Social Science Research 47 (2014) 108–120

Contents lists available at ScienceDirect

Social Science Research journal homepage: www.elsevier.com/locate/ssresearch

Subjective time pressure: General or domain specific? Sibyl Kleiner ⇑ Dept. of Public Affairs and Sociology, School of Economic, Political & Policy Sciences, The University of Texas at Dallas, 800 W. Campbell Road, GR-31, Richardson, TX 75080-3021, United States

a r t i c l e

i n f o

Article history: Received 19 September 2012 Revised 15 January 2014 Accepted 30 March 2014 Available online 8 April 2014 Keywords: Time pressure Work and home contexts Work stress Home stress

a b s t r a c t Chronic time pressure has been identified as a pervasive societal problem, exacerbated by high demands of the labor market and the home. Yet time pressure has not been disaggregated and examined separately across home and work contexts, leaving many unanswered questions regarding the sources and potentially stressful consequences of time pressure. Using data collected in the United States General Social Survey waves 2002 and 2004, this study disaggregates time pressure into the domains of home and work, and asks whether considering time pressures within distinct work and home contexts reveals distinct predictors or associations with stress. Findings show that both predictors and stress associations differ across work and home pressures, revealing both methodological and theoretical implications for the study of time pressure and work and family life more generally. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Subjective time pressure, or the sense that one’s duties and responsibilities exceed one’s ability to complete them in the time available, is arguably a persistent characteristic of modern life for many individuals (i.e., Gleick, 1999). Particularly in the United States, discourse on time scarcity is well-established and widespread (e.g., Clarkberg and Moen, 2001; Gleick, 1999; Gross and Sheth, 1989; Hochschild, 1997; Schor, 2004). Yet research on the experience of time pressure is relatively scarce. More is known, for example, about how time is experienced subjectively than how a lack of time is experienced (see, for example, Dapkus, 1985; Flaherty, 1991; Flaherty and Meer, 1994; Meck, 2005). In particular, we do not understand whether time pressure differs across important life settings, such as the contexts of work and home. At chronic levels, research suggests that feelings of time pressure may increase overall stress and challenge one’s mental health (Zuzanek, 1998; Roxburgh, 2004). Time pressure is conceptualized as a stressor, constituting one example of how social roles challenge or strain individuals, and potentially create a stress response (Goode, 1960). The social stress paradigm posits that potential stressors are, ideally, best interpreted and understood in context. For example, stressors may carry over from one domain to another (see Thoits, 1995). The domains of home and work, in particular, are distinct contexts that are highly demanding of individuals’ time (Coser, 1974). In these contexts individuals engage with different roles, behaviors, goals, networks, social interactions, and habits (Clark, 2000; Eby et al., 2010). The contexts of home and work are also, quite often, distinct places (Gieryn, 2000). Yet we do not know whether time pressure differs across these two settings. In this study I examine whether time pressures at home and work are distinct by asking (1) whether time pressures at home and at work are key components of an overall sense of pressure; (2) whether the social distribution of time pressures differ across home and work domains, and (3) whether time pressures at home and work have different associations with

⇑ Fax: +1 972 883 6297. E-mail address: [email protected] http://dx.doi.org/10.1016/j.ssresearch.2014.03.013 0049-089X/Ó 2014 Elsevier Inc. All rights reserved.

S. Kleiner / Social Science Research 47 (2014) 108–120

109

stress. We know relatively little about the experience of time pressure, particularly the contexts in which time pressure arises, and how social roles, time pressure, and stress are related across the distinct contextual domains of home and work.

2. Background 2.1. Definition, significance, and sources of subjective time pressure Subjective time pressure refers to the sense or awareness that there is too much to do and not enough time in which to do it (e.g., Roxburgh, 2002). Subjective time pressure, therefore, lies at the intersection of time experience and social roles: it involves both the perception of time, and of role obligations perceived as necessary to accomplish within a given time period. Phenomenological analysis confirms that the experience of time pressure involves both experiencing tempo (pacing of time) and experiencing limits and choices (having to choose one action versus the other) (Dapkus, 1985). The experience of time and its constant progression, the sense of the burden of one’s social obligations, and the reality that one cannot do everything at once, are all components of the perception of time pressure. The experience of time is intimately connected with society, yet a full sociological understanding of time pressure remains lacking. Societies imbue individuals with the tools and experiences that make up their consciousness and interpretation of time, and provide them with the resources to judge whether a given experience feels in sync with the normal pace and flow of daily life (Flaherty, 1991). Temporal aspects of daily life, such as what days or times of the week stores are open, whether to rest during the day, and even the modern calendar, are intertwined with social life and fade into the background of everyday experience (Sorokin and Merton, 1937; Zerubavel, 1981). Yet at times we feel that time is simply too short to finish everything we need to get done. Modern life, particularly in the United States and other similar nations, is often described as highly time pressured (see Hamermesh and Lee, 2007). Nonetheless the phenomenon of subjective time pressure – the perception of lack of time, relative to demands – is relatively understudied. Feelings of time pressure are of greatest theoretical and practical concern when they become frequent and characteristic of daily experience. The sense of time pressure seems to be typical of a normal, healthy human experience (Goode, 1960; Levine, 1997; Zuzanek, 1998). It is produced by social roles creating strains and challenges that individuals must work against – an instance of the general phenomenon referred to as role strain (Goode, 1960). It is at high (sustained/frequent, enduring) levels of occurrence that time pressure is thought to be a problem, and may be considered a chronic strain or stressor (Roxburgh, 2004; Szollos, 2009). At chronic levels, subjective time pressure is thought to potentially create unhealthy levels of stress, constituting one pathway by which social roles and institutions affect psychological well-being (also see Maule and Hockey, 1993; Pearlin et al., 2005). Empirically, researchers most often measure chronic subjective time pressure based on questions that tap into overall feelings of pressure. For instance, one common measure uses a question about how frequently one feels rushed. This question is found in several survey years of the General Social Survey, as well as in other national surveys (e.g., Hamermesh and Lee, 2007; Mattingly and Sayer, 2006; Robinson and Godbey, 1997). Other researchers use more complex measures of time pressure, combining multiple items, such as ‘‘How often do you wish there were more hours in a day?’’ ‘‘You feel rushed to do the things you have to do,’’ or ‘‘You never seem to have enough time to get everything done’’ into scales that taps into overall feelings of time pressure (Hilbrecht et al., 2008; Robinson and Godbey, 1997; Roxburgh, 2002, 2004; Van Emmerik and Jawahar, 2006; also see Tézli and Gauthier, 2009). These measures tap into on overall self-assessment of time pressure, which allows the researcher to rate total levels, but the measures do not directly address the contexts in which respondents feel time pressured. Studies of overall feelings of time pressure confirm that reported feelings of chronic time pressure are not purely perceptual, but implicate the greater presence of objective time constraints (Gimenez-Nadal and Sevilla-Sanz, 2011). Those who are time pressured have less time alone (Mattingly and Bianchi, 2003) to engage in leisure, relax, and recuperate from stress. For instance, parents of young children, and individuals who do more housework (usually women and girls; see also Hilbrecht et al., 2008; Sayer, 2005), have greater time demands and consequently report greater time pressures (Mattingly and Sayer, 2006; Roxburgh, 2002). Workers, and particularly those who work longer hours, also tend to report greater feelings of time pressure (Gimenez-Nadal and Sevilla-Sanz, 2011; Hamermesh and Lee, 2007; Roxburgh, 2002). Although work and home obligations are common sources of time pressure identified in the literature, other key dimensions of experience have been linked to greater overall feelings of subjective time pressure, particularly socio-economic status and emotional state. Experimental, demographic, and economic studies have reported various measures or proxies for socio-economic status as predictors of overall subjective time pressure. These studies suggest that time pressure is linked to socio-economic advantage either because greater earnings drive up consumption (Hamermesh and Lee, 2007), or because greater financial resources increase the perceived value of time (DeVoe and Pfeffer, 2011). Alternative explanations, such as the possibility that time pressure is linked to SES primarily through workplace demands (i.e., Schieman et al., 2006), have not been explored. Mood is another likely source of time pressure. Research suggests that although time pressure may have depressing consequences, particularly if one does not have the economic resources to deal with the pressure (Roxburgh, 2004), negative mood states – whether rooted in life circumstances or personality traits – may, in turn, make tasks seem insurmountable, increasing the sense of time pressure (i.e., Kangas and Meyerson, 2008). Those who are depressed also experience a distorted

110

S. Kleiner / Social Science Research 47 (2014) 108–120

sense of time (Bschor et al., 2004), and may be more likely to report negative emotions and perceptions as a symptom of depression. While a number of studies address an overall sense of time pressure, we know relatively little about subjective time pressure experienced within more specific contexts. 2.2. Subjective time pressure across work and family domains Work and home are key contexts for examining time pressure. Families and work organizations both represent ‘‘greedy institutions’’ that demand total attention and commitment (Coser, 1974). Research to date has, furthermore, consistently linked time pressure to family and work roles (see previous section). Although current theory on work and family recognizes that these two domains have permeable boundaries, with objects, thoughts, feelings, people, and conversations able to travel across work and family contexts, work and family institutions are also distinct (Clark, 2000). Work and family demands arise within different settings, focus on distinct goals and behaviors, and engage different networks. While it makes sense on an intuitive level to study overall time pressure, in part because time is always present in our lives, social institutions bracket our role activities into discrete settings. In particular, the empirical findings from the time pressure literature implicate the contexts of work and home as major sources of time pressure. The housework and childcare duties, multitasking, and lack of solitude disproportionately experienced by women primarily occur at home (Mattingly and Bianchi, 2003), and the home is a context more often experienced as positive by fathers than by mothers (Larson et al., 1994). The longer hours of work consistently associated with greater time pressures, even in cross-national research, usually take place in a separate work setting. Yet studies have not disaggregated home and work pressures to see how they differ from each other. Given that home and work often comprise distinct social networks, goals, and behaviors, we might expect divergent patterns of subjective time pressure across these domains. There are also several empirical paradoxes in the time pressure literature that would benefit from more nuanced attention to home and work contexts. First, while the American cultural discourse is that time pressure, overall, has grown, the academic literature suggests that time pressure has primarily increased for women. Mattingly and Sayer (2006) find that time pressure in the United States (during the period 1975–1998) increased for women but not men, resulting in a gender disparity in men’s and women’s time pressure. The lack of gender specificity in general cultural discourse about time pressure, however, is puzzling. It is possible, given women’s high contribution to household labor despite growing levels of participation in the labor force, that women’s time pressure has particularly increased at home. If this is the case, then analytically disaggregating subjective time pressure into time pressure at home and time pressure at work can address whether the gender difference in time pressure is driven by time pressure at home. If the gender difference in time pressure is driven by the home context, we should see a stronger gender difference in time pressure at home compared to time pressure at work. Another area of research that would benefit from greater attention to the work and home contexts of time pressure is the paradox of SES and time pressure. Several studies have found a relationship between higher income or work hours and higher levels of time pressure, and these findings have been replicated in multiple national contexts (Hamermesh and Lee, 2007). This finding is curious, however, because higher SES is related to better mental and physical health (Link and Phelan, 1995), and time is crucial for maintaining health. Those who have more financial resources and access to more prestigious jobs should have more control over their time, particularly through greater access to schedule control and the ability to purchase time-saving devices and services. One common argument among researchers finding a relationship between SES and time pressure is that greater consumption activities drives up time pressure among the wealthy (Becker, 1965). If consumption is responsible for the SES discrepancy in time pressure, disaggregating time pressure into home and work domains should reveal a strong SES distinction within the home domain, where consumption takes place, and where time spent on purchasing and consuming goods would take away from discretionary time. Alternatively, if time pressure is disproportionately experienced by higher status workers, the relationship between SES and time pressure should be strongest in the context of the workplace (i.e., Schieman et al., 2006), A more complex issue that could be best addressed by contextualizing time pressure, and particularly the ways that time pressure relates to stress, is the qualitatively different experiences that may constitute time pressure within work and home domains. Subjective time pressure is arguably a stressor, and one which has only been taken up as an important topic of study by sociologists of stress within the past decade (Roxburgh, 2004). A stressor is any internal or external demand that requires an individual to readjust their typical patterns of behavior in order to cope behaviorally and emotionally (see Thoits, 1995). Life events such as job loss or death of a loved one represent a type of stressor requiring rapid, major adaptation, while chronic strains such as perceived discrimination, time pressure, or job insecurity require ongoing readjustments for a prolonged period (see Thoits). Both forms of stressors may worsen mental and physical health, but chronic strains such as time pressure are studied less often than life events. Although studies have not directly and systematically examined time pressure as a stressor at home versus work, a range of studies implicate several important and divergent issues across the two domains. Time pressure at work is considered a stressor, with the potential to hinder task completion (DeDonno and Demaree, 2008; also see Teuchmann et al., 1999). Yet research finds that time pressure can also foster creativity. Particularly within organizations that foster creativity, time pressure and creativity are related by an inverse U-shaped function (Baer and Oldham, 2006). Furthermore, some work that is quite challenging and even time pressuring may also be a source of ecstatic, creative ‘‘flow’’ experiences

S. Kleiner / Social Science Research 47 (2014) 108–120

111

(Csikszentmihalyi, 1997). These findings suggest that time pressure within the work environment (and coping responses to this stressor) may not always produce a stress response, and may at times garner particular psychological rewards. In contrast, studies of time pressure within the home link it with feelings of chaos and guilt, to the extent that some individuals with high time pressures at home seek to escape the home environment altogether, even when it worsens their family relationships (Hochschild, 1997; see also Offer and Schneider, 2011). Recent qualitative research on women who homeschool their children, for example, reveals that they have so little time for themselves that many consider housework without children present as their only respite, and use these occasional quiet moments as a chance to reflect and unwind (Lois, 2010). These studies do not suggest that time pressure at home is accompanied by enriching rewards such as psychological growth, cognitive challenge, or creativity, as described in the work literature. Time pressure at home may indicate limited access to psychologically enriching activities that buffer overall stress, whereas time pressure at work may not be a stressor that consistently carries over to the home environment. Furthermore, because the home environment is critical for sleep and self-care, time pressure in this setting may be particularly pervasive in its stress effects. Recent scholarship critiques researchers in sociology and social psychology for failing to look across multiple domains of people’s lives, rather than focusing on one particular social role or context (McLeod and Lively, 2006. This study responds to this insight by examining time pressure across two life domains – home and work. These are the two primary contexts implicated by existing literature on time pressure. Yet there is reason to expect that time pressure in these two contexts might operate differently, and might hold different implications for stress. This study tests these possibilities by examining whether disaggregating time pressure into home and work pressures yields different predictors and different associations with stress.

3. Methods 3.1. Samples and sample composition Two data sources are required for this study. Home and work pressures have not been disaggregated and examined in previous research, so validity testing is needed in order to judge whether the novel measures available truly represent a disaggregation of a known measure of overall time pressure. Yet no single dataset exists that includes not only home and work pressures but also a standard, widely used overall measure of time pressure. The use of two very similar datasets helps establish construct validity by comparing predictors of a non-standard, constructed measure of overall time pressure – composed of home and work pressures combined – to predictors of the known standard. The data sources come from two survey waves of the General Social Survey (GSS). The GSS has been conducted annually within the United States since 1972 and biennially since 1994. Respondents are targeted using full probability sampling of English-speaking individuals over the age of 18 who are living in non-institutional settings, and they are interviewed for an average of one and a half hours (Davis et al., 2007). In 2002, the GSS included separate questions about time pressure at home and at work through the auspices of the International Social Survey Programme (ISSP). This is a separate, international initiative that provides a set of standardized questions that all participating countries agree to field on a national level. Because of the size and complexity of the GSS, only a certain proportion of respondents are asked specialized questions – such as those about time pressure, or those asked within the ISSP module. 1171 individuals were asked these questions, with 92% providing responses on all questions used in the following analyses, and 4% missing solely on income, yielding a total sample size of 1044 for the analyses using the GSS 2002. In 2004, the GSS included a commonly used measure of overall time pressure: how often the respondent felt ‘‘rushed.’’ In this sample, 984 respondents were asked about overall time pressure, and of these, 83% provided responses to all variables used in the main analyses, with 10% missing solely on whether they typically felt sad and blue, and another 2% missing solely on income. Overall, 818 cases from the GSS 2004, with full data on all questions, were used in this study. To examine the distinct predictors of home and work pressures, and their relationships with stress, a currently working subsample from the GSS 2002 is analyzed. The full sample is also examined where possible, for comparison purposes. The subset of currently working respondents (N = 659) is particularly important for examining work stress and time pressure at work, as these phenomena cannot be assessed for those who are not currently working for pay. Restricting the sample to only those who are currently working allows for greater comparability across analyses of home and work pressures, because this allows the same sample – and sample size – to be used across the two measures. For one outcome variable, additional restrictions are applied. Additional models examine married or cohabitating respondents’ answers to a question on disagreements on household work. As a result, models including this measure use a smaller sample size (N = 601, or for those currently working, N = 392) composed only of married or cohabitating respondents. Overall, data from the GSS 2002 and 2004 are quite similar, which is not surprising given that the same institution fielded them. Response rates for both survey years (2002 and 2004) were 70% (Davis et al., 2007), and approximately 63% of respondents in both the GSS 2002 and GSS 2004 were currently working. In both years and both samples, more than half the respondents were women. The average age of both full samples was 45, although the workers are somewhat younger – approximately 40 in 2002, and 41 in 2004. Average level of work, for those currently working, was approximately

112

S. Kleiner / Social Science Research 47 (2014) 108–120

Table 1 Descriptive statistics, full sample, GSS 2002 and 2004. Variable

a

GSS 2002

GSS 2004

Obs

Mean

Std. dev.

Min

Max

Obs

Mean

818

2.16

818 818 818 818 818 818 818 818

0.56 45.77 0.18 0.74 0.53 26.67 16.74 0.30

Time pressure measures Overall time pressure Pressure at home Stress indicators Stress at home Disagree about houseworka

1044 1044

6.30 3.80

2.11 1.15

2 1

10 5

1044 601

2.83 2.19

1.24 1.19

1 1

5 5

Independent variables Female Age Unhappy Kids Married Hours/wk Income College degree

1044 1044 1044 1044 1044 1044 1044 1044

0.59 44.89 0.04 0.31 0.53 26.16 2.60 0.25

0.49 16.75 0.20 0.46 0.50 22.88 1.14 0.43

0 18 0 0 0 0 1 0

1 89 1 1 1 89 4 1

Std. dev.

Min

Max

0.67

1

3

0.50 16.40 0.39 0.44 0.50 23.68 5.82 0.46

0 18 0 0 0 0 1 0

1 89 1 1 1 89 24 1

Married sample.

42 h/week in both 2002 and 2004. The 2004 samples contained somewhat more college graduates as compared to the 2002 sample. 3.2. Measures 3.2.1. Subjective time pressure In 2004, which is the closest and most recent wave of the GSS containing a measure of time pressure, a question was asked about feeling ‘‘rushed.’’ The exact question was: ‘‘In general, how do you feel about your time – would you say you always feel rushed even to do things you have to do, only sometimes feel rushed, or almost never feel rushed?’’ Response choices were always, sometimes, or never. On average, respondents tended toward more frequently feeling rushed than not, both among the full sample and the worker subsample. To examine time pressure across the separate domains of home and work, this study uses measures found in the GSS 2002. In this data source, time pressure at home and at work were assessed using the questions from the ISSP module ‘‘Gender and Family Roles.’’ A question specific to work asks: ‘‘To what extent do you agree or disagree? There are so many things to do at work, I often run out of time before I get them all done.’’ A question specific to home asks for a response to: ‘‘There are so many things to do at home, I often run out of time before I get them all done.’’ Possible responses were strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree, and are shown here with a high score indicating greater levels of pressure. As shown in Table 3, respondents, on average, agreed that they were chronically pressed for time at home – or, among workers, at work. The two time pressure measures have a low to moderate, and significant, correlation. For cases with valid responses on both time pressure questions (responses to current work pressure were considered valid only if the respondent was currently working), work and home pressures have a correlation of .28 (p < .0001). Since the home and work pressures have not been examined in previous research, an additional validity check is necessary. I use these home and work measures to construct an overall time pressure measure, in order to compare predictors of this overall measure to predictors of a known standard. To create this constructed measure of overall time pressure, a summed work and home pressure score was used for workers. For those who did not work, home pressure + 1 was used. This forces non-workers to ‘‘strongly disagree’’ that they currently experience time pressure in the workplace, and forces workers’ overall time pressure levels to be a sum of their reported home and work levels.1 These coding decisions were examined further by restricting the sample to workers in some of the analyses (keeping only those whose scores are based on self-report) to see if results changed, and by comparing results across the two data sources used in this study. 3.2.2. Predictors of time pressure Tables 1 and 2 show the descriptive statistics for all variables used in the analyses. Table 1 shows two samples for the GSS 2002, a full sample and a worker subsample, because the bulk of the analyses use the restricted work sample in order to compare findings across home and work pressure. Table 2 also shows two samples for the GSS 2004, a full sample and a worker subsample. Many independent variables are coded or re-coded identically in the GSS 2002 and 2004 data sources. Dummy variables indicate gender, whether one is currently married, and whether one has a college degree from a regular college or university. Hours worked per week, and age, are also included in the analyses. Children, typical mood, and income, are coded somewhat 1

In a regression analysis, this is the same as if the scores were averaged.

113

S. Kleiner / Social Science Research 47 (2014) 108–120 Table 2 Descriptive statistics, worker sample, GSS 2002 and 2004. Variable

a

GSS 2002

GSS 2004

Obs

Mean

Std. dev.

Min

Max

Obs

Mean

513

2.28

513 513 513 513 513 513 513 513

0.52 41.21 0.16 0.70 0.55 42.53 17.97 0.36

Time pressure measures Overall time pressure Pressure at home Pressure at work Stress indicators Stress at home Disagree about houseworka Stress at work Trouble concentrating at work

659 659 659

7.30 3.91 3.39

1.87 1.09 1.24

2 1 1

10 5 5

659 392 659 659

2.82 2.29 3.26 1.56

1.21 1.19 1.23 0.73

1 1 1 1

5 5 5 4

Independent variables Female Age Unhappy Kids Married Hours/wk Income College degree

659 659 659 659 659 659 659 659

0.53 40.08 0.04 0.33 0.56 41.44 2.87 0.27

0.50 12.42 0.19 0.47 0.50 13.98 1.03 0.44

0 18 0 0 0 1 1 0

1 74 1 1 1 89 4 1

Std. dev.

Min

Max

0.64

1

3

0.50 13.04 0.36 0.46 0.50 14.80 5.00 0.48

0 18 0 0 0 1 1 0

1 86 1 1 1 89 24 1

Married worker sample.

Table 3 OLS regressions predicting overall time pressure in the GSS 2002 and 2004. 2002

Female Age Unhappy/sad Kids Married Hours/wk Income College degree Constant Observations R-squared Adjusted R-squared

2004

Full sample

Worker sample

Full sample

Worker sample

0.169** (0.0512) 0.00215 (0.00163) 0.290* (0.122) 0.155** (0.0578) 0.00606 (0.0541) 0.0255*** (0.00122) 0.0508 (0.0290) 0.191** (0.0589) 0.780*** (0.108) 1044 0.390 0.384

0.177** (0.0685) 0.000764 (0.00281) 0.320 (0.172) 0.156* (0.0750) 0.00739 (0.0740) 0.0122*** (0.00248) 0.0486 (0.0419) 0.303*** (0.0777) 0.319 (0.176) 659 0.081 0.070

0.249*** (0.0672) 0.0118*** (0.00228) 0.0965 (0.0858) 0.369*** (0.0835) 0.115 (0.0754) 0.00986*** (0.00154) 0.0116 (0.0394) 0.267*** (0.0756) 0.295* (0.143) 818 0.158 0.150

0.326*** (0.0809) 0.00645 (0.00335) 0.119 (0.111) 0.380*** (0.0987) 0.0706 (0.0922) 0.0143*** (0.00274) 0.0112 (0.0540) 0.277** (0.0885) 0.759*** (0.192) 513 0.134 0.120

Standard errors in parentheses. Income is fully standardized. All other coefficients are Y-standardized. *** p < 0.001. ** p < 0.01. * p < 0.05.

differently in the two samples. The measure for children in the GSS 2002 (ISSP module) captures whether any child is living in the household, while in the GSS 2004 the variable codes for whether one has ever had a child. Hence, in the 2004 sample, many more respondents are coded as having children. The measure for being sad or unhappy is coded from a question about happiness versus unhappiness in 2002, and from a question about feeling sad in 2004.2 For these questions, a greater proportion of respondents responded that they typically felt sad (18%) compared to the proportion of respondents who reported

2 Sadness or unhappiness is included in the models for theoretical reasons, but the theoretical causality is bi-directional. Alternate models excluding these variables were run for all analyses, with no substantive difference in the findings.

114

S. Kleiner / Social Science Research 47 (2014) 108–120

rarely feeling happy (4%), and there was a higher refusal rate for the question about feeling sad as compared to the question about feeling happy. Family income was recoded into quartiles from the ISSP data, while in the GSS it is a categorical measure. 3.2.3. Stress indicators Finally, in the GSS 2002, additional measures are included to assess stress levels at work and at home. These are used to assess whether work and home pressures differ in their relationships with stress, both at work and home. For each domain, one psychological self-report and one behavioral self-report is used. At work, the two work stress questions used were ‘‘To what extent do you agree or disagree? My job is rarely stressful,’’ and ‘‘How often has each of the following happened to you during the past three months? I have found it difficult to concentrate at work because of my family responsibilities.’’ At home, the two questions used were ‘‘To what extent do you agree or disagree? My life at home is rarely stressful,’’ and ‘‘How often do you and your spouse/partner disagree about the sharing of household work?’’ While the psychological self-reports ask the respondent to estimate their current stress level at home or at work, the behavioral questions ask respondents to report on difficulty concentrating, or disagreements with a partner, both of which are affected by stress levels. Using two questions helps ensure that the conclusions drawn regarding time pressure and stress levels are not idiosyncratic to a particular measure. 3.3. Plan of analyses In the following analyses I address whether disaggregating time pressure into home and work domains might add to our understanding of the social processes that lead to feelings of time pressure. As a first step, I examine whether a combined measure of time pressure, using data on home and work pressures, yields predictors comparable to the often-used global time pressure measure capturing the extent to which one often feels ‘‘rushed.’’ This helps to establish whether time pressures experienced at home and at work are key components of an overall sense of time pressure. Second, I examine the predictors of home and work pressures separately. Because the disturbances from the two equations predicting home and work pressure are correlated (Breusch-Pagan test of independence rejected at p < .0001), as an additional check on the estimates I show small-sample statistics from Zellner’s seemingly unrelated regression to account for the shared error across the equations (Felmlee and Hargens, 1988; Zellner, 1962), with very similar results.3 Finally, using OLS, I examine how time pressure at home and work each relate to indicators of stress at home and at work. Alternate models, for all analyses, were calculated using ordinal logistic regression, with comparable results (see Appendix A for a subset of these supplemental analyses, showing all final models presented in the analytical tables). Together the final two analytic steps test for possible differences in the social roots and stress implications of subjective time pressure at home and work. Specifically, the analyses examine whether the predictors of home and work pressures differ, whether their associations with stress differ, and to what extent the associations with stress cross from one domain into another. 4. Results The purpose of the following analyses is to investigate whether time pressures at home and work differ. An initial examination of separate measures for time pressure at home and time pressure at work shows that while they are significantly correlated, the level of the correlation is low to moderate. This suggests that time pressures at home and at work may operate differently, and are driven by different phenomena. 4.1. Predictions of overall time pressure in each dataset In the first analytic step, I compare a constructed measure in the GSS 2002, composed of home and work pressures, to a known standard in the GSS 2004 (how often one feels ‘‘rushed’’). Table 3 shows OLS regression results for the social roles and resources thought to relate to feelings of time pressure. Looking across the findings for the constructed measure in the 2002 data, compared to the known standard in the 2004 data, the most notable pattern is that the predictors behave almost identically across the two datasets. Women, parents, the college educated, and those with longer weekly work schedules agree more strongly that they often feel rushed (2004) or that they have so many things to do they often run out of time before they get everything done (2002). Marital status and income did not predict time pressure in either dataset. There was some discrepancy between the two datasets for two of the eight variables: age, and feeling unhappy or sad. Age significantly predicted time pressure in 2004 for the full sample, though not the worker sample. In all models, however (including those using 2002 data), the coefficient for age was negative. Feeling fairly, very, or completely unhappy was significantly associated with time pressure in the full sample – though not the worker sample – in 2002, whereas feeling sad was not 3 In the presence of correlated equations, assuming homoscedastic residuals with a mean of zero, Zellner’s method can result in more efficient estimates (Felmlee and Hargens, 1988; Zellner, 1962). In practice, heteroscedasticity is frequently present in correlated equations, and SUR may still represent a gain in efficiency over OLS (Mandy and Martins-Filho, 1993). As this issue is not definitely resolved, due to heteroscedasticity concerns I estimated a separate model (not shown) using bootstrapping of the SUR estimates (2000 iterations), with comparable results.

115

S. Kleiner / Social Science Research 47 (2014) 108–120 Table 4 OLS and Seemingly Unrelated Regressions Predicting Time Pressure at Home and at Work, GSS 2002 . Time pressure at home

Female Age Unhappy Kids Married Hours/wk Income Degree

Time pressure at work

Full sample

Worker sample Model 1

Worker sample Model 2

Worker sample Model 3a

Worker sample Model 1

Worker sample Model 2

Worker sample Model 3a

0.261*** (0.0633) 0.00106 (0.00201) 0.353* (0.150) 0.296*** (0.0715) 0.0550 (0.0670) 0.00630*** (0.00151) 0.0375 (0.0314) 0.0359 (0.0728)

0.326*** (0.0738) 0.00133 (0.00302) 0.425* (0.185) 0.274*** (0.0808) 0.0815 (0.0797) 0.00725** (0.00268) 0.0200 (0.0395) 0.0866 (0.0838)

0.326*** (0.0709) 0.00202 (0.00290) 0.385* (0.178) 0.271*** (0.0776) 0.106 (0.0766) 0.00345 (0.00262) 0.00542 (0.0380) 0.0314 (0.0819) 0.272*** (0.0363)

0.326*** (0.0709) 0.00260 (0.00290) 0.351* (0.178) 0.268*** (0.0776) 0.126 (0.0766) 0.000216 (0.00261) 0.00703 (0.0380) 0.132 (0.0818) 0.503*** (0.0349)

0.00166 (0.0767) 0.00253 (0.00314) 0.148 (0.192) 0.0111 (0.0840) 0.0883 (0.0828) 0.0140*** (0.00278) 0.0538 (0.0411) 0.434*** (0.0870)

0.0971 (0.0747) 0.00292 (0.00302) 0.0236 (0.185) 0.0692 (0.0813) 0.112 (0.0796) 0.0119*** (0.00268) 0.0479 (0.0394) 0.409*** (0.0836)

0.179* (0.0747) 0.00326 (0.00302) 0.0827 (0.185) 0.138 (0.0813) 0.133 (0.0796) 0.0100*** (0.00268) 0.0429 (0.0394) 0.387*** (0.0836)

1.009*** (0.197) 659 0.096

0.293*** (0.0391) 0.874*** (0.190) 659 0.168

0.543*** (0.0376) 0.759*** (0.190) 659 0.116

Work pressure Home pressure Constant Observations R-squareda

0.453** (0.138) 1044 0.066

0.461* (0.189) 659 0.065

0.187 (0.186) 659 0.14

0.0467 (0.185) 659 0.086

Standard errors in parentheses. Where home and work pressures appear as predictors, the coefficients are fully standardized. All other regression coefficients are Y-standardized. *** p < 0.001. ** p < 0.01. * p < 0.05. a Model 3 estimated using Zellner’s seemingly unrelated regression.

associated with time pressure in 2004. This difference may be due to the wording differences across 2002 and 2004 for the independent variables capturing sadness or unhappiness. Overall, the similarities across the two datasets, and only minor differences, suggest that the time pressure measures used in 2002 and 2004 capture similar types of respondents. Importantly, this suggests that the measures for home and work pressures represent a disaggregation of overall time pressures. 4.2. Predictors of time pressure at home and at work To examine how home and work pressures differ, the next set of analyses examines their predictors. Table 4 presents the OLS and seemingly unrelated regression coefficients for a set of models predicting subjective time pressure across the two domains of home and work. The first four columns show the findings for time pressure at home, first using the full sample, then the worker sample Model 2 for the worker sample adds a control for work pressure. The purpose of this control is not to examine causal effects of time pressure at home and work on each other, but to account for the underlying relationship between work and home pressure when examining the other predictors. Model 3 uses seemingly unrelated regression, producing estimates that account for the error shared with Model 3 in the second set of columns. The second set of columns show the findings for time pressure at work within the worker sample, with model 2 including a control for time pressure at home, and the final model showing results for the seemingly unrelated regression. The analyses shown in Table 4 show a striking difference between the predictors of time pressure at home and the predictors of time pressure at work. Women, those who are unhappy, and parents, are more pressed for time at home, but not at work. In contrast, those who work longer hours, or hold a college degree, are more likely to feel pressed for time at work, but not at home. The only similarity across the two domains is for weekly work hours; those who work long hours are pressed for time at home as well as work, but the relationship between long work hours and time pressure at home is accounted for by controlling for time pressure at work. Overall, the two types of time pressure appear to have distinct predictors, with minimal overlap. 4.3. Time pressure at home and work as predictors of stress at home and work The final analyses examine whether the two types of time pressure, time pressure at home and time pressure at work, have not only distinct predictors, but distinct relationships with stress. Table 5 presents the results for a series of regressions

116

S. Kleiner / Social Science Research 47 (2014) 108–120

Table 5 Time pressure at home and at work as predictors of home stress. Predicting home stress Full sample Home pressure Home pressure

Female Age Unhappy Kids Married Hours/wk Income College degree Constant Observations R-squared

Worker sample Both pressures

Full sample Home pressure 0.153*** (0.0418)

0.0223 (0.0382) 0.176* (0.0748) 0.0104*** (0.00306) 0.588** (0.188) 0.337*** (0.0819) 0.0704 (0.0808) 0.000716 (0.00276) 0.0728 (0.0401) 0.0978 (0.0864) 0.0730 (0.196) 659 0.091

0.157*** (0.0410) 0.0650 (0.0394) 0.125 (0.0752) 0.0101*** (0.00303) 0.527** (0.186) 0.294*** (0.0818) 0.0870 (0.0801) 0.000173 (0.00274) 0.0719 (0.0397) 0.0929 (0.0855) 0.0436 (0.194) 659 0.111

0.120*** (0.0304)

Work pressure *

0.155 (0.0624) 0.00592** (0.00197) 0.626*** (0.147) 0.357*** (0.0703) 0.0832 (0.0654) 0.00247 (0.00149) 0.0194 (0.0307) 0.158* (0.0711) 0.0473 (0.136) 1044 0.110

Predicting housework disagreements (married respondents only)

Worker sample Work pressure

0.137 (0.0853) 0.0160*** (0.00299) 0.285 (0.237) 0.0711 (0.0893) 0.0963 (0.143) 0.000151 (0.00204) 0.0384 (0.0412) 0.00577 (0.0955) 0.786*** (0.219) 601 0.104

Worker sample Work pressure

Worker sample Both pressures

0.0294 (0.0537) 0.00911 (0.106) 0.0214*** (0.00461) 0.326 (0.275) 0.140 (0.105) 0.00500 (0.181) 0.000394 (0.00389) 0.0111 (0.0574) 0.0980 (0.119) 0.847** (0.312) 392 0.081

0.132* (0.0563) 0.00342 (0.0552) 0.0352 (0.107) 0.0208*** (0.00459) 0.238 (0.276) 0.101 (0.106) 0.00914 (0.180) 0.000346 (0.00388) 0.0111 (0.0571) 0.108 (0.118) 0.883** (0.311) 392 0.094

Standard errors in parentheses. Home and work pressures are fully standardized. All other regression coefficients are Y-standardized. *** p < 0.001. ** p < 0.01. * p < 0.05.

that model time pressure as a predictor of home stress, controlling for likely confounders. Table 6 shows the results for time pressure as a predictor of work stress. The models presented in Table 5 show time pressure as a predictor of two indicators of stress at home. These stress indicators include a measure of self-rated stress at home, and, among partnered respondents only, a report of how often the respondent argues with their spouse or partner about housework. Each dependent variable is shown as an outcome of home pressure (full sample), work pressure (worker sample), or both (worker sample). For both home stress and disagreements about housework, time pressure at home is consistently a significant predictor, whereas time pressure at work is not a significant predictor in any of the models (and the coefficient for work pressure is quite close to zero). Table 6 shows time pressure at home and work as a predictor of two indicators of stress in the workplace environment, the first a self-report of stress at work, and the second a self-rating of how often the respondent has trouble concentrating at work. As in Table 5, the present set of models tests each type of time pressure, as well as both simultaneously, as potential predictors of each dependent variable. The results show that both time pressure at work and at home are significant predictors of self-rated work stress or trouble concentrating at work. In contrast to Table 5, in which both types of time pressure were modeled as a predictor of stress within the home environment, but only home pressure predicted home stress, here in Table 6 both types of time pressure, at home as well as at work, are significantly associated with stress within the work environment. 5. Discussion and conclusions Subjective time pressure is of theoretical interest in part because it is conceptualized as a form of role strain, linking specific social roles with the challenges, struggles, and strains that individuals work to overcome. It is also of importance because chronic levels of such strains may potentially lead to stress and negative emotions. Existing research, however, has not fully examined subjective time pressure as a form of role strain. Chronic levels of subjective time pressure are generally conceptualized as pervasive and constant, rather than examined within the particular social contexts that produce the time pressure. Furthermore, very little research explicitly examines links between subjective time pressure and stress. The goal of this study was to disaggregate time pressure into the home and work contexts in which they are experienced. In particular, I examined whether these types of pressure have different predictors, as well as different relationships with stress. I also compared the home and work-specific time pressure measures with a more widely used measure from the

117

S. Kleiner / Social Science Research 47 (2014) 108–120 Table 6 Time pressure at work and at home as predictors of work stress (worker sample). Predicting work stress Work pressure Work pressure

Female Age Unhappy Kids Married Hours/wk Income Degree Constant Observations R-squared

Work and home pressure

Work pressure 0.131** (0.0399)

0.201*** (0.0403) 0.0944 (0.0770) 0.00213 (0.00311) 0.0242 (0.191) 0.0372 (0.0838) 0.0642 (0.0820) 0.0156*** (0.00277) 0.0606 (0.0406) 0.209* (0.0861) 0.907*** (0.196) 659 0.117

0.307*** (0.0386) 0.111** (0.0401) 0.124 (0.0737) 0.00302 (0.00297) 0.0314 (0.182) 0.0585 (0.0801) 0.0297 (0.0784) 0.0120*** (0.00268) 0.0459 (0.0389) 0.0831 (0.0838) 0.639*** (0.190) 659 0.195

0.337*** (0.0373)

Home pressure 0.160* (0.0729) 0.00325 (0.00298) 0.0112 (0.183) 0.0885 (0.0798) 0.0180 (0.0787) 0.0123*** (0.00269) 0.0465 (0.0391) 0.0796 (0.0842) 0.660*** (0.191) 659 0.186

Predicting trouble concentrating at work

Home pressure

0.157* (0.0780) 0.00677* (0.00319) 0.938*** (0.196) 0.0717 (0.0854) 0.0353 (0.0842) 0.000115 (0.00288) 0.0390 (0.0418) 0.0479 (0.0901) 0.0146 (0.204) 659 0.068

Home pressure

Work and home pressure

0.139*** (0.0414) 0.111 (0.0791) 0.00625 (0.00319) 0.898*** (0.196) 0.0350 (0.0861) 0.0582 (0.0842) 0.000698 (0.00284) 0.0432 (0.0417) 0.00332 (0.0885) 0.0528 (0.201) 659 0.069

0.101* (0.0414) 0.110* (0.0430) 0.121 (0.0789) 0.00655* (0.00318) 0.896*** (0.195) 0.0419 (0.0858) 0.0469 (0.0840) 0.000495 (0.00287) 0.0384 (0.0416) 0.0445 (0.0897) 0.0352 (0.203) 659 0.077

Standard errors in parentheses. Home and work pressures are fully standardized. All other regression coefficients are Y-standardized. *** p < 0.001. ** p < 0.01. * p < 0.05.

GSS 2004, but also found in other surveys and published research. This study yielded several findings with important implications for research on time pressure, work and family roles, and social stress. First, time pressures at home and at work appear to be distinct types of time pressure, with distinct predictors. Different social roles predicted the two types of pressure, with women and parents reporting more pressure at home, and the college educated and those who work longer hours reporting more time pressure at work. These social roles or statuses (gender, family, education, and work) were highly domain specific in their associations with time pressure. While men and non-parents were advantaged at home, they were not disadvantaged at work. Similarly, working lower hours, or lack of a college education, meant less time pressure at work, but did not confer any advantages at home. Second, time pressures at home and at work behave differently in their associations with stress. While time pressure at home predicted stress at home and at work, time pressure at work was only related to stress at work. This is a surprising finding, because numerous facets of employment affect stress and health. This study, however, finds no link between time pressure within the work environment and stress outside of it. It may be that time pressure at work is more common in high status jobs, and workers in these types of jobs may extend their work hours in response to time pressures, helping to confine time pressure to the work environment. It is also likely that these individuals have a higher quality of life at home, such as a safe, unpolluted neighborhood, a comfortable home, and lower debt, as a result of their high status jobs, all of which would contribute to lower stress at home. In contrast, low status jobs, though less time pressured, may expose workers to a host of other work stressors, such as low job control, occupational hazards, and boring, repetitive work. The finding that time pressure at home is associated with stress both at home and at work is intriguing. Some work/family researchers have made a distinction between time-based, strain-based, and boundary spanning role demands (Voydanoff, 2005). The present study, however, suggests that subjective time pressure is not only strain-based, but also time-based, as it constitutes pressure created by lack of time. Furthermore, it finds that time pressure can also be boundary-spanning, with the potential to affect stress levels across both home and work domains, depending on the source of the time pressure. There are several likely explanations for the finding that time pressure at home, in particular, is related to stress both at home and at work. One is that time pressure at home is more often created by tasks that are not only stressful, but less rewarding. Another is that home is where recuperation from stress takes place, and for this reason time pressure within the home environment may be particularly influential on stress levels. Future studies could further address these issues by examining the nature of time pressure-related stress in home and work environments, the extent to which it is psychologically rewarding versus emotionally draining, and the extent to which time pressure at home and at work is associated with changes in stress at home and work over time and over the life course.

118

S. Kleiner / Social Science Research 47 (2014) 108–120

Finally, in order to infer whether the home and work pressure measures used in this study are likely similar to disaggregating the commonly used time pressure measure of feeling ‘‘rushed,’’ the study compared a constructed measure of overall time pressure with the more widely used measure in the GSS. Results suggest that combining home and work pressures yields the same predictors as feeling ‘‘rushed.’’ This suggests that the two types of pressure do represent a disaggregation of overall time pressure into home and work domains, and that the two types likely represent the major sources of time pressure individuals consider when reporting on their overall levels of pressure. An additional source of time pressure not represented in this research, however, is the space between home and work domains, particularly commuting between the two. Assessing and integrating the role of commuting time pressures in time pressure research would add an important piece to this issue. There are several important limitations to the present study. First, the data are cross-sectional. Although time pressure and stress are theorized to co-occur, self-reports of stress or stressful experiences may reflect responses to life circumstances that have accrued over time. Longitudinal research would provide a critical next step in assessing relationships between chronic time pressure and stress levels. Furthermore, alternate measures of stress, beyond self-report, would add important dimensions to this area of study, and might particularly shed light on the health implications of time pressure. Another limitation to this study, highlighted by recent research in this journal, is the greater tendency for busy individuals to decline to participate in surveys (Vercruyssen et al., 2011). We can expect, therefore, that highly time pressured individuals might be under-sampled in this study. Future data collection projects should employ special effort to recruit time pressured individuals, and research should particularly attend to the intersection of time pressures across domains. Overall, this study demonstrates advantages to looking at time pressure separately according to life domain, and suggests that previous findings on general time pressure could be expanded by the use of more fine-grained questions tapping into work versus home contexts of time pressures. Future surveys encompassing subjective time pressure should consider including such domain-specific questions, particularly to further address the unique relationships with well-being found in the present study. The domain-specific measures examined here, when combined, yielded findings quite similar to the overall measure of feeling ‘‘rushed,’’ but when disaggregated provided new insights into the sources and stresses of perceived time pressure. Acknowledgments I am grateful to Joshua Klugman for his numerous insightful comments on this manuscript, and to Nick Vargas for his helpful feedback. I also thank Patricia McManus and Eliza Pavalko for their useful comments on an earlier draft. This material is based upon work supported by the National Science Foundation under Grant No. 1002635. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation (NSF). Appendix A Ordinal logistic regression coefficients for all final models (worker samples). Overall time pressure (Table 3)

Domain-specific pressure (Table 4)

Predicting stress at home and work (Tables 5 and 6)

Overall time Overall pressure time (constructed) pressure (rushed)

Time pressure at home

Time pressure at work

Home stress

Household disagreements (married respondents)

Work stress

Trouble concentrating at work

0.777*** (0.151) 0.00444 (0.00613) 1.078* (0.422) 0.554*** (0.166) 0.300 (0.162) 0.00768 (0.00564)

0.151 (0.148) 0.00635 (0.00599) 0.0421 (0.360) 0.163 (0.161) 0.219 (0.159) 0.0262*** (0.00558)

0.249 (0.148) 0.0193** (0.00595) 1.211** (0.387) 0.594*** (0.161) 0.201 (0.156) 0.000160 (0.00528)

0.0229 (0.204) 0.0409*** (0.00908) 0.341 (0.566) 0.181 (0.203) 0.146 (0.373) 0.000802 (0.00757)

0.242 (0.149) 0.00713 (0.00604) 0.0967 (0.365) 0.0772 (0.163) 0.0364 (0.160) 0.0238*** (0.00545)

0.225 (0.162) 0.0109 (0.00670) 1.642*** (0.395) 0.0408 (0.176) 0.154 (0.174) 7.23  10 (0.00588)

0.411** (0.143) Age 0.00430 (0.00585) Unhappy/sad 0.680 (0.353) Kids 0.330* (0.154) Married 0.0182 (0.153) Hours/wk 0.0286*** (0.00529) Female

0.728*** (0.184) 0.0135 (0.00761) 0.295 (0.250) 0.894*** (0.224) 0.0806 (0.208) 0.0331*** (0.00644)

6

119

S. Kleiner / Social Science Research 47 (2014) 108–120

Appendix A (continued)

Income College degree

Overall time pressure (Table 3)

Domain-specific pressure (Table 4)

Predicting stress at home and work (Tables 5 and 6)

Overall time Overall pressure time (constructed) pressure (rushed)

Time pressure at home

Time pressure at work

Home stress

Household disagreements (married respondents)

Work stress

Trouble concentrating at work

0.100 (0.0867) 0.676***

0.0157 (0.124) 0.589**

0.0242 (0.0794) 0.112

0.0596 (0.0785) 0.783***

0.156* (0.0772) 0.177

0.0895 (0.113) 0.0431

0.0883 (0.0795) 0.155

0.0677 (0.0860) 0.0120

(0.162)

(0.199)

(0.174) 0.631***

(0.168)

(0.168) 0.143

(0.216) 0.0404

(0.170) 0.665***

(0.182) 0.204*

0.646***

(0.0809) 0.354***

(0.107) 0.235*

(0.0848) 0.233**

(0.0876) 0.241**

(0.0859) 659 .0646

(0.0840) 659 .0419

(0.108) 392 .0310

(0.0841) 659 .0762

(0.0920) 659 .0351

Work pressure

(0.0839) Home pressure Observations 659 R-squareda .0251

513 .0754

659 .0637

Standard errors in parentheses. Home and work pressures are fully standardized. Income is fully standardized in columns 1 and 2. All other regression coefficients are Y-standardized. *** p < 0.001. ** p < 0.01. * p < 0.05. a The R-squared term reported is McKelvey and Zavoina’s R-squared (see Long and Freese, 2006).

References Baer, Markus, Oldham, Greg R., 2006. The curvilinear relation between experienced creative time pressure and creativity: moderating effects of openness to experience and support for creativity. J. Appl. Psychol. 91 (4), 963–970. Becker, Gary S., 1965. A theory of the allocation of time. Econ. J. 75 (299), 493–517. Bschor, T., Ising, M., Bauer, M., Lewitzka, U., Skerstupeit, M., Müller-Oerlinghausen, B., Baethge, C., 2004. Time experience and time judgment in major depression, mania and healthy subjects. A controlled study of 93 subjects. Acta Psychiatr. Scand. 109, 222–229. Clark, Sue Campbell, 2000. Work/family border theory: a new theory of work/family balance. Hum. Relat. 53, 747. Clarkberg, Marin, Moen, Phyllis, 2001. Understanding the time squeeze: married couples’ preferred and actual work-hour strategies. Am. Behav. Sci. 44 (7), 1115–1135. Coser, Lewis A., 1974. Greedy Institutions: Patterns of Undivided Commitment. The Free Press, New York, NY. Csikszentmihalyi, Mihalyi, 1997. Finding Flow: The Psychology of Engagement with Everyday Life. HarperCollins, New York. Dapkus, Marilyn A., 1985. A thematic analysis of the experience of time. J. Pers. Soc. Psychol. 49 (2), 408–419. Davis, James Allan, Smith, Tom W., Marsden, Peter V., 2007. General Social Surveys, 1972–2006: Cumulative Codebook. Principal Investigator, James A. Davis; Director and Co-Principal Investigator, Tom W. Smith. National Opinion Research Center, Chicago. DeDonno, Michael A., Demaree, Heath A., 2008. Perceived time pressure and the Iowa gambling task. Judgment Decis. Making 3 (8), 636–640. DeVoe, Sanford E., Pfeffer, Jeffrey, 2011. Time is tight: how higher economic value of time increases feelings of time pressure. J. Appl. Psychol. 96 (4), 665– 676. Eby, Lillian T., Maher, Charleen P., Butts, Marcus M., 2010. The intersection of work and family life: the role of affect. Annu. Rev. Psychol. 61, 599–622. Felmlee, Diane H., Hargens, Lowell L., 1988. Estimation and hypothesis testing for seemingly unrelated regressions: a sociological application. Soc. Sci. Res. 17, 384–399. Flaherty, Michael G., 1991. The perception of time and situated engrossment. Soc. Psychol. Quart. 54, 76–85. Flaherty, Michael G., Meer, Michelle D., 1994. How time flies: age, memory, and temporal compression. Sociol. Quart. 35 (4), 705–721. Gieryn, Thomas F., 2000. A space for place in sociology. Annu. Rev. Sociol. 26, 463–496. Gimenez-Nadal, Jose Ignacio, Sevilla-Sanz, Almudena, 2011. The time crunch paradox. Soc. Indic. Res. Gleick, James, 1999. Faster: The Acceleration of Just About Everything. Pantheon Books, New York. Goode, William J., 1960. A theory of role strain. Am. Sociol. Rev. 25 (4), 483–496. Gross, Barbara L., Sheth, Jagdish N., 1989. Time-oriented advertising: a content analysis of United States magazine advertising, 1890–1988. J. Market. 53 (October), 76–83. Hamermesh, Daniel C., Lee, Jungmin, 2007. Stressed out on four continents: time crunch or Yuppie Kvetch? Rev. Econ. Stat. 89 (2), 374–383. Hilbrecht, Margo, Zuzanek, Jiri, Mannell, Roger C., 2008. Time use, time pressure and gendered behavior in early and late adolescence. Sex Roles 58, 342– 357. Hochschild, Arlie Russell, 1997. The Time Bind: When Work Becomes Home and Home Becomes Work. Metropolitan Books, New York. Kangas, Nocole L., Meyerson, Debra E., 2008. The gendering of emotions and perceived work time: chicks and geeks at I.com. In: Fineman, Stephen (Ed.), The Emotional Organization: Passions and Power. Blackwell Publishing, Oxford, UK, pp. 102–120. Larson, Reed W., Richards, Maryse H., Perry-Jenkins, Maureen, 1994. Divergent worlds: the daily emotional experience of mothers and fathers in the domestic and public spheres. J. Pers. Soc. Psychol. 67 (6), 1034–1046. Levine, Robert, 1997. A Geography of Time: The Temporal Misadventures of a Social Psychologist. Basic Books, New York. Link, Bruce G., Phelan, Jo, 1995. Social conditions as fundamental causes of disease. J. Health Soc. Behav. 35, 80–94. Lois, Jennifer, 2010. The temporal work of motherhood: homeschoolers’ strategies for managing time shortage. Gender Soc 24, 421–446.

120

S. Kleiner / Social Science Research 47 (2014) 108–120

Long, J. Scott, Freese, Jeremy, 2006. Regression Models for Categorical Dependent Variables Using Stata. Stata Press, College Station, TX. Mandy, David M., Martins-Filho, Carlos, 1993. Seemingly unrelated regressions under additive heteroscedasticity: theory and share equation applications. J. Econom. 58, 315–346. Mattingly, Marybeth J., Bianchi, Suzanne M., 2003. Gender differences in the quantity and quality of free time: the U.S. experience. Soc. Forces 81 (3), 999– 1030. Mattingly, Marybeth J., Sayer, Liana C., 2006. Under pressure: gender differences in the relationship between free time and feeling rushed. J. Marriage Fam. 68 (February), 205–221. Maule, A. John, Hockey, G.Robert J., 1993. State, stress, and time pressure. In: Svenson, Ola, John Maule, A. (Eds.), Time Pressure and Stress in Human Judgment and Decision Making. Plenum Press, New York, NY, pp. 83–102. McLeod, Jane D., Lively, Kathryn J., 2006. Social structure and personality. In: Delamater, John (Ed.), Handbook of Social Psychology. Springer, New York, NY, pp. 77–102. Meck, Warren H., 2005. Neuropsychology of timing and time perception. Brain Cognition 58 (1), 1–8. Offer, Shira, Schneider, Barbara, 2011. Revisiting the gender gap in time-use patterns: Multitasking and well-being among mothers and fathers in dualearner families. Am. Sociol. Rev. 76 (6), 809–833. Pearlin, Leonard I., Schieman, Scott, Fazio, Elena M., Meersman, Stephen C., 2005. Stress, health, and the life course: some conceptual perspectives. J. Health Soc. Behav. 46 (June), 205–219. Robinson, John P., Godbey, Geoffrey, 1997. Time for Life: The Surprising Ways Americans Use Their Time. Pennsylvania State University Press. Roxburgh, Susan, 2002. Racing through life: the distribution of time pressures by roles and role resources among full-time workers. J. Fam. Econ. Issues 23 (2), 121–145. Roxburgh, Susan, 2004. ‘There Just Aren’t Enough Hours in the Day’: the mental health consequences of time pressure. J. Health Soc. Behav. 45 (2), 115–131. Sayer, Liana C., 2005. Gender, time and inequality: trends in women’s and men’s paid work, unpaid work and free time. Soc. Forces 84 (1), 285–303. Schieman, Scott, Whitestone, Yuko Kurashina, Van Gundy, Karen, 2006. The nature of work and the stress of higher status. J. Health Soc. Behav. 47, 242–257. Schor, Juliet B., 2004. Born to Buy: The Commercialized Child and the New Consumer Culture. Scribner, New York. Sorokin, Pitirim A., Merton, Robert K., 1937. Social time: a methodological and functional analysis. Am. J. Sociol. 42 (5), 615–629. Szollos, Alex, 2009. Toward a psychology of chronic time pressure. Time Soc. 18 (2/3), 332–350. Teuchmann, Katja, Totterdell, Peter, Parker, Sharon K., 1999. Rushed, unhappy, and drained: an experience sampling study of relations between time pressure, perceived control, mood, and emotional exhaustion in a group of accountants. J. Occup. Health Psychol. 4 (1), 37–54. Tézli, Annette, Gauthier, Anne H., 2009. Balancing work and family in Canada: an empirical examination of conceptualizations and measurements. Can. J. Sociol. 34 (2), 433–462. Thoits, Peggy A., 1995. Stress, coping, and social support processes: where are we? What next?’’. J. Health Soc. Behav. (Extra Issue), 53–79. Van Emmerik, I.J. Hetty, Jawahar, I.M., 2006. The independent relationships of objective and subjective workload with couples’ mood. Hum. Relat. 59 (10), 1371–1392. Vercruyssen, Anina, Roose, Henk, Van de Putte, Bart, 2011. Underestimating busyness: indications of nonresponse bias due to work-family conflict and time pressure. Soc. Sci. Res. 40, 1691–1701. Voydanoff, Patricia, 2005. Work demands and work-to-family and family-to-work conflict: direct and indirect relationships. J. Fam. Issues 26 (6), 707–726. Zellner, Arnold, 1962. An efficient method of estimating seemingly unrelated regression and tests for aggregation bias. J. Am. Stat. Assoc. 57 (298), 348–368. Zerubavel, Eviatar, 1981. Hidden Rythms: Schedules and Calendars in Social Life. University of Chicago Press, Chicago and London. Zuzanek, Jiri, 1998. Time use, time pressure, personal stress, mental health, and life satisfaction from a life cycle perspective. Journal of Occupational Science 5 (1), 26–39.