employee attitudes (Peccei, Giangreco & Sebastiano, 2011), but also as it may buffer the effects stress as shown in this study. In the face of organisational stress reduction programmes, fostering commitment is a key manner to succeed in managing stress at the workplace. Moreover, it comes forward that these programmes as well as work-life balance initiatives are not only of high importance in PBCs, but especially in SSCs this should be given increasing attention. Stress should be paid more attention to by organisations, in terms of theory, research as well as practice to raise awareness of the antecedents as well as consequences of high stress levels in SSCs and how they affect organisational outcomes.
Limitations
Although the data is highly generalisable due to the large, diverse sample, this study has its limitations. First, as it is the case with the majority of stress related research, the stress measurement relies on self-report. Those self-report measures on their own might be insufficient, as their validity can be questioned in particular with regard to ―negative affectivity‖. This is a general personality trait that reflects individual differences in negative emotionality and self-concept, in terms of concentrating on negative aspects only and experiencing distress in many situations (Watson & Clark, 1984). Negative affectivity might affect worker‘s perception of their work environment as well as their psychological health and wellbeing (European Agency for Safety and Health at Work, 2000). Hence, the stress measure used in this study might cause methodological problems.
A second limitation connected to this is fact that stress was just measured at one certain moment in time, where a person‘s mood might have strong influence on the judgement of his/ her stress levels. Additionally, it is highly individual at what point individuals start to feel stress. The measure is thus vulnerable to bias and individual differences.
A third limitation lies in the reliability of the variables, as the Cronbach‘s alphas were satisfactory, but not very high. This might have had an influence on the results of this study, and it is desirable to increase the reliability of the study to this regard.
A last limitation of this research is that it does not fully capture the complexity a cross-cultural study brings along. The understanding of the concept of stress differs across countries. Eastern philosophies have viewed stress as an absence of inner peace, whereas Western cultures see stress more as a loss of control (Seaward, 2006). In India for example, a country representing the Southern Asian cluster, there is not even an equivalent word for ‗stress‘ in any of the Indian languages (Laungani, 2002). What might be stressors for individuals in other cultures might not be stressors for Indian citizens due to their unique cultural circumstances as for instance their belief in fate and chance rather than control (Narayanan et al., 1999). Hence, stress is not a unique concept across cultures and the application of our Western concept and understanding in this study might not fully reflect why individuals across cultures experience different stress levels.
Directions for future research
Second, the missing influence of culture on work centrality should be further investigated. Contrary to the expectations, culture did not explain the different levels of work centrality across cultures, which requires a closer investigation of how and why this is the case. As suggested above, organisational culture rather than national culture as well as personality traits and demographics might have an influence instead, but empirical evidence and a further examination of how those affect work centrality is needed.
A third topic that could be studied further is the absence of a relationship between PBCs and stress, whereby differing stress levels across countries can generally be explained by culture. It seems paradox that SSCs, which contain the same cultural dimensions in different loadings as PBCs, can act as a predictor for stress, while PBCs cannot. This gives rise to the assumption that the relationship is more complex than initially assumed, and a thorough research is required on why PBCs do not interact with stress. The suggestion that personal wellbeing might be a determining factor could be a starting point for further investigation. Moreover, the two second-order factors could be broken down into their original dimensions for a more detailed investigation.
Another direction refers to the methodology of this study. As explained in the limitations, the utilised measurement of stress is subject to several criticisms. To avoid the influence of a person‘s negative affectivity and momentum mood on the evaluation of his/her levels of occupational stress, a longtitudinal study could be considered. Moreover, the measure of stress should not only rely on self-reports, but be complemented by more objective measures. Due to the complexity of cross-cultural studies whereby every country‘s unique cultural circumstances need to be considered, it should be investigated with care if the stress scales can be generalised across countries and cultures, as there is always a danger that this might not possible (Narayanan et al., 1999) and the scales need to be adapted.
Conclusion
This study rebuts the common conception that stress is a predominant phenomenon of modern western cultures. Stress is a concept that is indeed mainly conceptualised, researched,
REFERENCES
Ala-Mursula, Vahtera, J., Kivimaki, M., Kevin, M. C. & Pentti, J. (2002). Employee control over working times: associations with subjective health and sickness absences. Journal of Epidemiology and Community Health, 56, 272-278
Ardichvili, A.A. & Kuchine, K.P. (2009). International perspectives on the meanings of work and working: Current research and theory. Advances in Developing Human Resources, 11, 155
Aycan, Z., Kanungo, R.N., Mendonca, M., Yu, K., Deller, J., Stahl, G. & Kurshid, A. (2000). Impact of culture on Human Resource Management practices: A 10-country
comparison. Applied Psychology: An international review, 49 (1), 192-221
Baron, R. M. & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51 (6), 1173-1182
Begley, T. M. & Czajka, J. M. (1993). Panel analysis of the moderating effects of
commitment on job satisfaction, intent to quit, and health following organisational change. Journal of Applied Psychology, 78, 552–556
Bolino, M.C. & Turnely, W.H. (2005). The personal costs of citizenship behavior: The relationship between individual initiative and role overload, job stress, and work– family conflict. Journal of Applied Psychology, 90 (4),740–748
Brooke P.P., Russell D.W. & Price, J.L. (1988). Discriminant validation of measures of job satisfaction, job involvement, and organisational commitment. Journal of Applied Psychology, 73, 139-145
Clay, D.L., Anderson, W.P. & Dixon, W.A. (1993). Relationship between anger expression and stress in predicting depression. Journal of Counseling and Development, 72, 91- 94
Cohen, A.B. (2010). Just how many different forms of culture are there?. American psychologist, January, 59-60
Cooke, R. A. & Rousseau, D. M. (1984). Stress and strain from family roles and work role expectations. Journal of Applied Psychology, 69, 252–260
Diener, E., Diener, M. & Diener, C. (1995). Factors predicting the subjective well-being of nations. Journal of Personality and Social Psychology, 69 (5), 851 -864
