Abstract
Learning at work is usually seen as beneficial for the professional and personal lives of workers. In this article, we propose that learning’s relationship to worker well-being may be more complicated. We posit that learning can become a burden (instead of always being a benefit) in occupations that are learning intensive and tightly associated with the postindustrial economy. Results of analyses using data from the General Social Survey suggest that learning lessens work–family conflict by increasing job satisfaction, but at the same time, learning makes work–family conflict worse by leading people to work longer hours and exacerbating work-related stress.
Keywords
The dominant form of work in a society during a particular period of time is usually associated with specific problems that plague those who do such work. At the height of the second industrial revolution when manufacturing and various forms of physical labor occupied the largest percentage of the Americans, when husbands were the primary breadwinners, and when most wives managed the home and family, the physical, mental, and emotional costs of work for men and women differed. Subject to the rigors of the mine, the factory or the construction site, men’s bodies succumbed to physical strain, toxic environments and repetitive motions that induced a host of physical ailments ranging from pneumoconiosis (black lung disease), asbestosis, and carpal tunnel syndrome to what was perhaps the most common and ubiquitous of all: physical fatigue (Park, 1934; Walker & Guest, 1952). Indeed, the latter was so widespread that its study, mitigation, and treatment became a preoccupation of early industrial engineers (Gilbreth & Gilbreth, 1916) and, at least partially, motivated advocates of job design (Hackman & Oldham, 1980). Although studies of housewives were rare, it seems reasonable to argue that women too suffered from the physical demands of housework (Cowan, 1985), but perhaps more importantly, many women did not have the opportunity to use their untapped intellectual and interpersonal talents by working outside the home.
The coming of the postindustrial economy (Bell, 1973; Drucker, 1957; Machlup, 1962), which has brought about the decline of manufacturing jobs, the rise of dual earner couples, especially among the middle class, and the steady shift in the division of labor toward professional, managerial, and technical work, has significantly changed the costs of working (Mirowsky & Ross, 2007). Some physical ailments remain common, for example, carpal tunnel syndrome which was once predominant among manual laborers (particularly those who used picks and shovels) has now become pandemic among white collar workers chained to poorly designed desks, chairs, computers, and mice (Konz, 1983). Nevertheless, the costs of postindustrial work seem to be more psychosocial than physical. Both men and women are now more mentally than physically strained when they come home from work. Once at home, disconnecting from work can be extremely hard, so it becomes difficult to shift roles from worker to parent, spouse, and so on. Work–family conflict ensues as people try to juggle the demands of the workplace with the demands of home and family (Blair-Loy & Wharton, 2002; Jacobs & Gerson, 2004). Although volumes have been written about work–family conflict, work–life balance, and similar issues, we shall depart from most previous work, by asking: Are there specific problems that plague those in the postindustrial economy? In particular, what are the consequences of continually having to learn for the other aspects of people’s lives? Our thesis is that work in the postindustrial economy is more focused on continual learning and that continual learning is an unrecognized cause of work–family conflict.
We begin by briefly reviewing the literature on what increases the probability that working men and women will report work–family conflict, the term that we will use to gloss a number of related but subtly different concepts such as work–family balance (Clark, 2000) and work-family spillover (Grzywacz, Almeida, & McDonald, 2002). We then turn to what is known about postindustrial work to pinpoint attributes typically associated with jobs that are now more common in the knowledge economy. We then hone in on one aspect of such work that has attracted little attention: the continual demand to learn new things. We marshal data to show that after controlling for the most prominent predictors of work–family conflict, including occupation, education, and other work conditions, continually having to learn both directly and indirectly increases the odds that Americans will report work–family conflict even though learning ironically also enhances job satisfaction. The irony may well underwrite what other researchers have called the “stress of higher status”: the empirical finding that members of higher status occupations experience more stress and work–family conflict (Schieman, Whitestone, & Van Gundy, 2006). In short, were it not for the greater job satisfaction that comes from continually learning new things at work, learning would unambiguously increase the odds of work–family conflict.
Work–Family Conflict
Both academic and popular commentators usually associate work–family conflict with families in which both spouses are employed. However, work–family conflict did not become a topic of widespread research until dual-earner couples became common among the middle and upper class (Grzywacz et al., 2002; Jacobs & Gerson, 2004; Nätti, Anttila, & Tammelin, 2012). In fact, dual-earner families were once more prevalent among blue- and lower white collar workers than among managers and professional occupations, but during the 1980s, more middle and upper class women entered the labor force so that by 1997 the situation had reversed (Waite & Nielsen, 2001). As Figure 1 shows, the percentage of academic articles mentioning work–family conflict also began to increase substantially after the 1980s with a noticeable jump in rate of publication after 1998.
1
Percentage of articles in JSTOR mentioning work–family conflict.
Significantly, this upturn in publication rate corresponds to the point in time when the percentage of women employed in managerial, professional, and technical jobs began to equal and then surpass the percentage of women employed in sales and administrative support, which had previously been the type of occupations that employed the most women. Figure 2(a), which plots the percentage of American women employed in various occupational categories, illustrates this trend and indicates that the crossover finally occurred in 2002.
2
Managerial, professional, and technical work had been the largest occupational category for American men since 1975 (see Figure 2(b)). Thus, today managerial, technical, and professional occupations account for more employed American men and women than any other occupational category.
(a) Annual percentage of women in the U.S. labor force by occupational categories. (b) Annual percentage of men in the U.S. labor force by occupational categories.
Although scholars recognize that work–family conflict attracted more attention as middle and upper class women entered the workforce in the 1970s and 1980s (Jacobs & Gerson, 2004), they have paid less attention to the fact that women’s entry into the workforce occurred simultaneously with the growing tendency for both men and women to work in professional, technical, and managerial jobs. This entanglement is important because it suggests that the difficulty of balancing work and family may not only be explained by the fact that both partners work, but that both are now more likely to work in occupations intimately associated with the so called “knowledge economy.” Could there be something about postindustrial work that is particularly conducive to work–family conflict?
When explaining work–family conflict, numerous scholars implicitly or explicitly distinguish between the “demands” and “resources” associated with a job or occupation (Glavin & Schieman, 2010; Schieman & Glavin, 2011; Voydanoff, 2004). 3 They treat demands as those characteristics of a job that exacerbate the tendency for work to spill over into family life. In contrast, they view resources as those aspects of work that reduce the probability of spillover either by mitigating demands or by allowing individuals to have more control over how they allocate their time across the two domains. A representative inventory of work demands that have been theoretically or empirically associated with work–family conflict include the following: role overload (Clarkberg & Moen, 2001; Hochschild, 1997; Hochschild & Machung, 1989), role conflict (Voydanoff, 2004), time pressure (Major, Klein, & Ehrhart, 2002; Ritti, 1971; Wallace, 1999), working long hours (Grzywacz et al., 2002; Judge, Boudreau, & Bretz, 1994), job insecurity (Batt & Valcour, 2003; Kinnunen & Mauno, 1998), irregular shift work (Bellavia & Frone, 2005; Grosswald, 2003), unpaid overtime (Brinkley, Fauth, Mahdon, & Theodoropoulou, 2009), work-related contacts outside normal work hours (Schieman & Glavin, 2011) and the use of information technologies, particularly email and teleconferencing technologies (Chesley, 2005). Conversely, among the resources that researchers have thought might ameliorate work–family conflict are: the ability to control one’s schedule (Schieman & Glavin, 2008), authority (Bakker & Demerouti, 2007), job autonomy (Schieman, Milkie, & Glavin, 2009), latitude to make decisions (Karasek, 1979), and access to flextime or flexplace programs (Hill, Hawkins, Ferris, & Weitzman, 2001). 4
There are two problems with using traditional notions of work demands and resources to understand what it is about postindustrial work that may exacerbate work–family conflict. First, resources can sometimes act as demands and vice versa, which creates theoretical trouble for demand-resource theory. A number of scholars have observed this conundrum. Schaufeli and Taris (2014) have noted that whether a work condition serves as a resource or a demand depends on how individuals value and perceive particular work conditions. Although it might seem that workers would value autonomy and discretion and, hence, view them as resources, there are some individuals for whom autonomy and discretion are stressful. Moreover, Schieman (2013) found that challenging work and job autonomy, two work conditions that are normally characterized as resources, were positively associated with higher (not lower) levels of job pressure. Schieman and Young (2010) showed that people who engage in creative work are more likely to report work–family conflict because they also experience greater work demands, more boundary-spanning pressures and, hence, frequently attempt to multitask to handle their work and family duties simultaneously. The second problem which goes to the crux of our research is that the job demands and resources discussed in the literature could apply to almost any job and, therefore, do not highlight what might be special or different about postindustrial occupations.
To understand why postindustrial workers might be particularly susceptible to work–family conflict, ideally, we would like to find the postindustrial equivalents of repetitive motion, demands to work overtime, and narrow job duties that characterized (and still characterize) most factory, clerical, and routine service work. We do not simply want to know that workers in the knowledge economy work long hours, bring work home, or feel overloaded and pressured; we want to know why they do and feel these things. What is it about the nature of postindustrial work that makes managers, professionals, and technical workers particularly susceptible to having work spill over into the other domains of their lives? The continual need to learn found among those who hold such jobs is a strong candidate.
Learning
Learning is integral to, if not definitional of, the idea of a knowledge economy. Although some commentators speak of the knowledge economy as primarily involving a shift in the occupational structure toward occupations that require more years of formal education, specialized training, and credentials (Bell, 1973; Drucker, 1957; Reich, 1991), the majority of the literature highlights the need for professionals, managers, and technicians to do more than apply what they learned in school. Those who hold postindustrial jobs are not only required to confront unstructured and new problems but also to engage in independent decision making (Blackler, 1995; Pyoria, 2005), think creatively (Mirowsky & Ross, 2007; Reinhardt, Schmidt, Sloep, & Drachsler, 2011), acquire new information, learn new tasks (Brinkley et al., 2009), and learn from as well as teach others (Bailey & Barley, 2011; Perlow, 1997). Even accomplished professionals must continually learn, because the problems associated with a particular case, such as designing a building or diagnosing the symptoms of a patient, never neatly match the generalizations and templates on which the professional draws because no two projects or patients are alike (Schön, 1983; Waisberg, 2012).
The need to learn is a frequent component of how researchers characterize postindustrial work. For instance, operationalizations of challenging work and creative work frequently entail the need to continually learn (Mirowsky & Ross, 2007; Schieman, 2013; Schieman & Young, 2010). Learning also figures prominently in ethnographies of postindustrial occupations. Barley and Kunda’s (2004) ethnography of high tech contractors explored in depth the tactics that software developers, IT experts, and various engineers used to avoid technological obsolescence and to remain productive. Although few took formal courses, they developed networks of similarly skilled individuals on whose expertise they could draw. They scanned and read the technical literature, built extensive home computer labs to test new software and equipment in their off hours, volunteered as beta testers and, most importantly, chose to work on projects that would “stretch” their knowledge and skills as much as possible (O’Mahony & Bechky, 2006). In short, to be a productive member of the knowledge economy is to become committed to life-long learning both on and off the job. Unlike schedule control, role overload, time pressure, and other characteristics often associated with the demands of an organization on its workers, learning is an intrinsic characteristic of some occupations and is relentlessly demanded of managerial, professional, and technical workers in much the same way that repetitive motion was demanded of factory workers on assembly lines, attention to detail of clerks, and emotional labor of personal service workers (Halle, 1984; Hochschild, 1983; Walker & Guest, 1952).
Many people perceive opportunities to learn continually to be a highly desirable characteristic of a job and a source of fulfillment. In fact, researchers have argued that opportunities to learn at work contribute to people’s ability to thrive and grow as individuals (Sonenshein, Dutton, Grant, Spreitzer, & Sutcliffe, 2013; Spreitzer, Sutcliffe, Dutton, Sonenshein, & Grant, 2005). This is especially the case if the workplace offers a climate of psychological safety in which the exploration involved in learning new things is not punished (Edmondson, 1999). Indeed, there is some empirical evidence that learning at work increases job satisfaction (Ford & Wooldridge, 2012). Ritti (1971), for example, found that people were more satisfied with their jobs when their work posed intellectual demands. Mikkelsen, Ogaard, and Lovrich (2000) found that job satisfaction was enhanced when organizations created a positive climate for learning. Moreover, Voydanoff (2004) found that opportunities to learn improved workers’ moods at home. From this perspective, one would expect that learning should be a resource in managing work–family conflict.
However, the role of learning in work–family conflict is not so clear cut. The need to learn usually arises when one faces an unstructured or ill-defined problem whose solution requires acquiring new information or learning new tasks or bodies of knowledge. In contemporary organizations, the need to learn usually occurs under time pressure and requires the help of others (Perlow, 1997). When attempting to puzzle through a problem or create new knowledge, workers have a difficult time “checking out,” as Schieman and Young (2010) put it: They continue to be absorbed by work-related thoughts long after they have left the workplace. As Moen, Lam, Ammons, and Kelly (2013) observed, “the professional employees we studied describe how they must decide when they are not working: most of the time they sense pressure to engage in or at least be available for job-related tasks” (p. 83).
From this perspective, learning could be viewed as a demand (rather than a resource) that might increase work–family conflict. Indeed, researchers have found that jobs that require learning also exacerbate stress (Ford & Wooldridge, 2012; Mikkelsen, Ogaard, Lindoe, & Olsen, 2002). Because the phenomenon is so prominent among professionals and managers, researchers have dubbed it “the stress of high status” (Moen et al., 2013; Schieman et al., 2006). Voydanoff (2004) and Schieman and his collaborators (Schieman & Glavin, 2008, 2011; Schieman et al., 2009) have also found that holding a managerial or professional job and possessing an advanced degree increases the probability of work–family conflict. Moreover, Schieman and Glavin (2011) have shown that autonomy, schedule control, and having supervisory authority—attributes common among professionals and managers—exacerbate rather than ameliorate work–family conflict. Finally, even though Voydanoff found that learning may put workers in a better mood when they come home, she found no evidence that it reduces work–family conflict.
Based on previous research, it would seem that learning may have a more complicated relationship to work–family conflict than one might at first expect. On the one hand, researchers have suggested that learning functions as a job resource because it enhances job satisfaction, contributes to a sense of personal growth, and potentially puts people in a better mood at home. On the other hand, learning can become preoccupying, and it can also lead to working longer hours and experiencing more work-related stress: both of which are known to be associated with heightened work–family conflict. One might conceptualize this complicated relationship as depicted in Figure 3. In this model, learning directly exacerbates work–family conflict because it can become mentally consuming, but it also does so indirectly to the degree that it increases the tendency to work more regular hours, work overtime, and experience stress. At the same time, learning may reduce work–family conflict via enhancing job satisfaction which may flow over into the home environment. In other words, long hours, working overtime, and work-related stress mediate and job satisfaction suppresses learning’s effect on work–family conflict.
Direct and indirect (partial mediation) effects of learning on work–family conflict.
In the remainder of this article, we draw on data from the General Social Survey to answer three questions. First, are professionals, managers, and technicians more likely to report the need to learn than members of other occupations? Second, after controlling for a variety of influences known to exacerbate work–family conflict, does the need to learn increase the probability of work–family conflict? Finally, and perhaps most importantly, does the need to learn increase work–family conflict by the way it effects other attributes of the job, such as the number of hours worked, the tendency to work beyond the end of the workday and work-related stress while also reducing work–family conflict by enhancing job satisfaction? In short, if learning is integral to the knowledge economy, could it be a double-edged sword?
Methods
Data and Measures
To examine the preceding questions, we employed data from the General Social Survey (GSS). The GSS is a nationally representative survey of randomly selected households in which interviews are conducted face-to-face with respondents (Smith, Marsden, & Hout, 2015). The National Opinion Research Center (NORC) administered the GSS annually from 1972 to 1994 (except in 1979, 1981, and 1992), then, in 1996, it began collecting the data biannually. Questions cover respondents’ views of their lives, work and other topics as well as household attributes. The GSS also includes modules that focus on specific issues but that are not asked during every administration. We made use of the “Quality of Working Life” module, which was included in 2002, 2006, 2010, and 2014. This module contains many questions that focus on the respondent’s experience at work that are pertinent to the subject of our inquiry. Because the same questions were administered repeatedly, we could pool the data to create a data set of 5,460 unique respondents who were between the ages of 18 and 65. We chose these age boundaries because 18 is typically the age at which Americans graduate from high school, and after 65 most Americans have finished rearing children, have lighter workloads, and often retire (Munnell, 2011).
Variables
To measure work–family conflict, our dependent variable, we used responses to the question: “How often do the demands of your job interfere with your family life?” Respondents had the option of answering with the following four responses: (1) never, (2) rarely, (3) sometimes, and (4) often. Although we shall refer to this variable as “work-family conflict,” readers should keep in mind that we are measuring how work spills into family life, not the reverse.
The independent variable in which we are primarily interested is the degree to which people must engage in learning to do their work. To assess learning, we employed the question: “My job requires that I keep learning new things.” Respondents assessed their jobs’ demand for learning by answering: (1) strongly disagree, (2) disagree, (3) agree, or (4) strongly agree. Note that the question assesses the respondents’ perceptions of how critical learning is for what they are required to do, rather than either the actual time they spent learning or whether their attempts to learn were successful. In our analyses, we call this variable learning.
Occupations and Educational Attainment of Individuals in the Sample.
Source: General Social Surveys 2002, 2006, 2010, and 2014.
The occupational categorization is from the 1988 International Standard Classification of Occupations (Smith, Marsden, and Hout, 2015).
Researchers also frequently use education as an indicator of who is and is not a member of the knowledge economy (Brinkley et al., 2009; Davenport, 2005; Drucker, 1957; Kelloway & Barling, 2000; Nätti et al., 2012; Pyoria, 2005). Scholars have shown that education is also a predictor of work–family conflict (Grzywacz et al., 2002; Schieman & Glavin, 2008, 2011). Accordingly, we included a variable that classified the respondents’ education into five categories based on the highest degree the respondent had obtained: (a) less than high school, (b) high school degree, (c) associate or junior college degree, (d) bachelor’s degree, or (e) graduate degree. Panel B in Table 1 displays the number and percentage of respondents at each level of education.
We assessed the influence of numerous work-related variables that other researchers have found to be associated with work–family conflict. For example, Hill et al. (2001), Grzywacz et al. (2002), and Voydanoff (2004), among others, have shown that long hours of work are associated with more work–family conflict. Accordingly, we used the question: “If working full or part time: How many hours did you work last week, at all jobs.” When answering respondents gave the specific number of hours they worked. We also measured workload with a question assessing overtime, “How many days per month do you work extra hours beyond your usual schedule,” to which respondents replied with a number of days. Finally, we measured whether respondents worked a second job or not, reasoning that holding down two jobs would make it more difficult to attend to family matters. The question was phrased, “Do you have any jobs besides your main job or do any other work for pay?” Respondents answered “yes” or “no.”
We measured job satisfaction using the following question: “On the whole, how satisfied are you with the work you do—would you say you are very satisfied, moderately satisfied, a little dissatisfied, or very dissatisfied?” In addition, we assessed work-related stress using the following question: “How often do you find your work stressful?” Respondents could reply: (1) never, (2) hardly ever, (3) sometimes, (4) often, and (5) very often.
One’s experience in the workplace is not the only contributor to work–family conflict. The structure of a family and its activities are also important. In particular, researchers have found that work–family conflict is strongly associated with being a parent and having school aged children (Grzywacz et al., 2002; Hill et al., 2001; Jacobs & Gerson, 1998; Moen et al., 2013). We measured parenthood using the question: “How many children have you ever had? Please count all that were born alive at any time (including any you had from a previous marriage).” We call this variable parenthood which we treated as dichotomous: either the respondent was or was not a parent. We also used three variables to capture the age of children currently living in the household: (a) number of children less than 6 years old, (b) number of children between 7 and 12 years old, and (c) number of children between 13 and 17 years old. We also assessed whether respondents were married.
Finally, we controlled for gender, age, and race. Gender and race were dichotomous variables in which “1,” respectively, indicated male and White. Age was coded in years ranging from 18 to 65. We also included controls for the year in which the survey was administered.
Analysis
We used Ordinary Least Square (OLS) regressions to assess (a) whether managers, professionals, and technicians are more likely than members of other occupations to report the need to learn and (b) whether the need to learn at work increases the probability of work–family conflict. To assess whether the number of hours worked, the need to work overtime, feelings of work-related stress, and job satisfaction enhance or suppress the relationship between learning and work–family conflict, we conducted mediation analyses (Baron & Kenny, 1986; MacKinnon, Fairchild, & Fritz, 2007).
Results
Do Members of Some Occupations Report More Learning Than Others?
Descriptive Statistics.
Source: General Social Surveys 2002, 2006, 2010 and 2014.
The mean values for these variables are not readily interpretable because large numbers of respondents had no children in a particular age bracket.
p < .05.
Linear Regression Models of Several Independent Variables on Learning.
Source: General Social Surveys 2002, 2006, 2010 and 2014. Standard errors are in parentheses.
p < .001, **p < .01, *p < .05 (two-tailed tests).
Model 2 adds levels of education to the mix and treats those with less than a high school education as the omitted variable. The R2 increased from .116 to .131, and people with all degrees reported needing to learn more at work than people who did not graduate from high school. Model 3 adds work demands and resources to the analysis. The more hours a person worked, the more days they worked overtime, whether they held a second job, whether they were satisfied with their work, and whether they experienced work-related stress were all significantly associated with the need to learn at work. The fact that the R2 for Model 3 rose from .131 to .189 indicates that work conditions matter substantially for learning. Models 4 and 5, respectively, add family attributes and all remaining control variables to the regression.
The results in Table 3 strongly indicate that professionals, managers, and technicians were more likely to have to learn at work to do their jobs. Moreover, this relationship held regardless of education and the demands and resources associated with a job.
Does the Need to Learn at Work Increase Work–Family Conflict?
Linear Regression Models of Several Independent Variables on Work–Family Conflict.
Source: General Social Surveys 2002, 2006, 2010, and 2014. Standard errors are in parentheses.
p < .001, **p < .01, *p < .05 (two-tailed tests).
The overall pattern of results suggests that the need to learn at work does have a complicated relationship with work–family conflict. On the one hand, learning seems to function as a job demand. As such, it should be classed along with long hours, working overtime, and work related stress. On the other hand, as some researchers have noted, learning serves as a resource which enhances job satisfaction. As such, learning should be classed as a resource along with job satisfaction because it may reduce work–family conflict. To determine learning’s net effect, we tested the relationships depicted in Figure 3. We undertook a mediation analysis to determine when learning is likely to prove detrimental or beneficial to the workers and their families.
By What Paths Does the Need to Learn Exert its Effect on Work–Family Conflict?
Linear Regression Models for the Mediation Analysis.
Source: General Social Surveys 2002, 2006, 2010, and 2014. Standard errors are in parentheses.
p < .001, **p < .01, *p < .05 (two-tailed tests).
Models 1 through 4 in the left hand panel of Table 5, respectively, show the effect of learning on the proposed mediators and the suppressor: work hours, overtime, work-related stress, and job satisfaction. In each case, learning remains significant (p < .001) even after all controls are included. Hence, Baron and Kenny’s second condition holds: learning significantly affects the mediators. Models 5 through 9 of Table 5 address Baron and Kenny’s third condition. Models 5 through 8, respectively, exclude one of the mediators or the suppressor. When work hours are included in the regression, the coefficient of learning (.103 in Model 5) falls to .086 (in Model 9). Similarly, when overtime is included, the coefficient drops from .094 (Model 6) to .086 (Model 9). The coefficient also falls when stress in included: .124 (Model 7) to .086 (Model 9). In contrast, when job satisfaction is included, the coefficient for learning increases from .063 (Model 8) to .086 (Model 9). Hence, the analysis suggests that work hours, overtime, and stress act as mediating variables, while job satisfaction serves as a suppressor variable, as depicted in Figure 3.
To confirm the model further, we conducted Sobel tests on the coefficients of learning in Models 5 through 9. A Sobel test indicates whether a drop (or rise) in the coefficients for learning are statistically significant and provides an estimate of the proportion of the total effect that is mediated (or suppressed) by each variable. The results of the Sobel tests were significant for all four changes in the coefficient for learning. Specifically, the Sobel tests estimated that work hours mediated 12% of the total effect of learning on work–family conflict (z = 3.95, p < .001); overtime mediated 10% of the total effect (z = 4.04, p < .001); stress mediated 31% (z = 7.42, p < .001), while job satisfaction suppressed 34% of learning’s effect on work–family conflict (z = –5.54, p < .001).
Conclusion
Many researchers and commentators have portrayed learning as a distinctive characteristic of work in the knowledge economy (e.g., Blackler, 1995; Davenport, 2005; Garrick & Clegg, 2000; Lewis, 2004; Perlow, 1999; Schement, 1990). Learning is also often seen as a positive job characteristic that induces personal growth and satisfaction. College students long for jobs that will allow them to continually learn throughout their lives. Indeed, learning associated with work is portrayed in the academic literature and in the popular press as a goal for which we should strive, if for no other reason that it is thought to benefit the economy as well as the individual. Yet, few researchers have put these assumptions to the test. Ours is the first study to statistically test whether knowledge occupations require more work-related learning than other occupations and whether such learning is always positive for the individual.
In contrast to what many might expect, we found that learning exacerbates work–family conflict, especially among professional, managerial, and technical occupations. Moreover, learning does so both directly and indirectly. The need to learn directly contributes to work–family conflict most likely by continually preoccupying the learner’s attention. Learning enhances work–family conflict indirectly by leading people to work longer hours and more days while exacerbating the work-related stress that they experience. We also found that learning increases job satisfaction which, in turn, moderates learning’s effects on work–family conflict.
Some might argue that increasing demands to learn in a knowledge economy represent a form of work intensification. This is particularly likely to be true as project work becomes increasingly common. Project work is likely to place a heavy demand for learning on workers because projects usually focus on unique and one-off products or goals, and they often require coordination of people who have not previously worked together and who share few routines. It is well known that managers tend to underestimate the amount of time it will take to complete the project (Brooks, 1975). Of course, there are practical and cognitive limitations involved in our inability to make satisfactory estimates of what projects require (Heath & Staudenmayer, 2000; Kahneman & Tversky, 1979); nevertheless there are also multiple economic reasons for doing so. When companies compete for projects, they may underestimate the project’s duration so that the client will choose them. Internally, project managers may underestimate time frames to convince top executives that they can minimize time to market. Finally, managers may underestimate project durations because they cannot predict the roadblocks that a project will encounter on its way to completion. In any case, the tendency to underestimate time creates pressures for workers to intensify their learning when problems threaten deadlines. In such scenarios, learning becomes less of an opportunity and more of a burden.
The upshot is that postindustrial workers often have little choice but to allow work to spill over into family and leisure time. The cost is born not simply by the worker but by his or her family as well. As the knowledge economy expands, the potential negative effects of learning on work–family conflict may increase. In this regard, it is particularly telling that managers, professionals, and technicians are the most likely to report the need to learn and that the need to learn is consistently associated with the spillover of work into family life. Accordingly, it is not unreasonable to argue that learning is a double-edged sword and may represent the postindustrial equivalent of the manufacturing era’s physical demands. In the postindustrial economy, we may be more intellectually stimulated, but our lives may become as monothematic as that of the man on the assembly line. In fact, we may be worse off than assembly line workers, because the latter were typically protected by unions, they knew with whom to negotiate and they were granted two to four weeks of annual vacation which the professionals and managers are notorious for foregoing (Alesina, Glaeser, & Sacerdote, 2006; Bell & Freeman, 2001).
The results of this study pose an important caveat for arguments that ground work–family conflict in the rise of dual-earner families. Although it is undoubtedly true that work–family conflict is greater when both spouses work, the nature of their work also matters. Our contention is that the so-called “stress of higher status,” observed by Schieman and Glavin (2011), is intimately tied to the fact that people must cope with continual and intensified demands to learn. Because so many dual earner couples are highly educated managers and professionals, researchers need to begin untangling the effects of the nature of work from the effects of family, education, and occupation per se.
Researchers also need to begin to attend to how workers successfully or less successfully manage the demands of learning and their implications for work–family conflict. Family friendly programs such as parental leave and flextime are unlikely to address the pressures of learning, even when workers take advantage of them. One cannot stop thinking about problems simply because one has an afternoon or a day off to take care of family business. In fact, if demands to learn remain constant, such programs may actually exacerbate the problem because they remove the worker from the work context where more help with learning is likely to be available.
Organizations and researchers also need to consider what structures and systems might provide some protection from the pressure of continual learning. As Perlow (1997) has noted, workers are rarely rewarded for helping each other solve problems nor are they given time to focus on learning what they need to learn. Perlow (1999) showed that an intervention as simple as “quiet time,” setting aside a few hours every other day for the individual to focus solely on his or her own tasks, improved productivity, and satisfaction by eliminating interruptions which disrupted concentration and the ability to learn. Sabbaticals might be another programmatic option. Finally, and most importantly, to deal with the deleterious consequences of incessant demands to learn new things, organizations need to learn how to set more realistic project timelines, perhaps by routinely, and realistically, adding cushion time into their estimates. What our research on learning and work–family conflict shows is that we should be more circumspect about what we wish for in our jobs. Even opportunities that at first glance appear to be advantageous and desirable may have a hidden downside that undermines the obvious benefits of the opportunity.
Footnotes
Occupational Classification of Individuals in the Samplea
| Occupational category | Examples |
|---|---|
| Elementary occupations | Street vendors, door-to-door salespersons, domestic and office cleaners, sweepers, garbage collectors, and freight handlers. |
| Plant and machine operators | Ore and metal furnace operators, well drillers and borers, paper-pulp plant operators, machine-tool operators, and petroleum plant operators. |
| Service workers | Travel guides, housekeepers, cooks, waitresses, bartenders, child-care workers, barbers, undertakers, fire-fighters, and police officers. |
| Clerks | Office clerks, secretaries, receptionists, store clerks, mail carriers, cashiers, tellers, and croupiers. |
| Craftsmen and trades workers | Carpenters, blacksmiths, bricklayers, plumbers, electronic fitters, painters, welders, miners and quarry workers, musical instrument makers, handicraft workers in wood, leather, textile, etc., glassmakers, butchers, and upholsterers. |
| Managers | CIOs, general managers in manufacturing, construction, transportation, restaurants, hotels, etc.; miscellaneous office supervisors; and department managers of production, operations, finance, R&D, personnel, sales, marketing, etc. |
| Technicians | Computer assistants, medical equipment operators, air traffic controllers, optometrists and opticians, police inspectors and detectives, trade brokers, and bookkeepers. |
| Professionals | Engineers, programmers, architects, biologists, physicists, medical doctors, dentists, lawyers, sociologists, economists, political scientists, psychologists, writers, and journalists. |
The occupational categorization is from the 1988 International Standard Classification of Occupations. These are just a few representative examples of each occupational category. The complete list can be found in Smith, Marsden, and Hout (2015).
Acknowledgments
The authors wish to thank the Fulbright Program and Becas Chile for supporting Gonzalo in pursuing his research. The authors would also like to thank Paul Leonardi, Renee Rottner, Melissa Valentine, Pam Hinds, Katy DeCelles, Zach Rodgers, Ece Kaynak, Hatim Rahman, and the participants of WTO labs at Stanford for their insightful comments throughout this project. The authors are also grateful to Michael Rosenfeld for his help and encouragement on the first drafts of this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.