Keywords

Work is a central activity and a principal source of identity for most adults. It is also frequently described as a source of stress, anxiety, and hardship. As such, the relationship between work and mental and emotional well-being is of substantial interest. The effects of work on well-being, however, cannot be effectively understood simply by examining individual experiences in particular jobs. Rather, from a social structural perspective, work-related well-being is substantially influenced by macroeconomic (the way the economy is structured and changes) and labor market (the way jobs and employees are matched) structures that define opportunities for employment in particular kinds of jobs, workers’ positions in social stratification systems that affect labor market positions, and the intersection of work roles and other major roles, especially marital and parental roles. Indeed, the sociological study of work and mental health emphasizes that social and economic structures routinely and normatively affect exposure to work-related stressors and the consequences of that exposure (Fenwick & Tausig, 2007; Tausig & Fenwick, 2011).

Four research foci have addressed the ways in which work and psychological well-being are related, and collectively, they can be linked to provide a social structural explanation for work-related well-being. Most research on work and mental health examines the relationship between job conditions and individual strain or distress. These studies examine how features of jobs – such as the level of job demands, decision latitude, autonomy, substantive complexity, coworker support, and job insecurity – are related to individual levels of strain or distress (Häusser, Mojzisch, Niesle, & Schulz-Hardt, 2010; van der Doef & Maes, 1999). The studies generally do not connect job conditions to larger economic and social conditions, but treat job conditions as stressors and/or sources of work-related support. The job demand/control (support) model (Johnson & Hall, 1988; Karasek, 1979) and the job demands-resources model (Bakker & Demerouti, 2007) that are frequently used to account for individual levels of work stress provide theoretically useful ways for sociologists to understand job-related stress because they can be interpreted as stress-support-distress models.

Second, some studies examine the effects of macroeconomic structures and change on aggregate or individual mental health. These studies generally assess the relationship between aggregate macroeconomic conditions, such as unemployment rates, and aggregate rates of disorder (Brenner, 1973, 1976, 1984, 1995; Brenner & Mooney, 1983), but occasionally link aggregate economic conditions such as changes in unemployment rates (e.g., economic recessions) to individual psychological outcomes (Catalano & Dooley, 1983; Fenwick & Tausig, 1994; Tausig & Fenwick, 1999; Turner, 1995).

In the last three decades, the nature of work has changed substantially as has the relationship between employers and employees. Recent discussions of the restructuring of the employment relationship include consideration of how downsizing, nonstandard work arrangements, labor market segmentation, “new forms of work,” and the proliferation of low-wage jobs – all macroeconomic changes – affect job conditions and well-being. Many of these work-related changes appear to be independent of economic cycles and to represent historic changes in the way in which workers are exposed to and cope with work-related stressors.

A third focus of research is reflected in studies that attempt to explain the relationship between positions in social structures of inequality, work, and well-being. This literature is based on the sociological study of labor markets that is principally used to explain economic outcomes but can be extended to account for psychological outcomes (Fenwick & Tausig, 2007). Social status differences (including gender, race, SES, and citizenship status) affect participation in the labor market and consequent worker exposure to stressful job conditions. The “social status as a fundamental cause of disease” perspective (Link & Phelan, 1995) can be usefully applied to understand the relationship between social status, work, and mental health. Moreover, differences in work-related stress based on social status can be understood as providing a partial explanation for status-based health disparities.

A fourth focus of research examines the intersection of work with the family. This literature has developed, in part, because of increased female participation in the labor force and, in part, because of the more general recognition that the impact of work on mental health cannot be properly understood without accounting for other social contexts (Fenwick & Tausig, 2001, 2004; Schieman, Milkie, & Glavin, 2009; Tausig & Fenwick, 2001).

In this chapter, I summarize what each of these four research areas tell us about the relationship between work and mental health. Each approach reflects subdisciplinary interests, but collectively, they present a broad sociological perspective on the relationship between work and psychological well-being.

The Social Structural Explanation of Job Stress

The juxtaposition of these four research foci yields a social structural explanation of job stress that views stress as a fundamental product of the economic system, labor markets, social structures of inequality, and intersection of social institutions. Macroeconomic structures and change, labor markets, structures of social inequality, and family provide a context for understanding how immediate job conditions affect psychological well-being (see Fig. 21.1).

Fig. 21.1
figure 1

A social structural explanation for job stress

Full size image

This conception, for example, allows us to directly link the broad changes in the nature of work and the relationship between workers and employers that are due to macroeconomic change to increased levels of anxiety (insecurity) as well as economic hardship – forms of psychological well-being/distress (Appelbaum, Bernhardt, & Murnane, 2003; Kalleberg, 2009).

Macroeconomic structures and change also affect well-being in at least two other ways. First, macroeconomic conditions influence the types of jobs (good jobs, bad jobs) that are available in the labor market (Kalleberg, Reskin, & Hudson, 2000), and second, the restructuring of jobs that can be attributed to macroeconomic changes directly affects the nature of job conditions (Cappelli et al., 1997) that are related to job stress.

Social structures of inequality function primarily by defining locations in labor market segments that, in turn, affect experienced job conditions. The US labor market is segmented into standard and nonstandard jobs (with corresponding good and bad job conditions), and women, racial and ethnic minorities, those with low educational attainment, and noncitizens are more likely to compete for nonstandard (bad) jobs (Hudson, 2007; Kalleberg et al., 2000). Hence, positions in status systems of inequality affect exposure to stressful job conditions and subsequent mental health outcomes.

Finally, the family is a particularly salient social institution that intersects work life. In this instance, family obligations affect labor market participation and hence experienced job conditions. Further, work organizations sometimes establish “family-friendly” job conditions that are intended to affect levels of job-related distress. The impact of family life on work-related well-being can also be assessed as a conflicting set of role obligations that cause work-life imbalance or role overload. Of course, family directly affects well-being through nonwork mechanisms as well, but these are not discussed here.

Job Conditions and Distress

What we actually do on our jobs and how we are able to do it have strong effects on well-being. The bulk of research concerning the relationship between work and mental health is focused on specific job conditions and how they affect individuals. In particular, the relationship between the demands of work and the ability to meet those demands is of crucial importance both to well-being and to the development of identity and intellectual flexibility (Karasek, 1979; Karasek & Theorell, 1990; Kohn & Schooler, 1983).

Robert Karasek and his colleagues (Karasek, 1979; Karasek & Theorell, 1990) have outlined a “demand/control” model for explaining worker’s well-being that has received widespread empirical support (Häusser et al., 2010; van der Doef & Maes, 1999). In this schema, the way that a worker can balance work demands with decision latitude (autonomy) in the way work is done is strongly related to worker mental health. The worker who experiences a high level of demands on the job but has little flexibility in the way he/she can meet these demands is at higher risk of developing signs and symptoms of psychological distress. In this model, stress comes from the structured inability of the worker to manage (cope with) high levels of demand. Karasek argued that job demands interact with decision latitude to create job strain but there is also strong evidence that job demands and lack of decision latitude can be regarded as independent stressors (van der Doef & Maes). There is a substantial literature that debates the precise way in which job demands and decision latitude might interact and how that relationship should be modeled, but it is clearer to discuss the research by treating each construct separately. The value of the demand-control model is its emphasis on how job structures affect worker’s well-being.

Job demands are usually indexed by asking workers if they must work very fast on their job, if they have too much work, or if they have enough time to get everything done. Job demands can also be used to indicate if the work is paced by machine and whether it is boring and repetitive. The effects of machine pacing have been of concern for some time. In the stereotypical image of assembly line manufacturing, a worker’s rate of activity is determined by the speed of the assembly line, and the image of the worker falling behind the pace of the machine is a symbol of the stress of manufacturing jobs. Machine pacing has been associated with higher levels of boredom, anxiety, and depression (Caplan, Cobb, French, Van Harrison, & Pinneau, 1975; Hurrell, 1985). In addition, Link, Dohrenwend, and Skodol (1986) have shown that “noisome” physical occupational conditions are linked to psychological disorder.

Decision latitude appears to be the most crucial variable related to work satisfaction and also distress. It is central to the notion of personal control and autonomy. In its simplest version, decision latitude assesses whether the worker has the ability to complete assigned tasks in a way that permits individual preferences to be respected. That is, to what extent can an individual participate in the design and execution of his/her work? Decision latitude is typically measured by questions about a worker’s belief that he/she has the freedom to decide what to do on the job, has a lot of say about what happens on the job, feels that he/she has responsibility to decide how the job gets done, and that the job requires some creativity. Low decision latitude also contains the notion of “closeness of supervision.” Findings suggest that persons who are closely and constantly monitored by their supervisors, who perceive that they are unable to make decisions about their work on their own, and who have no opportunity to disagree with their supervisors will display increased levels of anxiety, low self-confidence, and low job satisfaction (Kohn & Schooler, 1983).

Link, Lennon, and Dohrenwend (1993) have shown that the ability to control the work-related activities of others is also important for well-being. This ability is a job characteristic defined by the Dictionary of Occupational Titles as “direction, control, and planning.” The construct is clearly related to decision latitude, but applies as a description of the job only insofar as the employee has control over other employees.

The “substantive complexity” of jobs is another feature of work that affects psychological well-being. Jobs that require more thinking to complete or which are more complicated to complete are associated with lower rates of anxiety, higher self-esteem, and higher life satisfaction (Caplan et al., 1975; Kohn & Schooler, 1983; Kornhauser, 1965; LaRocco, House, & French, 1980).

Work is also a social setting. Generally, we talk with our coworkers and our supervisors during the day. Often people develop important friendships among coworkers that are carried on after working hours. The opportunity to interact with one’s coworkers fills a general human need for socializing. As well, interactions with coworkers and supervisors offer the possibility of receiving support in times of strain or distress. Jobs that permit workers to interact and to form relationships (e.g., those in which one does not work alone or where the surrounding noise is not too great) also permit workers to obtain support and advice regarding work-related (and, maybe, family-related) problems. Having someone who is trusted to consult about problems is essential to well-being. Thus, opportunities to make friends and to obtain social support from coworkers and supervisors on the job can have a positive effect on well-being (Billings & Moos, 1982; Etzion, 1984; Karasek, Triantis & Chaudhry, 1982; LaRocco et al., 1980; Winnubst, Marcelissen, & Kleber, 1982). Karasek’s (1979) job demands/control model was extended by Johnson and Hall (1988) to include coworker social support as an additional element of the model, and it has been found to interact with demands and control to predict well-being as well as coronary heart disease (de Lange, Taris, Kompier, Houtman, & Bongers, 2003). This latter model is directly compatible with stressor-support-distress models of psychological well-being, although there has been little discussion of the exact ways that these work-related constructs are related to one another (see Lin (1986) for a description of alternate general models of stress, support, and distress relationships).

Although the vast majority of studies relating job conditions to well-being utilize the demand/control (support) argument, an alternative (but not contradictory) argument has recently been advanced, the demand/resource model (Bakker & Demerouti, 2007; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). In this formulation, job demands and resources are conceived more broadly than in the Karasek (1979) demand/control argument. Demands include work pressures, the physical environment, and emotional labor requirements. Resources include pay, promotion opportunities, job security, supervisor and coworker social support, access to information, participation in decision making, skill variety, autonomy, and performance feedback. This model is intended to broaden the range of job conditions that can be related to job stress. The effects of demands and resources can be additive and/or interactive in this model, making it consistent with stress-mediating and stress-buffering models in the stress-illness literature.

Changes in the nature of work and labor markets that will be discussed later in this chapter have increased the salience of job insecurity as it is related to job stress (Kalleberg, 2009). Job insecurity is a job characteristic too, but it differs somewhat from dimensions such as job demands and rewards or decision latitude because it refers to the perceived stability of a job and not to inherent job properties. Job insecurity is defined as “…perceived powerlessness to maintain desired continuity in a threatened job situation” (Greenhalgh & Rosenblatt, 1984). Workers’ beliefs that their jobs will still exist in a year and that they can expect to keep the job if they choose are important to a sense of well-being. Even when economic times are generally good, employees worry about the stability of their employment. When times are bad, fear of unemployment can have severe psychological effects on individuals (Heaney, Israel, & House, 1994; Joelson & Wahlquist, 1987). Job insecurity has become a major feature of the work setting because of the large-scale restructuring of work and its context.

There is strong evidence that employees who regard their current employment as unstable (i.e., insecure) are more likely to experience physical health problems and psychological distress (Burgard, Brand, & House, 2009; de Witte, 1999; Ferrie, Shipley, Stansfeld, & Marmot, 2002; McDonough, 2000; Sverke, Hellgren, & Näswall, 2002). Among other explanations is the hypothesis that during periods of decreased demand for labor (i.e., recessions), workers will feel more vulnerable to layoffs even when they remain employed and, thus, levels of perceived job insecurity (as well as distress) will rise (Catalano, Rook, & Dooley, 1986). The significance of this explanation is that workers need not directly experience unemployment to experience distress (Brenner & Mooney, 1983; Fenwick & Tausig, 1994; Tausig & Fenwick, 1999). Empirically, changes in levels of job insecurity associated with general increases in unemployment have not been shown to be directly related to distress (Tausig & Fenwick). However, there is now evidence that overall job security has declined for other, more systematic reasons and that this decline is associated with elevated distress (Fullerton & Wallace, 2007; Burgard et al., 2009).

Job insecurity is a salient characteristic of “new forms of work” that give workers more autonomy but also make continued employment contingent on their successful contributions to organizational productivity (Cappelli et al., 1997; Fullerton & Wallace, 2007). As well, job insecurity is inherent when organizations attempt to maintain workforce flexibility in a competitive context by using temporary workers or by downsizing.

The globalization of the economy, deregulation of US businesses, technological changes, and worldwide surplus of labor has created a general and enduring “precarity” of employment (Kalleberg, 2009). The growth of precarious work has decreased employees’ attachment to their employers, increased long-term unemployment, and increased perceived job insecurity (Kalleberg). Precarious work leads to insecure workers and to greater distress (Benach, Benavides, Platt, Diez-Roux, & Muntaner, 2000; Burgard et al., 2009; Kivimäki et al., 2003; Quinlan, Mayhew, & Bohle, 2001). This is to say that job insecurity has become a ubiquitous and highly relevant condition of work. Cappelli et al. (1997) suggest that new forms of work have removed the “insulation” from jobs that used to shield workers from the vagaries of the labor market (e.g., through the existence of internal labor markets and a social contract defining employer-employee obligations). The direct exposure to the precarious labor market that results from new forms of work makes job insecurity a significant condition of the job and not solely an individual perception.

Understanding the relationship between job conditions and well-being is useful, but it cannot explain where those job conditions come from. In fact, worker exposure to job stressors and distress is the result of systematic social and economic structures and associated processes.

Macroeconomic Structures, Change, and Distress

The plight of workers in the context of industrial economies has been of concern and interest at least since Engels analyzed the condition of the working class in England in 1844 (Engels, 1958 [1844]). Marxist studies of labor under capitalism show a relationship between this mode of economic production and both societal and individual alienation, and they suggest a direct link between economic organization and well-being (Marx, 1964 [1843–1844]; Mészáros, 1970).

Research by Brenner (1973, 1976, 1984, 1987), Marshall and Funch (1979), and especially Catalano, Dooley, and their associates (Catalano et al., 1986; Catalano & Dooley, 1977, 1979; Catalano, Dooley, & Jackson, 1985; Dooley & Catalano, 1984; Dooley, Catalano, & Rook, 1988) shows that a direct relationship does indeed exist between aggregate indicators of the state of the economy (generally unemployment rates) and aggregate indicators of stress-related poor health (rate of psychiatric hospital admissions, cardiovascular illness, mortality). Changes in unemployment rates (mainly, increases) increase risk of exposure to negative work and financial-related events and reduce social tolerance for deviant behavior. Greater exposure to stressors and reduced tolerance, in turn, lead to higher aggregate rates of morbidity or mortality (Catalano, 1989).

Brenner (1987) suggests that when macroeconomic conditions force a firm to reduce its labor force, remaining employees will experience fear of employment loss and destruction of careers, as well as increased work stress. Starrin, Lunberg, Angelow, and Wall (1989) suggest that fear of unemployment causes employed workers to work harder and that, at least in certain industries, as unemployment rates increase, owners of capital will find it efficient to extract more labor by requiring overtime work from a smaller number of workers instead of obtaining cheaper labor from the growing pool of the unemployed. As job demands and job insecurity are increased, these authors suggest, worker distress increases. Kivimäki, Vahtera, Pentti, Thomson, Griffiths, and Cox (2001) have shown that downsizing results in the restructuring of remaining jobs in such a way that job insecurity is increased, job demands are increased, and decision latitude is reduced. Further, these changes are linked to decreases in perceived health. Although these latter arguments make a case for the existence of direct effects of macroeconomic structure and change on job conditions related to distress, the study of unemployment most clearly illustrates the direct effects of macroeconomic conditions on mental health.

Unemployment

Unemployment is a stressor that clearly leads to greater physical and mental distress (Dooley, Catalano, & Wilson, 1994; Horwitz, 1984; Jahoda, 1988; Liem & Rayman, 1984; Pearlin & Schooler, 1978). Given that we often define ourselves by our job titles, the loss of a job can mean that our identity, based on our employment, is threatened. Unemployment, of course, also has significant financial effects. Most of the research on the effects of unemployment on worker’s well-being focuses on these two matters: threats to identity and financial strain. The typical study of the health effects of unemployment shows that unemployment is related to increases in drinking, more physical illness, higher rates of depression, anxiety, “bad days,” suicidal ideation, and increased use of tranquilizers (Kessler, House, & Turner, 1987). This research also shows that becoming reemployed largely wipes out the effects of not being employed (Kessler, Turner, & House, 1989).

If the health effects of unemployment are not much debated, then the question turns to the causes of unemployment. It is here that we can see some of the ways in which social and economic structures affect worker’s well-being by affecting opportunities for work. The main reason for unemployment is “structural” and involuntary. That is, the economy goes through cycles of growth and decline, and, during decline, jobs are lost simply because employers cannot afford the labor force costs they incurred when times were better or because they close economically marginal plants or relocate production to lower-wage areas. During recessions, the number of unemployed people swells, and prospects for quick reemployment are poor. By definition, involuntary unemployment means that workers do not have control over the basic condition of their access to financial and identity security. There is also evidence that even among workers who do not lose their jobs during recessions, elevated levels of insecurity brought on by concern over the economy increase symptoms of depression and other forms of psychological distress (Heaney et al., 1994; Kuhnert & Vance, 1992).

A number of authors, however, have noted that personal reactions to unemployment can be affected by the aggregate economic context as well. Although Dooley et al. (1988) did not find such a relationship, Perrucci, Perrucci, Targ, and Targ (1988) and Turner (1995) have shown that community-level reactions to plant closures and/or local unemployment rates interact with personal unemployment experiences to affect psychological reactions. For example, Turner found that it is better to lose a job – in terms of less physical and psychological distress associated with unemployment – when the chances for reemployment in the local community are good. Dooley et al. (1994) found that community-level unemployment rates had an indirect effect on individual depression by raising the risk of unemployment.

Studies of the consequences of reemployment show that the distress attributed to unemployment largely disappears (Kessler et al., 1989; Kessler, Turner, & House, 1988; Liem & Liem, 1988; Payne & Jones, 1987; Turner, 1995; Warr & Jackson, 1985). This effect can be attributed to personal job-seeking efforts and the consequent feelings of efficacy as well as improved finances. In some instances, however, reemployment may not improve well-being. Perrucci et al. (1988) report that the well-being of those reemployed following a plant closing was no better than for those who remained unemployed largely because those who became reemployed did so in jobs that offered substantially lower wages and less job security than their previous employment. In this instance, reemployment addressed neither financial nor identity issues.

This latter point has increased relevance as the restructuring of the economy and jobs that has been occurring over the last three decades has created more involuntary part-time employment and increased low-wage jobs (Bernhardt, Morris, Handcock, & Scott, 2001). Dooley, Prause, and Ham-Rowbottom (2000), for example, have shown that underemployment such as occurs with involuntary part-time employment is associated with as much increased depression as unemployment. Virtanen, Liukkonen, Vahtera, Kivimäki, and Koskenvuo (2003) found that contingent workers with uncertain employment contracts experienced higher rates of both physical and mental illness, although these rates were not as high as those among low-income unemployed workers. Both studies of the effects of unemployment and reemployment suggest the importance of accounting for the macroeconomic context in understanding effects on distress. This concern now leads us to examine some recent changes in the basic relationship between workers and employers and the effects these changes may have on employee well-being.

The Changing Nature of Work

There is broad agreement that work organizations and work have undergone considerable restructuring in the past 30 years. The changes have been both radical and widespread (Osterman, 1994, 2000; Vallas, 1999). The impetus for these changes is variously tied to global economic competition, changes in employment law and regulatory and trade policies, the shift away from manufacturing (in the US), technological change (i.e., computerization), and fundamental shifts in the nature of capitalism (Cappelli et al., 1997; Smith, 1997; Vallas, 1999).

The image evoked earlier of the hapless worker whose tasks were tied to the assembly line (high demand, low control equaled job stress) no longer describes the typical full-time, core worker or his/her job in advanced economies. “New forms of work” that are characterized by the recognition of worker knowledge and judgment, the use of teams, and minimal supervision have replaced the “Fordist” model of hierarchy, formalization and supervision (Cappelli et al., 1997; Smith, 1997). And, in principal, this form of work increases work control and decreases job demands (Macky & Boxall, 2008). Work in the “high performance work organization” (HPWO) is one way in which the organization attempts to increase its “functional flexibility” (Smith) relative to work tasks and productivity demands (Kalleberg, 2003). New forms of work describe “core” workers in organizations who work “standard” full-time, Monday to Friday jobs. The very limited empirical examinations of how new forms of work affect worker’s well-being seem to suggest that the effects on job stress vary by industry (Berg & Kalleberg, 2002; Parker, 2003). But these studies also suggest that high performance work practices have either no effect or increase job stress rather than decrease job stress as might be expected from the increase in decision latitude and more interesting work. Cappelli et al., for example, have suggested that new forms of work and work organization contain contradictions that can potentially create job stress. New forms of work often demand substantially more from the worker. Different skills such as those related to interpersonal relationships (team play) and logistics may be called for. Workers may find that the level of job demands has increased dramatically. Workers may also discover that the greater autonomy promised by the reorganization of work is illusory or offset by normative processes within work groups (Barker, 1993). Smith reports that research studies do not generally find that true decentralization of authority occurs in redesigned work. In this case, meaningful decision latitude may not increase. Indeed, Parker found that lean production systems increased job depression due to perceived decreases in job autonomy, skill utilization, and participation in decision making. Landsbergis, Cahill, and Schnall (1999) reviewed studies of the impact of lean production systems in the auto industry and concluded that lean production intensified job demands and that decision latitude did not increase.

The Changing Labor Market

Employers have effectively restructured their workforce into a standard, permanent “core” set of workers (increasingly organized under high performance work practices discussed above) and nonpermanent, nonstandard “peripheral” workers. This organization of work and workers creates the “numerically flexible” firm (Smith, 1997) that is intended to give employers the ability to compete in global markets and to maintain profits by quickly increasing or decreasing its workforce as conditions dictate (Kalleberg, 2003). Kalleberg et al. (2000) and Hudson (2007) have shown that the US labor market is now segmented into a core segment characterized by “standard,” full-time work with good pay and benefits, “good jobs,” and a peripheral segment characterized by nonstandard, part-time, and contingent labor with low pay and few or no benefits, “bad jobs.”

Sizable numbers of those employed in the US now work in jobs that are intentionally structured to last a limited period of time or to provide limited hours of work (i.e., temporary, contingent, or part-time). In 2005, the Bureau of Labor Statistics (BLS) estimated that up to four point 1% of the current labor force was working in jobs that meet various definitions of contingent employment (BLS, 2005). Adding those who are self-employed and those who are employed part-time brings the total closer to one in three workers (Parker, 1993). For large numbers of workers, employment is not permanent, income is not predictable, and traditional employee benefits such as retirement and health insurance are highly uncertain (Kalleberg et al., 2000). Moreover, the prospect of “downsizing” hangs over many permanently employed workers who no longer regard any job as permanent even as their own employment continues. In short, many persons in the labor force are likely to feel insecure about their jobs, and many will feel that they have little control over the conditions of their employment.

Further, it is important to note that these changes in the distribution of permanent and nonpermanent jobs are occurring independently of economic cycles. While jobs are lost during recessions as organizations cope with the poor economic climate, downsizing and the expansion of temporary and contingent jobs is an intentional (and permanent?) feature of work in the US and other advanced economies (Kalleberg, 2009). Moreover, high-paying and mid-level jobs are increasingly being replaced by low-wage jobs that contain both economic and psychological stressors (Appelbaum et al., 2003; Bernhardt et al., 2001; Luo, 2010).

Contingent workers (other than independent contractors and the self-employed) share a number of characteristics. Their wages are typically lower than permanent workers; they receive few, if any, fringe benefits; they have few opportunities for career advancement; and they have no chance to exert control over the conditions of their work. These are characteristics which may increase job-related stress among contingent workers. Kivimäki et al. (2003) found that temporary employment is associated with higher mortality than permanent employment. Virtanen et al. (2005) reviewed 27 studies of the health effects of temporary employment and concluded that there is an association between temporary employment and increased psychological morbidity. Parker, Griffin, Sprigg, and Wall (2002), however, reported that while temporary status and lower participation in decision making that accompanies temporary job status increases strain, this effect is offset by lower levels of job demands so that the net effect of temporary employment was to reduce job strain.

Not all nonstandard employment is identical in form or in its implications for personal sense of control. For example, part-time work represents the largest category of nonstandard work, but much of it is voluntary on the part of workers. Negrey (1993) concluded that voluntary part-time employment may enhance worker sense of control by permitting scheduling and participation in other social activities. Tilly (1991), on the other hand, found that most of the increase in part-time employment since 1970 is among “involuntary” part-timers, workers who prefer full-time employment but cannot find it. Dooley et al. (2000) found that among involuntary part-time workers, depression levels were as high as among unemployed workers. Further, the Bureau of Labor Statistics (BLS, 2005) reported that 54% of contingent workers would prefer to have a permanent job.

Fenwick and Tausig (2001, 2004) concluded that schedule control was a key determinant of health outcomes among nonstandard workers net of other job characteristics. Kalleberg (2003) concluded that the degree to which a worker in a nonstandard employment context can exercise control over his/her skills determines the benefit derived from nonstandard employment. Virtanen et al. (2003, 2005), and Saloniemi, Virtanen, and Vahtera (2004) found that fixed term employment was not associated with elevated distress but, rather, only non-fixed term contingent work (where the worker had no control over length of employment) was related to distress.

There is also a small set of studies that examine the effects of contingent workers on permanent workers in the same firm. These studies suggest that the presence of certain types of contingent workers may increase insecurity among permanent workers and decrease job satisfaction, loyalty, and attachment to the firm (Chen & Brudney, 2009; Davis-Blake, Broschak, & George, 2003; de Cuyper, Sora, de Witte, Caballer, & Peiró, 2009; George, 2003). These studies, however, do not explicitly examine stress or mental-health-related outcomes.

In summary, the macroeconomy and changes in the macroeconomy can directly affect mental health through changes in unemployment levels, both personal and aggregate. Long-term changes in the context (precarious employment) and the nature of work (high performance work practices) and its organization (standard, core jobs vs. nonstandard, contingent jobs) that are due to macroeconomic factors also affect mental health indirectly through changes in the structure of the labor market and the stressful qualities of restructured jobs.

Social Stratification and Job Stress

If the labor market is now segmented into good jobs and bad jobs based on standard versus nonstandard employment, workers have different “risks” of being found in each segment, in part, based on social status (Hudson, 2007). In turn, workers employed in different segments of the labor market have different risks for ill-health (Virtanen et al., 2003).

Women, those with high school educations or less, racial/ethnic minorities, and noncitizens are more likely to be found in nonstandard (bad) jobs. Hence, these groups are also more likely to be exposed to the stressful elements of work – particularly low wages, absence of benefits, insecurity, and low decision latitude. To put this in another way, social stratification affects exposure to stressful job conditions and may be regarded as one mechanism that links work-related distress to the observed social gradient in health (Marmot, Bosma, Hemingway, Brunner, & Stansfeld, 1997; Marmot, Ryff, Bumpass, Shipley, & Marks, 1997; Warren, Hoonakker, Carayon, & Brand, 2004). Indeed, it is possible to suggest that some health disparities attributed to structures of inequality occur because of the differences in risk exposure to work-related stressors that follow from differences in labor market positions (see Fig. 21.1). Not only are jobs in the peripheral segment of the labor market less secure and apt to provide less decision latitude, but they are also low wage and rarely include health insurance benefits leading also to differences in health-care access and health outcomes. This account is completely compatible with the social status as fundamental cause of illness argument (Link & Phelan, 1995) and is seen as increasingly relevant for explaining the social gradient in health generally (Clougherty, Souza, & Cullen, 2010).

Women, African Americans, part-time workers, and those with less than a high school diploma (and those with advanced degrees) are more likely to be employed in contingent jobs (Hipple, 2001). White women make only 81.1% of the salary of their male counterparts, Asian women 75.6% of what Asian men earn, Hispanic women 89.9% of what Hispanic men earn, and African American women make 96.2% of the salary of their male counterparts (BLS, 2011b). The median earnings for African American men are only 73.4% of the median for white men; median earnings of Hispanics were lower than those of African Americans, whites, and Asians; and persons with low educational attainment earn from 38% to 56% of the median weekly earnings of college graduates (BLS). Foreign-born men earn 70% as much as native-born men, and foreign-born women earn 80% as much as native-born women. At all education levels, the median weekly earnings of foreign-born workers who work full-time were less than those of their native-born counterparts in 2004 (Mosisa, 2006). Foreign-born workers and especially non-US citizens are more likely to be employed in contingent, time-limited jobs (Hipple). These data, then, suggest that lower status workers are more likely to be exposed to economic and insecurity-related stressors as well as the stressors associated with nonstandard work regardless of specific job characteristics.

It is also worth noting that there is another indirect relationship between structures of inequality and well-being through the labor market. African Americans, Hispanics, and those with no college education (or less than a high school degree) have higher unemployment rates (BLS, 2011a) and are, thus, more exposed to the negative emotional consequences of unemployment as well.

Gender

Men and women still work in very different jobs. Today, those differences are captured by the differences in allocation by gender into standard and nonstandard jobs that, in turn, affect specific job characteristics. These differences have well-documented effects on job outcomes ranging from differences in income (Blau & Beller, 1988) and authority (Wright, Baxter, & Birkelund, 1995) to distress (Barnett & Marshall, 1991).

Compared to men, women constitute one group of workers whose employment is typified by job characteristics that have been found to be stressful. Women’s work is concentrated in low-paying occupations, smaller organizations, and peripheral, nonunionized industries (Beck, Horan, & Tolbert, 1978; England & McCreary, 1987; Gabriel & Schmitz, 2007). This occupational segregation is also related to characteristics of the jobs that women typically encounter. Women tend to predominate in occupations that are less flexible and that permit less autonomy than those occupied by men – precisely the characteristics related to high levels of job-related distress (Glass, 1990; Hachen, 1988; Rosenfield, 1989; Tomaskovic-Devey, 1993). These stressful job conditions are now associated with forms of nonstandard employment in which women predominate. It has been suggested that part of the persistent occupational segregation observed by gender is related to the preferences of women related to a desire for more flexible work arrangements (scheduling), especially for family-related considerations (Gabriel & Schmitz, 2007). We will take up this question in the next section on work and family.

Women’s job-related distress is, therefore, affected both by the macroeconomic and social conditions that channel women into specific jobs and to the specific job characteristics they encounter within those jobs.

Education

A similar argument can be made for differences in educational status. Just as I suggested that the typical jobs that women hold have characteristics that make them more stressful, the same is true for persons with lower educational attainment (Karasek, 1991; Link et al., 1993). Educational attainment affects job-related distress by sorting workers into jobs with different levels of stressful characteristics. Low educational attainment, for example, is one of the personal characteristics that make it more likely that a worker will compete for nonstandard jobs (Hudson, 2007). Karasek also found that while not all occupations that could be characterized as either “blue collar” or “white collar” contained the same basic job characteristics, jobs with high levels of demand and low levels of decision latitude are more prevalent in blue collar occupational categories. Kohn, Naoi, Schoenbach, Schooler, and Slomczynski (1990) suggest that the traditional indicators of SES (education, income, and occupational prestige) are consistently related to distress only for manual workers. They found that manual workers differ from others largely because their jobs lack the dimension of control of one’s own labor. Link et al. found that the crucial characteristic of work which connects socioeconomic status to distress is the extent to which occupations permit workers to control the work of others. They showed that persons in occupations containing the characteristic of direction, control, and planning are less likely to experience depression and that such jobs are linked to higher SES.

Socioeconomic status is related to social class but in a complicated fashion (Kohn et al., 1990). The concept of class distinguishes between those who own the means of production and those who work for owners. Research confirms a relationship (although not a linear relationship) between class-related positions, job characteristics, and distress (Kohn et al., 1990; Tausig & Fenwick, 1993). Tausig and Fenwick showed both that the characteristics of jobs in terms of demands and decision latitude differ by class and that the impact of macroeconomic change on workers is conditioned by class status. Those who work for others and do not supervise others (the proletariat) are more likely to work in jobs with high demands and low decision latitude. During economic downturns, their decision latitude decreases, and their levels of anxiety and depression increase. Owners and supervisors, however, are not immune to the effects of macroeconomic change. The owners of smaller businesses in the peripheral sector of the economy experience increased anxiety following economic downturns, and they also report decreases in decision latitude that affect depression. Depression and anxiety levels for supervisors also increase during recessions because job demands increase and decision latitude decreases. Kohn et al. argued that position in the class structure determines the degree of control one has over the conditions of one’s work, especially regarding occupational self-direction (decision latitude), that is related to psychological functioning.

Race/Ethnicity

Membership in nonwhite racial categories has substantial effects on the likelihood that an individual will be employed in a job that contains stressful characteristics (i.e., a nonstandard job) (Hipple, 2001; Kalleberg et al., 2000; Presser, 2003).

African Americans are more likely to be employed in jobs with nonstandard work characteristics, more likely to do shift work, and more likely to work in contingent, time-limited jobs. They also earn less than whites on average and within identical occupations. African American men earn about 73% of the amounts earned by white men (BLS, 2011b). African American women earn almost as much as white women, but earn substantially less than men. In addition, African Americans are more likely to report experiencing racial discrimination in their jobs, and perceived discrimination is related to well-being (Jackson & Saunders, 2006). Studies show that African Americans have less access to “good,” well-paying jobs that are high in decision latitude and lower in job demands (Tomaskovic-Devey, 1993). Tomaskovic-Devey found that African American employees are more closely supervised and have less complex tasks, less managerial authority, and less supervisory responsibility than whites.

African Americans and whites also have different risks for unemployment. Unemployment rates for African Americans are routinely nearly twice those for white Americans (BLS, 2011a). Whether this is the result of human capital differences or racist employment policies, the experience of unemployment also contributes to observed rates of distress among African Americans.

Presser (2003) has shown that Hispanic workers are also more likely than non-Hispanic whites to work nonstandard work schedules and shifts, and Hipple (2001) found that Hispanics are more likely to work in contingent jobs. Mosisa (2006) found that foreign-born workers (mainly Mexican born) are more likely to be employed in service occupations; natural resources, construction, and maintenance jobs; and in production, transportation, and material moving occupations. As a result, foreign-born workers make only 76% of the wages and salary of native-born workers. These labor-market-related factors are partially explained by the lower average educational attainment of migrants from Mexico, but Portes and Zhou (1993) have shown that second-generation Hispanic workers appear to lack the usual occupational mobility expected among second-generation workers. Relative to whites, even second-generation Mexicans have been found to be working in jobs with low earnings and benefits (Waldinger, Lim, & Cort, 2007). In short, some racial and ethnic minorities are more likely to work in peripheral, nonstandard jobs with known stressful characteristics.

Citizenship

Immigration status is associated with the segmented labor market, and immigrants are far more likely to obtain jobs that are nonstandard (bad) in nature (Hudson, 2007; Kalleberg et al., 2000). Hence, immigrants are more likely to be exposed to stressful job conditions, including insecurity associated with nonstandard employment. According to Hipple (2001), non-US citizens are almost twice as likely to be employed in contingent jobs as US-born workers. Noncitizens are twice as likely to work in agriculture and five times as likely to work in private household services compared to US natives and naturalized citizens. The relevance of citizenship as a predictor of labor market position has increased substantially in the last 20 years or so. It is a primary predictor of labor market status in the peripheral, nonstandard segment of the labor market. I must note, however, there are no studies of work stress that link immigration status (and especially noncitizenship status) to work conditions and mental health. At this time, the relationship is speculative but is presented here because of the relevance of labor market segmentation to job conditions and stress.

Gender, education, race, ethnicity, and citizenship status (all structures of inequality) affect the exposure and vulnerability of workers to distress by affecting the labor market participation of individuals based on these status characteristics. Women, persons with low educational attainment, racial and ethnic minorities, and immigrants make up groups that are matched to nonstandard work arrangements through the labor market. These nonstandard jobs are shown to contain stressful characteristics including low wages, low benefits, low decision latitude, and high insecurity. Indeed, the way in which social status functions as a distal cause of ill-health is likely to include the way in which social status differences structure exposure to stressors (and support) based on those status differences. This would certainly include structured access to jobs with various characteristics. And, in this case, the exposure to stressful job conditions also includes access to health insurance (as an absent benefit) so that a partial explanation for health disparities related to social status would include exposure to more stressors as well as the inability to get care for illness conditions that arise from them.

Work and Family

The increased participation of women in the paid labor force and the psychological effects on women of that participation can be partly understood as a function of the consequences of social structures of inequality on work-related stress and distress (above). But, particularly, because of the increased participation of women in the labor force (but not exclusively so), the intersection of the family as a social institution with work has also been studied as a source of work and/or family-related stress (family-work interference, work-family interference).

Figure 21.1 suggests two pathways by which family intersects with work to affect work-related distress: family considerations affect the participation of family members in the labor market, and family also affects job characteristics. (Figure 21.1 also indicates a direct relationship between family and well-being, but this nonwork-related connection is not discussed in this chapter.) There is also a very sizable research literature on the relationships between work and family that centers on how each social institution creates conflict or interference for the individual family member as worker or worker as family member. This literature shows that work-family interference and family-work interference significantly affect worker’s well-being (Frone, 2000; Greenhaus & Beutell, 1985; Grzywacz & Bass, 2003).

In the section above, I explained that one way that gender inequality affects well-being is through the different labor market positions of men and women and the consequent differences in exposure to stress-related job characteristics that follow. Specifically, women are more likely to work in nonstandard jobs that are part-time, temporary, and/or contingent. Hence, women are more often exposed to work-related stressors such as low decision latitude, high job insecurity, low wages, and absent benefits such as health insurance that are characteristic of nonstandard jobs. I also noted that between 50% and 60% of workers with nonstandard schedules would prefer standard, full-time, and more permanent employment.

But when we look at those workers who prefer working nonstandard work schedules, we find that family-related reasons are often given to explain such preferences (Presser, 1995, 2003). Both men and women (but more often, women) indicate that nonstandard work arrangements are preferable because such employment allows for better child-care arrangements and/or better arrangements for care of other family members. Moreover, for women particularly, the presence of one or more children over the age of five is associated with a greater preference for nonstandard work arrangements. If we view the family and work as “greedy” institutions that both demand participation and time commitments, then it is clear that voluntarily choosing nonstandard work hours is one way to solve this time bind and, therefore, to reduce stressors and ill-health-related outcomes created by the need to meet both work and family demands (Fenwick & Tausig, 2001; Tausig & Fenwick, 2001; Voydanoff, 1988). In short, one way that the family affects work-related stress is that family conditions affect self-selection into the standard work arrangement, primary segment of the labor market with “good” jobs, or into the nonstandard work arrangement, secondary segment of the labor market with “bad” jobs.

The participation of women in the labor force has also directly affected some aspects of work organization and subsequent job conditions. In order to retain permanent workers who have conflicting or demanding family obligations, some firms have introduced “family-friendly” work policies that include flexible work scheduling, provisions for child care, and extended maternity or paternity leave (Berg, Kalleberg, & Appelbaum, 2003; Davis & Kalleberg, 2006; Glass & Fujimoto, 1995; Osterman, 1995). Hammer, Saksvik, Nytrø, Torvatn, and Bayazit (2004) suggest that family-friendly work norms may be regarded as job conditions related to work stress exactly in the sense that job demands, decision latitude, and coworker and supervisor support have been. Family-friendly work policies should reduce work stress and work-family distress.

Glass and Fujimoto (1995) and Anderson, Coffey, and Byerly (2002) argue that family-friendly work policies reflect an effort to counteract absenteeism, turnover, and job dissatisfaction, especially in organizations with a high proportion of female workers (Davis & Kalleberg, 2006). Osterman (1995) and Berg et al. (2003) have observed that family-friendly work organization policies are associated with the presence of high performance work organizations (HPWO) since such policies appear to increase worker commitment to the organization and high levels of involvement that are essential to the success of high performance work organizations.

Family-friendly work policies are thus offered to core employees in HPWOs and particularly to professional and managerial-level employees (Davis & Kalleberg, 2006; Glass & Estes, 1997), and such policies function as a job condition that affects work-related stress and strain (Hammer et al., 2004). Indeed, Thomas and Ganster (1995) have shown that flextime is related to decreased depression and somatic complaints by workers, although a meta-analytic review of family-friendly work environments by Mesmer-Magnus and Viswesvaran (2006) finds that overall family-friendly work environments have few positive effects on worker’s well-being. It is worth noting in this context that family-friendly work policies are generally not available to part-time, temporary, and contingent employees, but, as I noted earlier, such workers may self-select nonstandard work arrangements precisely because these forms of work effectively help manage family-work interference even if it is at the cost of lower wages, job insecurity, and fewer formal benefits.

Thus, one consequence of attempts to balance work and family roles is exposure to labor market conditions that do not favor positive job conditions. “The very job characteristics that would reduce stress and job-family tension among employed mothers are difficult for them to obtain because these rewards are linked to an authority and reward structure that places women in marginalized ‘women’s jobs’…” (Glass & Camarigg, 1992, p. 148).

In addition to family effects on labor market participation and the availability of family-friendly job conditions, there is an enormous research literature on the psychological consequences of work-to-family interference and family-to-work interference. What may generically be called work-family conflict can be viewed as leading to work-related or family-related stress and is intended to describe the literal intersection of work and family demands and the negative psychological outcomes based on that intersection. While the discussion above clearly suggests that the levels of work-family conflict are a function of labor market position, job conditions, and the organizational context, much work-family conflict is a function of competing demands and the management of those demands. Work-family conflict has been shown to be related to psychological distress (O’Driscoll, Ilgen, & Hildreth, 1992), stress (Kelloway, Gottlieb, & Barham, 1999), mood, anxiety, and substance use disorders (Frone, 2000; Grzywacz & Bass, 2003) and illness symptoms (Hammer et al., 2004; Klitzman, House, Israel, & Mero, 1990).

The specific structures of paid work and family work may cause distress that spills over or contaminates the level of psychological well-being associated with the other role. Meeting expectations in both the paid labor force and in families requires the management of job demands and scheduling demands in both spheres of activities (Voydanoff, 1988). For both paid labor and household labor, the balance of demands and decision latitude in each sphere can be used to estimate overall distress (Lennon & Rosenfield, 1992). Rosenfield (1989) showed that a woman’s ability to control demands in the work sphere improved her ability to control demands in the domestic sphere. Hughes, Galinsky, and Morris (1992) reported that workers in jobs with high demands and low supervisor support have more frequent marital arguments because high job demands increase the pressure to also complete family-related demands. Pleck and Staines (1985) reported that longer work hours for women lead to greater negative effects on family well-being, and Bolger, DeLongis, Kessler, and Wethington (1989) found that high levels of work hours, for husbands or wives, lead to increased strain for both husbands and wives. Similarly, Sears and Galambos (1992) found that high job demands and low pay for women lead to increased work-related distress, which, in turn, affects marital adjustment. Piotrkowski (1979) and Kanter (1977) found that control over scheduling at work is most crucial for determining whether work hours conflict with family demands.

Byron (2005) reported a meta-analytic review of over 60 studies of work-family conflict that reviewed the antecedents of work-family conflict. She concluded that job stress, family stress, and family conflict affect both work-to-family interference and family-to-work interference. In short, the simultaneous demands of work and family plus the existing degrees of conflict and strain in each sphere can be used to predict well-being. To the extent that structural features linked to job conditions affect job stress, it may be inferred that those same features will indirectly affect work-to-family-related distress. This is precisely what Schieman et al. (2009) argue. Using the demands/resources model of Bakker and Demerouti (2007), they predict work-nonwork interference (as opposed to work stress) as a function of job conditions that are themselves partially determined by social status. This explanation is based on a model that is quite similar to the one outlined in this chapter and strongly suggests the value of conceptualizing work-life interference/conflict/stress research using a social structural explanation to more comprehensively account for well-being.

Conclusion and Prospects

The sociological study of stress reflects the recognition of the importance of social structures and context for understanding this ubiquitous phenomenon. The principle has been articulated frequently and convincingly to define the sociology of health and mental health (Aneshensel, 2009; Aneshensel, Rutter, & Lachenbruch, 1991; Link & Phelan, 1995; Pearlin, 1989, 1999).

This chapter, then, has explored a social structural explanation for the relationship between work and psychological well-being. I have shown how macroeconomic structures and change, labor market structures, social structures of inequality, the organization of work, and the intersection of work with family affect the stressfulness of jobs. These social structures affect exposure to risk (work-related stressors) and access to resources that contribute to feelings of well-being or distress. The macroeconomy defines the overall demand for labor and its form. The labor market distributes those jobs. Social structures of inequality influence labor market participation, and family situations affect labor market participation and preferences. The outcome of these structural effects defines the immediate work context of employees including their exposure to stressful job conditions and coping resources and, hence, stress.

I have organized the discussion in such a way that researchers who are focused on one particular aspect of the work-stress relationship might see how that work articulates with others working in related areas. The articulation between these approaches is not seamless. However, it is also clear that researchers are increasingly aware of the need to account for these dimensions of structure as they develop a more complete understanding of how work affects well-being. This enterprise is also consistent with the notion of a sociological stress process (Pearlin, 1989, 1999), with the notion of social structure as a fundamental cause of illness (Link & Phelan, 1995), and with the need for medical sociology to establish clear connections with the larger discipline of sociology (Pescosolido & Kronenfeld, 1995). Tausig and Fenwick (2011) and Fenwick and Tausig (2007) argue, for instance, that the work-well-being model, as it is mapped onto arguments about the political economy, opens the possibility to think of health outcomes in the same way that we think of economic outcomes (status and income attainment, social mobility). In this way, for example, the sociology of mental health becomes sociology in general.

Limits and Prospects

There is an enormous volume of research on the relationships between work and health/mental health, and the review here has been selective among that research. For example, the emphasis on structural effects should not belie the importance of understanding the relationship between work and emotions. Hochschild (1983) argues that one of the consequences of the macroeconomic shift to service-related jobs in postindustrial societies is the increasing frequency with which jobs require the transformation of human raw material (the customer) via a process of “emotional” labor. She argues further that “emotional labor” exacts a direct cost on the emotional well-being of the worker by estranging the worker from his/her own emotional identity. Precisely because jobs increasingly require “working with people,” we need to incorporate an understanding of how people processing affects well-being (Erickson & Ritter, 2001; Pugliesi, 1999).

Similarly, the current shift to contingent employment and the “flexible” work force represents a historical shift in the relationship between employer and employee. Job characteristics have become moving targets in terms of their relevance to job stress. I have restricted my discussion to only a few of those job characteristics, but it is clear that the simple model of job demands/control should be elaborated to account for a larger set of job conditions.

A growing literature suggests that we need to think of work in a life course perspective. For instance, middle-aged workers sometimes lose their jobs because of the decline of certain industries or plant relocations. These workers may attempt to “retrain” to qualify for existing jobs, but we know almost nothing about the psychological consequences of this increasingly prevalent situation (Geller & Stroh, 1995). Hudson (2007) notes that there is a great deal of mobility from peripheral to core jobs that has consequences for the deterministic way that I have discussed the relationship between social structures of inequality and job stress. And while Hudson also notes that many current jobs may be classified as neither good nor bad, Rich (2010), for example, suggests that the current deep recession is leading to the loss of those “middle wage-middle quality” jobs. Indeed, we probably need to understand work roles as highly fluid and changeable (permanent, temporary, unemployed, underemployed, involuntary, and voluntary). As a result, we need to be much more sensitive to the dynamics of jobs and their structural context (family, neighborhood, social networks) to understand the relationships between work and work-related well-being.

As we think more in terms of social structures and how they affect job stress, we need to elaborate our general theory to better account for how social structures of inequality function as distal causes of illness. In fact, the study of work and stress makes it clear that social institutions such as work represent the day-to-day context in which the injuries of inequality play out.