Exploring the relationship of workplace flexibility, gender, and life stage to family-to-work conflict, and stress and burnout

Exploring the relationship of workplace flexibility, gender, and life stage tofamily-to-work conflict, and stress and burnout

E. Jeffrey Hilla*, Jenet I. Jacoba, Laurie L. Shannonb, Robert T. Brennanc,Victoria L. Blancharda, and Giuseppe Martinengoa

aBrigham Young University, Provo, UT, USA; bWFD Consulting, Newton, MA, USA; cHarvardMedical School, Boston, MA, USA

This study explores how the reported use and perceived value of five workplaceflexibility options differ by gender and life stage. It also examines the relationship ofperceived workplace flexibility, gender, and life stages to family-to-work conflict andstress and burnout. Data are from a multi-company database created by WFDConsulting containing the results of studies conducted in multiple organizationsbetween 1996 and 2006. Results revealed a curvilinear relationship in differencesbetween men and women in the use of workplace flexibility options over the life course.Hierarchical linear modeling (HLM) revealed that perceived flexibility consistentlypredicted less family-to-work conflict and less stress and burnout. Gender (being female)was also modestly associated with greater family-to-work conflict and greater stress andburnout. The addition of life stage variables significantly increased the proportion of thevariance explained in family-to-work conflict and stress and burnout. The addition ofthe interaction of life stage and gender significantly increased the proportion of thevariance explained in family-to-work conflict but not in stress and burnout. Theseresults support the idea that gender, life stage, and their interactions are importantvariables in research related to workplace flexibility.

Keywords: family-to-work conflict; gender; life stages; stress and burnout; workplaceflexibility

Este estudio explora como el uso informado y valor percibido de cinco opciones de laflexibilidad de lugar de trabajo difieren por la etapa del genero y la vida. Examinatambien la relacion de la flexibilidad percibida de lugar de trabajo, del genero, y de lasetapas de la vida al conflicto de la familia a trabajo y el enfasis y a fundicion. Los datosson de una base de datos de la multi-companıa creada por Consultar de WFD quecontiene los resultados de estudios realizados en multiples organizaciones entre 1996 y2006. Los resultados revelaron una relacion curvilınea en diferencias entre hombres ymujeres en el uso de opciones de flexibilidad de lugar de trabajo sobre el curso de la vida.Jerarquico lineal modelando (HLM) revelo que la flexibilidad percibida predijocoherentemente menos conflicto de familia a trabajo y menos enfasis y fundicion. Elgenero (es femenino) tambien fue asociado modestamente con el conflicto mas grande dela familia a trabajo y el enfasis y con fundicion mas grande. La adicion de la vida lasvariables del teatro aumentaron apreciablemente la proporcion de la variacion explicadaen el conflicto de la familia a trabajo y el enfasis y en fundicion. La adicion de lainteraccion de la etapa de la vida y el genero aumento apreciablemente la proporcion de lavariacion explicada en el conflicto de la familia a trabajo pero no en enfasis y fundicion.Estos resultados sostienen la idea ese genero, la etapa de la vida, y sus interacciones sonvariables importantes en la investigacion relacionada a la flexibilidad del lugar de trabajo.

*Corresponding author. Email: [email protected]

Community, Work & Family,

Vol. 11, No. 2, May 2008, 165�181

ISSN 1366-8803 print/ISSN 1469-3615

# 2008 Taylor & Francis

DOI: 10.1080/13668800802027564

http://www.informaworld.com

Palabras claves: la familia a trabajo conflicto; el genero; las etapas de la vida; stress andburnout; la flexibilidad de lugar de trabajo

Introduction

The suite of work�life policies now considered core to workplace flexibility (e.g., flextime,

compressed work week, telecommuting, part-time employment, job sharing) was originally

targeted at mothers with small children. Thus, flexible work arrangements were generally

called work�family programs. Coupled with strategies for child care, they were designed to

ameliorate mothers’ stress and burnout in order to retain women in the workforce and

reduce women’s family-to-work conflict to enable them to be more effective on the job.

Research generally supports that such arrangements facilitate these goals. Mothers

consistently report that they are less stressed, more likely to stay with their employer

after childbirth, and more satisfied on the job if flexible work arrangements are available

(Hill et al., 2006; Jacob, Bond, Galinsky, & Hill, 2008).

However, since the 1980s there has been a growing recognition that all employees, not

just mothers of young children, have lives outside of work. Workplace flexibility is now

seen as an integral work�life strategy to meet changing personal needs of both men and

women during all life stages (Hill, Jackson, & Martinengo, 2006). A young single person

might use flextime to train for a triathlon and value it because it dissipates the stress and

burnout associated with a high-pressure early career job. A middle-aged father might

telecommute in order to accept a promotion to headquarters without having to relocate his

family while his teenagers are in high school, thus circumventing significant family-to-work

conflict. A young mother and a mature-age woman might job share so that the one might

have added time to care for preschool children and the other might have the flexibility to

care for aging parents with less impact to their jobs. A grandmother in her mid-50s might

opt for a compressed work week so that she would have Friday afternoons to be renewed

by coaching her granddaughter’s soccer team. Finally, an executive in his late 60s might

choose part-time employment because he is financially independent and wants the

continued stimulation of employment along with time for his community and grand-

children. Understanding the ways in which the use and value of workplace flexibility

options vary by gender and life stage, and how the latter (gender and life stage) are related

to family-to-work conflict and stress and burnout, would facilitate the development of

effective programs, policies, and work�life strategies to meet employee and organizational

needs.

The first objective of this study is to explore how the reported use and perceived value

of workplace flexibility options differ by gender during five different life stages using

descriptive statistics. The second objective is to explore how gender and life stage are

related to perceived flexibility, family-to-work conflict, and stress and burnout.

Related research

Definitions

We define workplace flexibility as ‘the ability of workers to make choices influencing when,

where, and for how long they engage in work-related tasks’ (Hill et al., 2008, p. 152). In this

paper we will examine five specific types of workplace flexibility as measured in the WFD

database: flextime, compressed work week, telecommuting, part-time employment, and job

sharing (Civian, Richman, Shannon, Shulkin, & Brennan, 2008). Flextime is defined as a

166 E. J. Hill et al.

work schedule that enables employees to have flexibility in determining, within certain

limits, when their regular workday begins and ends. The total hours worked each day

remain the same. Compressed work week is defined as a work schedule that compresses the

standard eight-hour day, five-day work week into fewer, longer days, typically 40 hours

worked in four days or sometimes 80 hours spread out over nine days. Telecommuting is

defined as an arrangement in which an employee works at a location other than the normal

worksite such as at home or a satellite office. This may occur on a regular basis or on an as-

needed, occasional basis. Part-time employment is defined as working less than a company’s

standard full time. Job sharing refers to two employees voluntarily sharing the

responsibilities of one full-time job, while salary and benefits are pro-rated, for example,

a 50/50 or 60/40 split, or 60/60 when one day of overlap occurs.

Theoretical perspective

A life course, role context theoretical approach informs this study through its emphasis onthe dynamic interplay of the contexts of gender and family life stage in the experience of

the work�family interface. Changes in family needs, resources, vulnerabilities, and shifts in

family roles, relationships, responsibilities, and circumstances are posited to vary by age

and gender as well as life stage (Moen & Sweet, 2004). This perspective shifts the

discussion of the work�family interface from individuals to a more complex view of career

development in evolving family life stage contexts (Moen & Sweet, 2004). This research

anchors life course stages to the presence and ages of children because of the shifts in roles,

relationships, and responsibilities associated with parenthood (Mattessich & Hill, 1987).

Age of the employee’s youngest child and employee age are used as the markers of life stage

to account for potential changes in the experience of the work�family interface across the

life course.

Use of workplace flexibility for women and men at different life stages

Studies have identified several differences in the use of flexible work options for men and

women at different life stages. There has been an increase in both access to and use of

flexible work arrangements by women and men in the last two decades (Bond, Galinsky,

Kim, & Brownfield, 2005; Bond, Thompson, Galinsky, & Prottas, 2002). When compared

to men in the same company, however, women were more likely to have used flexible work

arrangements (Hill, Martinson, Hawkins, & Ferris, 2003).

Women and men also differed in the type of flexibility they used. The rate of women’s

part-time employment was estimated to be triple that of men’s (Bond et al., 2002; Comfort,

Johnson, & Wallace, 2003), and women were more likely to use flextime, although these

gender differences were much smaller (Galinsky, Bond, & Hill, 2004). According to a

recent Census Report (US Census Bureau, 2004), men and women were comparable in

their use of telecommuting at least one day a week. But when comparing parents in the

same company, 63% of mothers used telecommuting one day per week, compared to just

48% of fathers (Hill et al., 2003).

Life stage differences were also apparent. Married men and women were significantly

more likely than unmarried workers to use flexible work arrangements (Golden, 2001),

although differences were small. Greater differences were found when comparing thosewith and without young children. Having a child less than six years of age significantly

increased the odds of working a flextime schedule for both men and women (Billings &

Sharpe, 1999). Women ages 25�34 years were more likely than women of other ages to have

Community, Work & Family 167

flexible work arrangements (DiNatale & Boraas, 2002), although use of part-time

employment peaked in the prime childrearing years for women ages 35�44 years (Comfort

et al., 2003, p. 4). Younger women (under 40 years of age) without children were least likely

to differ from their male partners in the use of flexible work arrangements.

Value of workplace flexibility for women and men at different life stages

Findings from the few studies comparing the value men and women place on different

flexible work options suggested that both women and men value flexibility (RadcliffePublic Policy Center with Harris Interactive, 2000), but women were more likely to highly

value flexible work options (Catalyst, 2001). Married women with young children were

most likely to be interested in part-time work compared to their male and female

counterparts, while unmarried, childless men were least likely to be interested (Wharton &

Blair-Loy, 2002).

The female trend for highly valuing flexible work options may be explained partly by

evidence that couples with children were more likely than those without children to

prioritize the man’s occupation, preferring reduced work hours for women (Bond et al.,2002). And prior to having children, female graduate students explained that they expected

more diverse employment pathways than their male counterparts because the priority of

their economic and occupational commitment remained secondary to their husbands’

(Hakim, 2002; Moen & Sweet, 2002).

Family-to-work conflict for men and women at different life stages

Conflict in the work�family interface has been defined as a type of inter-role stress thatresults from incompatible demands in the work and family domains (Greenhaus & Beutell,

1985). Because inter-role conflict may originate from either the work or family domain, a

distinction has been made between work-to-family conflict and family-to-work conflict

(Gutek, Searle, & Klepa, 1991). Family-to-work conflict occurs when participation in the

work role is made more difficult because of participation in the family role (Greenhaus &

Beutell, 1985).

Differences between men and women in family-to-work conflict have been the focus of

a substantial body of work�family interface research, although little is known aboutchanges by gender across the life course. Early theoretical discussions of the work�family

interface hypothesized that women would experience more family-to-work conflict than

men, while men would experience more work-to-family conflict than women (Pleck, 1977).

The division of household labor along with traditional ideology about men’s and women’s

responsibilities were identified as contributing factors (Dilworth, 2004). Crouter’s (1984)

early qualitative exploration of the effect of family on work life supported this hypothesis.

Employed mothers with young children were more likely to experience family-to-work

conflict than their male counterparts. Dilworth (2004) also found that mothers from anationally representative sample experienced greater family-to-work conflict than fathers

when they had a young child at home. Other findings, however, found greater negative

spillover from home to work for men (Barnett & Marshall, 1992; Forthofer, Markman,

Cox, Stanley, & Kessler, 2004).

The presence and ages of children are potentially important factors in understanding

the inconsistencies in these findings. Both men and women with a child under the age of

6 and those with a child aged 6�18 had higher levels of negative spillover from family to

work than those who were childless (Grzywacz, Almeida, & McDonald, 2002; Winslow,

168 E. J. Hill et al.

2005). But mothers in the life stage of having young children seemed particularly

vulnerable to family-to-work conflict. The presence of preschool children significantly

predicted greater family-to-work conflict for women (Hundley, 2001), and mothers of

young children were more likely to miss days of work than their male counterparts

(Crouter, 1984; Dilworth, 2004).

A life stage that included elder care responsibilities was also associated with increased

family-to-work conflict (Anastas, Gibeau, & Larson, 1990), particularly for women (Gutek

et al., 1991). Recent analyses, however, found identical levels of family-to-work conflict for

male and female caregivers (Barrah, Shultz, Baltes, & Stoltz, 2004). Women and men

sandwiched between both early child care and elder care responsibilities have been

identified as particularly likely to experience family-to-work conflict (Barrah et al., 2004).

Flexibility and family-to-work conflict

Access to and use of flexible work arrangements has consistently been associated with

reduced conflict in the work�family interface (Hill, Martinson, & Ferris, 2004). Workers

who had greater access to flexible work arrangements were significantly more likely to

report no interference with job and family life and low levels of family-to-work conflict in a

study of a large sample of employed and self-employed workers (Bond et al., 2002). But it is

not known whether the strength of this relationship differs by gender and life stage. The

strongest reductions of family-to-work conflict seemed to involve flexibility that allowed a

reduction in the number of hours spent at work (Rau, 2006). Women may be particularly

likely to benefit from this type of flexibility because of the increased likelihood of their

experiencing family-to-work conflict, particularly during the peak child bearing years.

Stress and burnout for men and women at different life stages

As with family-to-work conflict, stress and burnout has increasingly been identified as the

psychological difficulty resulting from strain in the work�family interface (Maslach, 2006).

Defined as ‘a psychological syndrome that involves a prolonged response to chronic

interpersonal stressors’, it involves overwhelming exhaustion, feelings of cynicism and

detachment, and a sense of ineffectiveness and lack of accomplishment (Maslach, 1993,

p. 37). Studies of burnout have not systematically studied differences between men and

women (Rossi, 2006). Most previous studies have been done with American male

professionals (Rossie, Perrewe, Sauter, & Jex, 2006). But a growing body of research

suggests that family-to-work conflict is related to burnout (Halbesleben & Zellars, 2006).

Thus, women may be more likely to experience stress and burnout at particular stages in

the life cycle when there is greater strain resulting from caregiving and household

responsibilities (Dilworth, 2004).

Relationship between workplace flexibility and stress and burnout

Studies of the relationship between workplace flexibility and stress and burnout have

repeatedly identified positive mental and emotional health outcomes. Bond et al.’s

analysis (2002) of a nationally representative sample found that employees with more

access to flexible work arrangements reported fewer mental health problems. A similar

analysis found positive mental health outcomes related to access to flexibility (Jacob

et al., 2008). Scheduling practices that afforded high levels of flexibility showed a

Community, Work & Family 169

significant direct effect on reduced somatic health complaints, and enhanced mental and

physical health outcomes (Thomas & Ganster, 1995). In this special issue Grzywacz,

Carlson, and Shulkin (2008) reported that schedule flexibility was associated with lower

stress and burnout. Other mental and emotional benefits were found to be associated

with telecommuting in a study of employees from large companies (Richman, Noble, &

Johnson, 2002). Those who used telework experienced less emotional and physical fatigue

than their on-site counterparts and were less likely to report that they were drained after

work (Richman et al., 2002).

These studies provide little information, however, about the relationship between

flexibility and stress and burnout for women and men at different life stages. Findings

indicating that women are more likely to highly value flexible work arrangements may

suggest potentially greater benefits to women than men (Catalyst, 2001; Radcliffe Public

Policy Center with Harris Interactive, 2000). Similarly women and men in life stages

that involve greater caregiving responsibilities for dependent children or elders may be

more likely to experience mental and emotional benefits relative to those in other life

stages.

The findings in this paper indicate the need for further exploration of differences in the

use of and value placed upon workplace flexibility options by women and men at different

life stages as well as the need for further understanding of how the relationship between

gender, life stage, and use of flexibility may be related to family-to-work conflict and stress

and burnout.

Research question and hypotheses

R1: Explore how the use of and value placed upon workplace flexibility options vary bygender and life stage.

H1: Having the flexibility one needs will negatively predict family-to-work conflict andstress and burnout.

H2: Gender (being female) will predict greater family-to-work conflict and greater stressand burnout; and adding it to the models will significantly increase the proportion ofthe variance explained.

H3: Life stage will predict family-to-work conflict and stress and burnout, and addingthem to the models will significantly increase the proportion of the variance explained.

H4: Life stage by gender interactions will predict family-to-work conflict and stress andburnout, and adding them to the models will significantly increase the proportion ofthe variance explained.

Method

Sample

Data for these analyses consist of several sub-samples from a large multi-company dataset

created by WFD Consulting. This dataset is described thoroughly earlier in this special

issue (Civian et al., 2008). The largest sub-sample used in this paper includes 41,118

observations representing 143,391 employees in 20 distinct companies (4 pharmaceuticals,

3 technology, 4 manufacturing, 5 financial and professional services, and 4 universities).

Men (53%) and women (47%) are nearly equally represented in the sample. About three-

quarters (77%) are married/partnered. Nearly one-half of the respondents (48%) have a

170 E. J. Hill et al.

spouse/partner who works full time, about one-fifth (19%) have a spouse/partner who is

not employed, about one-tenth (10%) have a spouse/partner who is employed part time,

and about one-fourth (23%) do not have a spouse/partner. About half (48%) of the

participants have children under the age of 18. In addition, 11% have dependent elders,

91% work full time, 46% are non-exempt employees who are eligible for overtime pay while

54% are paid fixed salaries regardless of hours worked, and 59% had been employed by

their company for more than five years.

Measures

Family-to-work conflict. Family-to-work conflict was measured with a five-item index

measuring the frequency with which family responsibilities affect work (range of 1�4) (see

Civian et al., 2008). Higher values indicate greater family-to-work conflict.

Stress and burnout. Stress and burnout was measured with a validated seven-item index

that has a range from 0 to 14 (see Civian et al., 2008). Higher values indicate greater stress

and burnout.

Perceived flexibility. Perceived flexibility was assessed with a single item using a binary (yes/

no) response option (see Civian et al., 2008). Higher values indicate greater perceived

flexibility at work.

Gender. Gender was coded as a dummy variable: males were coded as 0 and females were

coded as 1.

Life stage. Central to these analyses is the concept of life stage. It is designed to capture

the ebb and flow of family responsibilities over the life course. Those identified in the

first life stage are those who do not have children and are under the age of 35. The next

three life stages are defined by the ages of the respondents’ children. The second lifestage is delineated by the presence of children ages 0�5 and no older children. The third

life stage is characterized by the presence of children ages 0�5 (preschool) as well as

children ages 6�17 (school-age). The fourth life stage is differentiated by the presence of

children ages 6�17 only (no younger children). The fifth and final life stage identifies

those who do not have children under age 18 and are age 45 and older. This age was

selected as a cut-off point after which it would be unlikely for the respondent to have a

first child.

Covariates. The covariates, or control variables, in this study include marital/earner status,

elder care responsibilities, part-time/full-time status, salary status (exempt vs. non-exempt),

tenure, and supportive policies. These variables were selected because they have beenshown to be related to employees’ access to flexible work options (Pitt-Catsouphes &

Matz-Costa, 2008). Marital/earner status was measured using a series of dummy variables

representing four categories: (1) not married/partnered; (2) married/partnered, partner not

employed; (3) married/partnered, partner employed full time; and (4) married/partnered,

partner employed part time. The category ‘not married/partnered’ served as a reference

group. Covariates were dummy coded as follows: elder care status (1�elder care

responsibility, 0�none); full-time work schedule (1�full time, 0�part time); salary

status (1�exempt, 0�non-exempt); and tenure (1�five or more years, 0�less than fiveyears). Supportive policies were measured by one of the items in the WFD validated Work�Life Support Index: ‘The company’s policies are supportive of my personal/family

responsibilities’ (0�disagree, 1�agree). Descriptive statistics for the dependent and

independent variables and the covariates are summarized in Table 1. Variables without

Community, Work & Family 171

sub-sample Ns are summarized based on their inclusion in the analysis for the Stress and

Burnout Index.

Analyses

To answer the first research question, we created dummy variables for the value(1�valued, 0�not valued) and use (1�using, 0�not using) of each of the following

flexibility options: flextime, compressed work week, telecommuting, part time, and job

sharing. For each flexibility option, respondents were asked if they were currently using the

option and the value of each option on a scale from (1) ‘little or no value’, (2) ‘some value’,

or (3) ‘great value’. For the analysis, we defined those of ‘great value’ as ‘valued’�1. All

respondents were asked about the value of each flexibility option, not just those who

currently used the option. We then compared the means of males and females for each of

the five life stages. Because of the large sample sizes we used effect sizes to determinewhether meaningful differences existed. Effect sizes were calculated by subtracting the male

mean from the female mean and dividing by the pooled standard deviation. Effect sizes of

0.20 or more were considered meaningful (see Cohen, 1988). The number of studies

examining the current use of flexibility options ranged from 7 studies (N�84,493) to

8 (N�98,762), and for value of flexibility options the number ranged from 13 studies

(N�67,985) to 18 (N�121,280).

To test hypotheses 1�4 we followed the general plan for using hierarchical linear

modelling (HLM) outlined for all papers in this special issue (see Civian et al., 2008).

Table 1. Descriptive statistics for analysis variables.a

Mean SD Range

Stress and Burnout Index (N�143,341) 3.00 3.09 0�14

Family-to-work conflict (N�65,427) 1.59 0.59 1�4

Expected retention (N�130,628) 0.86 0.35 0�1

Perceived flexibility 0.83 0.37 0�1

Marital/earner status

Not married/partnered 0.23 0.42 0�1

Married/partnered, partner not employed 0.19 0.39 0�1

Married/partnered, partner employed full time 0.48 0.50 0�1

Married/partnered, partner employed part time 0.10 0.30 0�1

Elder care status 0.11 0.31 0�1

Part-time or full-time work (full time�1) 0.91 0.29 0�1

Salary status (exempt�1) 0.54 0.50 0�1

Tenure (five years or more�1) 0.59 0.49 0�1

Company policies supportive of personal responsibilities

(agree�1)

0.68 0.47 0�1

Gender (female�1) 0.47 0.50 0�1

Life stage

Under age 35 with no children 0.24 0.43 0�1

Oldest child B6 years old 0.14 0.35 0�1

Children B6 and 6� 0.09 0.28 0�1

Only school-age children 6�17 0.25 0.43 0�1

Age 45 or older with no children 0.29 0.45 0�1

aThe Ns provided are for the sub-sample of data for the analysis that uses the accompanying variable as thedependent variable.

172 E. J. Hill et al.

Estimations began with a model that included only the covariates and then the analytical

variables of interest (gender, life stage, and life stage by gender) were added one step at a

time to determine the change in the proportion of the variance explained beyond what was

accounted for by the covariates. Following Kreft and de Leeuw’s recommendation (1998),

we calculated the proportion of the variance explained at each step by subtracting the s2 of

a model that includes only the independent variables of interest from the s2 of a model that

includes only the covariates and dividing by the s2 of a model that includes only the

covariates. Deviance tests were performed to determine if the change in the proportion ofvariance explained was statistically significant.

Results

In this part of the results section we answer research question 1 by exploring the differences

by gender at the five different life stages related to the reported use and perceived value of

workplace flexibility (see Table 2). We then test hypotheses 1�4 using HLM (see Tables 3

and 4).

Differences by gender and life stage on the use of workplace flexibility

As shown in Table 2, data reveal a curvilinear relationship in gender differences by life

stage in the use of workplace flexibility. In Life Stage 1 (B 35 years old and no children)

there were no meaningful gender differences in the use of workplace flexibility. However, in

Life Stage 2 (oldest childBsix years old) there were a number of gender differences.

Women were more likely than men to use part-time work, job sharing, flextime, and

telecommuting. The meaningful effect sizes for the gender differences for this life stageranged from 0.23 to 0.48. In Life Stage 3 (at least one childBsix years old and at least one

child six plus years old), women were even more likely than men to use these flexible work

arrangements. The effect sizes for differences in this life stage ranged from 0.24 to 0.59. In

Life Stage 4 (no childrenBsix years old and at least one child six plus years old), the

differences between men and women were less pronounced. Women were still more likely

than men to use part-time work (ES�0.45) and job sharing (ES�0.32), but there were no

other meaningful gender differences. In Life Stage 5 (age 45� and no children), the

differences virtually disappeared altogether.

Differences by gender by life stage on the perceived value of workplace flexibility

There were many more meaningful differences between men and women in the perceived

value of workplace flexibility options. At every life stage women valued almost every

flexible work option more than men valued them (see Table 2). These differences showed a

curvilinear pattern with the largest differences being in Life Stages 2, 3, and 4, and

somewhat smaller differences in Life Stages 1 and 5. It is important to note that thoughdifferences exist, flextime is the type of workplace flexibility valued most by men and by

women at every life stage. Depending on life stage, 48�63% of the men and 64�85% of the

women valued flextime.

Hypothesis testing

H1: Perceived flexibility will predict less family-to-work conflict and less stress and burnoutin all models.

Community, Work & Family 173

Table 2. Differences between means of female and male respondents on the value and use of workplace flexibility by life stage.

Life Stage 1

(no children B35)

Life Stage 2

(oldest child B6)

Life Stage 3

(children B6 and 6�)

Life Stage 4

(children 6�17 only)

Life Stage 5

(no children 45�)

M F ES M F ES M F ES M F ES M F ES

Use of workplace flexibility

Flextime 0.14 0.12 �0.06 0.10 0.19 0.25* 0.14 0.24 0.25* 0.14 0.19 0.14 0.14 0.17 0.08

Compressed work week 0.03 0.04 0.05 0.03 0.04 0.06 0.05 0.03 �0.10 0.04 0.04 0.00 0.05 0.05 0.00

Telecommuting 0.03 0.04 0.05 0.04 0.10 0.23* 0.06 0.13 0.24* 0.10 0.06 �0.15 0.05 0.08 0.12

Part time 0.01 0.01 0.00 0.00 0.12 0.48* 0.01 0.19 0.59** 0.00 0.10 0.45* 0.01 0.03 0.14

Job sharing 0.00 0.00 0.00 0.01 0.06 0.28* 0.00 0.16 0.58** 0.00 0.05 0.32* 0.02 0.00 �0.22*

Value of workplace flexible

Flextime 0.61 0.70 0.19 0.63 0.78 0.33* 0.55 0.85 0.66** 0.54 0.76 0.46* 0.48 0.64 0.32*

Compressed work week 0.41 0.53 0.24* 0.43 0.59 0.32* 0.38 0.57 0.38* 0.34 0.51 0.34* 0.28 0.45 0.35*

Telecommuting 0.46 0.61 0.30* 0.54 0.74 0.42* 0.43 0.74 0.63** 0.42 0.64 0.44* 0.35 0.44 0.18

Part time 0.22 0.45 0.48** 0.19 0.56 0.76** 0.17 0.63 0.93*** 0.17 0.44 0.58** 0.18 0.33 0.34*

Job sharing 0.17 0.40 0.50*** 0.14 0.48 0.73** 0.14 0.56 0.87*** 0.15 0.39 0.54** 0.13 0.23 0.26*

*Effect size�0.20�0.49; **effect size�0.50�0.79; ***effect size ]0.80.

17

4E

.J.

Hill

eta

l.

Table 3. Estimated associations of perceived flexibility, gender, life stage, and gender�life stage with family-to-work conflict.

Family-to-work conflict Family-to-work conflict Family-to-work conflict Family-to-work conflictEstimated coefficient (SE) Estimated coefficient (SE) Estimated coefficient (SE) Estimated coefficient (SE)

DemographicsMarital/earner status

Not married/partnered Reference Reference Reference ReferenceMarried/partnered, partner not employed �0.019 (0.015) �0.024 (0.015) �0.044* (0.018) �0.031$ (0.018)Married/partnered, partner employed part time �0.005 (0.011) �0.005 (0.011) �0.025* (0.013) �0.027* (0.013)Married/partnered, partner employed full time 0.005 (0.018) �0.001 (0.019) �0.026 (0.021) �0.015 (0.021)

Elder care responsibilities (yes�1) 0.068*** (0.015) 0.069*** (0.015) 0.092*** (0.017) 0.091*** (0.017)

Work demographicsFull-time work schedule (yes�1) �0.062*** (0.018) �0.063*** (0.018) �0.056** (0.019) �0.052** (0.019)Salary status (exempt�1) 0.062*** (0.010) 0.059*** (0.010) 0.049*** (0.011) 0.049*** (0.011)Tenure (5 years or more�1) 0.084*** (0.010) 0.085*** (0.010) 0.110*** (0.012) 0.111*** (0.012)

Supportive policies �0.178*** (0.011) �0.179*** (0.011) �0.143*** (0.012) �0.143*** (0.012)Perceived flexibility �0.062*** (0.013) �0.062*** (0.013) �0.071*** (0.014) �0.071*** (0.014)Gender (female�1) �0.019$ (0.011) �0.014 (0.012) �0.039* (0.019)

Life stageLife Stage 1 (no children, age B35) Reference ReferenceLife Stage 2 (oldest child B6) 0.075*** (0.017) 0.041$ (0.023)Life Stage 3 (children B6 and 6�) 0.060** (0.021) 0.005 (0.028)Life Stage 4 (children 6�17 only) 0.018 (0.015) �0.019 (0.021)Life Stage 5 (no children, age 45�) �0.102*** (0.016) �0.081*** (0.022)

Life stage�Gender interactionsLife Stage 1�Gender interaction ReferenceLife Stage 2�Gender interaction 0.067* (0.033)Life Stage 3�Gender interaction 0.113** (0.040)Life Stage 4�Gender interaction 0.070* (0.028)Life Stage 5�Gender interaction �0.037 (0.029)

Constant 1.734*** (0.046) 1.747*** (0.046) 1.733*** (0.044) 1.740*** (0.045)Proportion of variance explained 0.0979 0.0995 0.1171 0.1190Change in proportion of variance explained 0.0015 0.0176 0.0019

$pB0.10; *pB0.05; **pB0.01; ***pB0.001 (two-tailed).

Co

mm

un

ity,W

ork

&F

am

ily1

75

Table 4. Estimated associations of perceived flexibility, gender, life stage, and gender�life stage with stress and burnout.

Stress and burnout Stress and burnout Stress and burnout Stress and burnout

Estimated coefficient (SE) Estimated coefficient (SE) Estimated coefficient (SE) Estimated coefficient (SE)

Demographics

Marital/earner status

Not married/partnered Reference Reference Reference Reference

Married/partnered, partner not employed �0.062 (0.044) 0.077$ (0.045) �0.106* (0.048) �0.103* (0.048)

Married/partnered, partner part time 0.073* (0.035) 0.073* (0.035) �0.060$ (0.036) �0.056 (0.036)

Married/partnered, partner full time 0.101* (0.052) 0.246*** (0.053) 0.077 (0.055) 0.088 (0.056)

Elder care responsibilities (yes�1) 0.444*** (0.044) 0.426*** (0.044) 0.494*** (0.044) 0.492*** (0.044)

Work demographics

Full-time work schedule (yes�1) 0.227*** (0.051) 0.291*** (0.054) 0.351*** (0.054) 0.348*** (0.054)

Salary status (exempt�1) 1.360*** (0.030) 1.409*** (0.030) 1.388*** (0.030) 1.388*** (0.030)

Tenure (5 years or more�1) 0.639*** (0.030) 0.622*** (0.030) 0.638*** (0.033) 0.640*** (0.033)

Supportive policies �1.535*** (0.032) �1.543*** (0.032) �1.534*** (0.032) �1.534*** (0.032)

Perceived flexibility �1.633*** (0.039) �1.625*** (0.039) �1.622*** (0.039) �1.623*** (0.039)

Gender (female�1) 0.418*** (0.031) 0.407*** (0.031) 0.378*** (0.056)

Life stage

Life Stage 1 (no children, age B35) Reference Reference

Life Stage 2 (oldest child B6) 0.412*** (0.049) 0.431*** (0.067)

Life Stage 3 (children B6 and 6�) 0.555*** (0.058) 0.572*** (0.077)

Life Stage 4 (children 6�17 only) 0.304*** (0.044) 0.235*** (0.060)

Life Stage 5 (no children, age 45�) �0.098* (0.044) �0.115* (0.060)

Life stage�Gender interactions

Life Stage 1�Gender interaction Reference

Life Stage 2�Gender interaction �0.053 (0.093)

Life Stage 3�Gender interaction �0.056 (0.110)

Life Stage 4�Gender interaction 0.145$ (0.080)

Life Stage 5�Gender interaction 0.026 (0.076)

Constant 4.032*** (0.170) 3.727*** (0.179) 3.626*** (0.178) 3.641*** (0.179)

Proportion of variance explained 0.1508 0.1487 0.1560 0.1562

Change in proportion of variance explained �0.0022 0.0073 0.0001

$pB0.10; *pB0.05; **pB0.01; ***pB0.001 (two-tailed).

17

6E

.J.

Hill

eta

l.

Analyses provided universal support for hypothesis 1. Perceived flexibility (i.e., having the

flexibility one needs) significantly (pB0.001) predicted less family-to-work conflict (see

Table 3) and less stress and burnout (see Table 4) in each of the HLM models.

H2: Gender (being female) will predict greater family-to-work conflict and greater stress andburnout and adding it to the models will significantly increase the proportion of thevariance explained.

Analyses provided partial support for this hypothesis. There was no main effect of gender on

family-to-work conflict (see Table 3); when added to the model predicting family-to-work

conflict, gender did not significantly improve the model fit over the previous model (x2�1.61983, df�1, ns). As described below in H4, it was found, however, that gender interacted

with life stage in predicting family-to-work conflict. In harmony with the hypothesis, gender

(being female) predicted significantly greater stress and burnout (see Table 4) in all of the

models (pB0.001). When added to the model predicting stress and burnout, gender

increased the proportion of the variance explained by 0.22% and significantly improved the

model fit over the previous model (x2�178.7933, df�2, pB0.000).

H3: Life stages will predict family-to-work conflict and stress and burnout and adding themto the models will significantly increase the proportion of the variance explained.

When added to the model predicting family-to-work conflict, life stages increased the

proportion of the variance explained by 1.76% and significantly improved the model fit

over the previous model (x2�129.261, df�4, pB0.000). Being in Life Stage 2 (pB0.001)

or Life Stage 3 (pB0.01), the two stages including childrenBsix years of age, was

associated with greater work-to-family conflict than the reference group (Life Stage 1).

Being in Life Stage 5 (being at least 45 years old without any children) was associated with

less work-to-family conflict than the reference group (pB0.001). When added to the model

predicting stress and burnout, the life stage variables increased the proportion of the

variance explained by 0.73% and significantly improved the model fit over the previous

model (x2�256.446, df�4, pB0.000). Life Stage 2, Life Stage 3, and Life Stage 4 were all

associated with significantly greater stress and burnout than the reference group in Life

Stage 1 (pB0.001). Life Stage 5 was associated with significantly less stress and burnout

than the reference group (pB0.05).

H4: Life stage by gender interactions will predict family-to-work conflict, and stress andburnout, and adding them to the models will significantly increase the proportion of thevariance explained.

When added to the model predicting family-to-work conflict, the life stage by gender

(being female) interactions increased the proportion of the variance explained by 0.19%

and significantly improved the model fit over the previous model (x2�24.134, df�4, pB

0.000). Life Stage 2 (pB0.05), Life Stage 3 (pB0.01), and Life Stage 4 (pB0.01), the three

life stages including children in the home, were associated with greater family-to-work

conflict. When added to the models predicting stress and burnout, the life stage by gender

interactions did not significantly improve the model fit over the previous model (x2�10.651, df�4, ns).

Discussion

Workplace flexibility, family-to-work conflict, and stress and burnout have frequently

been included in research models of the work�family interface. In addition, it is not

Community, Work & Family 177

uncommon for gender to be included and found to influence such models. However, life

stage (as defined by presence or absence of children and their ages) and its interaction

with gender, have not been utilized. A major purpose of this study is to investigate the

applicability of life stage and gender to research about workplace flexibility. A second

purpose is to look at how the use and value placed upon workplace flexibility varies by

gender and by life stage.

The most important finding of this study is that gender, life stage, and their interaction

do matter in predicting family-to-work conflict and stress and burnout in conjunction withworkplace flexibility, and that life stage matters most. The addition of gender significantly

improved the fit of the model of stress and burnout, and addition of the gender by life

stage interactions significantly improved the fit of the model predicting family-to-work

conflict, but addition of life stage significantly improved the fit of both models. In

addition, the largest changes in the proportion of the variance explained in each model

emerged when life stage was added.

These data reveal that men and women are not homogeneous groups when it comes to

workplace flexibility. Gender can best be understood in the context of life stage. In lifestages without children in the home, men and women are more similar in their use and

value of workplace flexibility options than in life stages with children in the home. It

appears that because women shoulder greater responsibility for child care, they are more

likely to use and value these options than men while children are in the home.

Results revealed that over the life course there was a curvilinear relationship in

differences between men and women in the use of workplace flexibility options. There

were no meaningful differences between young, childless men and women, yet mothers of

preschoolers were more likely than fathers of preschoolers to use flexibility options. Themagnitude of these gender differences increased when both preschoolers and school-age

children were in the home. The gender differences decreased again for those with just

school-age children and almost disappeared altogether for older, childless men and

women.

It is interesting to note the men have very little interest in and almost no use of

options that would decrease their pay (i.e., job sharing, part-time employment). While

women are fairly similar to men in not using these options during Life Stages 1 and 5,

during the life stages in which children are in the home women value and use part-timework and job sharing more than other flexible work options. This may be an indication

that in couple relationships the provider role is still gendered, at least for men, so they do

not want to choose options that would reduce their pay (e.g., part-time employment, job

sharing).

An important finding of these data is that flextime is the type of workplace flexibility

valued most by men and by women at every life stage. Depending on life stage, 48�63% of

the men and 64�85% of the women valued flextime. This is important information for

policy makers because flextime is relatively easy to implement, is inexpensive, and can havebeneficial outcomes (Grzywacz et al., 2008). For example, flextime reduces stress because it

better enables employees to mesh the circadian rhythms of their own personal and family

life with the needs of the business. These data support the adoption of flextime programs as

the first way to implement workplace flexibility.

Limitations

There are limitations to the findings of this study. First, the data are cross-sectional and

inferences about the influences of life stage would be best tested with a longitudinal design.

178 E. J. Hill et al.

It is possible that some of the findings reflect cohort effects, rather than the influence of life

stage. However, in the study of workplace flexibility and life stage we do not have the

option of a longitudinal design because options like telecommuting have not been available

throughout the life course of the present cohort of workers. Second, the size of some of the

associations of gender, life stage, and their interactions though significant, are quite small.

Likewise, the significant increases in the proportion of the variance explained after adding

these variables are also small in magnitude, and in no case more than 2%. Still, because we

have such a large sample even small changes can achieve statistical significance. Third,generalizability of the results may be limited because data were collected from a set of

companies that sought the services of a consulting company and fielded a survey asking

questions about workplace flexibility, family-to-work conflict, and stress and burnout. This

critique can be minimized because the data represent a spectrum of companies from

diverse industries. Fourth, and finally, the data were collected between 1996 and 2006.

Companies are continually improving workplace flexibility options and so data from 1996

may be dated.

Conclusion

Academics and policy makers seek to understand workplace flexibility and its antecedents

and consequences to better predict the work�family interface as well as inform workplace

flexibility policy initiatives. These findings suggest that gender, life stage, and especially

their interaction are important in the quest for that understanding. Men and women are

not homogeneous groups on matters related to workplace flexibility, and can only be

understood when life stage is taken into account.

Acknowledgements

The research was supported by a grant from the Alfred P. Sloan Foundation (2005-3-33) and by theFamily Studies Center in the School of Family Life at Brigham Young University. We acknowledgethe contribution of Ann Crouter, at the Pennsylvania State University, for her contribution inoriginally conceptualizing this article.

Notes on contributors

E. Jeffrey Hill, PhD, is associate professor of family life at Brigham Young University where he

teaches in the School of Family Life and the Marriott School of Management. He has published

more than 40 scholarly articles and book chapters on work and family, with an emphasis on the

influence of flexible work arrangements on work and family outcomes.

Jenet I. Jacob, PhD, received a doctorate in Family Social Science at the University of Minnesota and

is an associate professor of family life at Brigham Young University. She publishes articles on

mothering and work and family.

Laurie L. Shannon, PhD, received a doctorate in Social and Developmental Psychology from

Brandeis University and is a research psychologist for WFD Consulting. Her research focuses on

work�life issues of flexibility, engagement, stress and burnout, and dependent care.

Robert T. Brennan, EdD, is a research associate at Harvard Medical School. He has been studying

issues pertaining to work and family for over 15 years, particularly in the context of dual-earner

couples, where he is known for his contributions to the analysis of data from dyads using multilevel

modeling.

Community, Work & Family 179

Victoria L. Blanchard, MS, received a Master’s degree in Marriage, Family, and Human

Development from Brigham Young University and is project manager for the IBM�BYU Work�Life Data Analysis venture.

Giuseppe Martinengo, PhD, received a doctorate in Marriage, Family, and Human Development

from Brigham Young University and now works for the More Good Foundation.

References

Anastas, J. W., Gibeau, J. L., & Larson, P. J. (1990). Working families and eldercare: A nationalperspective in an aging America. Social Work, 35(5), 405�411.

Barnett, R. C., & Marshall, N. L. (1992). Men’s job and partner roles: Spillover effects andpsychological distress. Sex Roles, 27, 455�472.

Barrah, J. L., Shultz, K., Baltes, B., & Stoltz, H. E. (2004). Men’s and women’s eldercare-based work�family conflict: Antecedents and work-related outcomes. Fathering, 2(3), 305�330.

Billings, J. R., & Sharpe, D. L. (1999). Factors influencing flextime usage among employed marriedwomen. Consumer Interests Annual, 45, 89�95.

Bond, J., Galinsky, E., Kim, S., & Brownfield, E. (2005). National study of employers: Highlights offindings. New York: Families and Work Institute.

Bond, J. T., Thompson, C., Galinsky, E., & Prottas, D. (2002). Highlights of the national study of thechanging workforce. New York: Families and Work Institute.

Catalyst (2001). The next generation: Today’s professionals, tomorrow’s leaders. New York: Catalyst.Civian, J. T., Richman, A. L., Shannon, L. L., Shulkin, S., & Brennan, R.T. (2008). Using a multi-

organization database: Research methods, strengths and limitations. Community, Work & Family,11(2), 139�148.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences, 2nd ed. New York: AcademicPress.

Comfort, D., Johnson, K., & Wallace, D. (2003). Part-time work and family friendly practices inCanadian workplaces (The Evolving Workplace Series). Statistics Canada. Retrieved January 12,2008, from http://www.statcan.ca/english/freepub/71-584-MIE/free.htm

Crouter, A. C. (1984). Spillover from family to work: The neglected side of the work�family interface.Human Relations, 37, 425�442.

Dilworth, J. E. L. (2004). Predictors of negative spillover from family to work. Journal of FamilyIssues, 3(25), 241�261.

DiNatale, M., & Boraas, S. (2002). The labor force experience of women from ‘Generation X’.Monthly Labor Review, 125(3), 3�15.

Forthofer, M. S., Markman, H. J., Cox, M., Stanley, S., & Kessler, R. C. (1996). Associations betweenmarital distress and work loss in a national sample. Journal of Marriage and the Family, 58, 597�605.

Galinsky, E., Bond, J. T., & Hill, E. J. (2004). When work works: A status report on workplaceflexibility: Who has it? Who wants it? What difference does it make? Retrieved November 21, 2005,from http://familiesandwork.org/3w/research/downloads/status.pdf

Golden, L. (2001). Flexible work schedules: What are we trading off to get them? Monthly LaborReview, 124(3), 50�67.

Greenhaus, J. H., & Beutell, N. J. (1985). Sources of conflict between work and family roles. Academyof Management Review, 10, 76�88.

Grzywacz, J. G., Almeida, D. M., & McDonald, D. A. (2002). Work�family spillover and dailyreports of work and family stress in the adult labor force. Family Relations, 51, 28�36.

Grzywacz, J. G., Carlson, D. S., & Shulkin, S. (2008). Schedule flexibility and stress: Linking formalflexible arrangements and perceived flexibility to employee health. Community, Work & Family,11(2), 199�214.

Gutek, B., Searle, S., & Klepa, L. (1991). Rational versus gender role expectations for work�familyconflict. Journal of Applied Psychology, 76, 560�568.

Hakim, C. (2002). Lifestyle preferences as determinants of women’s differentiated labour marketcareers. Work and Occupations, 29, 428�459.

Halbesleben, J. R. B., & Zellars, K. L. (2006). Stress and the work�family interface. In A. M. Rossie,P. L. Perrewe, S. L. Sauter, & S. M. Jex (Eds.), Stress and quality of working life (pp. 53�68).Charlotte, NC: Information Age Publishing.

180 E. J. Hill et al.

Hill, E. J., Grzywacz, J. G., Allen, S., Blanchard, V. L., Matz-Costa, C., Shulkin, S., et al. (2008).Defining and conceptualizing workplace flexibility. Community, Work & Family, 11(2), 149�163.

Hill, E. J., Jackson, A. D., & Martinengo, G. (2006). Twenty years of work and family atInternational Business Machines Corporation. American Behavioral Scientist, 49, 1165�1183.

Hill, E. J., Martinson, V., & Ferris, M. (2004). New-concept part-time employment as a work�familyadaptive strategy for women professionals with small children. Family Relations, 53, 282�292.

Hill, E. J., Martinson, V., Hawkins, A., & Ferris, M. (2003). Studying ‘working fathers’: Comparingfathers’ and mothers’ work�family conflict, fit, and adaptive strategies in a global high-techcompany. Fathering, 1, 239�261.

Hill, E. J., Mead, N. T., Dean, L. R., Hafen, D. M., Gadd, R., Palmer, A. A., et al. (2006).Researching the 60-hour dual-earner work week: An alternative to the ‘opt out revolution’.American Behavioral Scientist, 49, 1184�1203.

Hundley, G. (2001). Domestic division of labor and self/organizationally employed differences in jobattitudes and earnings. Journal of Family and Economic Issues, 22, 121�139.

Jacob, J. I., Bond, J. T., Galinsky, E., & Hill, E. J. (2008). Six critical ingredients in creating aneffective workplace. The Psychologist Manager Journal, 11(1), 141�161.

Kreft, I. & de Leeuw, J. (1998). Introducing multileveling modelling. Thousands Oaks, CA: Sage.Maslach, C. (1993). Burnout: A multidimensional perspective. In W. B. Schaufeli, C. Maslach &

T. Marek (Eds.), Professional burnout: Recent developments in theory and research (pp. 19�32).Washington, DC: Taylor & Francis.

Maslach, C. (2006). Understanding job burnout. In A. M. Rossie, P. L. Perrewe, S. L. Sauter &S. M. Jex (Eds.), Stress and quality of working life (pp. 37�52). Charlotte, NC: Information AgePublishing.

Mattessich, P., & Hill, R. (1987). Family development and life cycle research and theory revisited. InM. Sussman & S. Steinmetz (Eds.), Handbook of marriage and family research. New York: PlenumPress.

Moen, P., & Sweet, S. (2002). Two careers, one employer: Couples working for the same corporation.Journal of Vocational Behavior, 61(3), 466�483.

Moen, P., & Sweet, S. (2004). From ‘Work-family’ to ‘flexible careers’: A life course reframing.Community, Work & Family, 7(2), 209�225.

Pitt-Catsouphes, M., & Matz-Costa, C. (2008). The multi-generational workforce: Workplaceflexibility and engagement. Community, Work & Family, 11(2), 215�229.

Pleck, J. H. (1977). The work�family role system. Social Problems, 24, 417�427.Radcliffe Public Policy Center with Harris Interactive. (2000). Life’s work: Generational attitudes

toward work and life integration. Cambridge, MA: Radcliffe Public Policy Center.Rau, R. (2006). The association between blood pressure and work stress: The importance of

measuring isolated systolic hypertension. Work and Stress, 20(1), 84�97.Richman, A., Noble, K., & Johnson, A. (2002). When the workplace is many places. The extent and

nature of off-site work today. Watertown, MA: WFD Consulting Inc.Rossie, A. M., Perrewe, P. L., Sauter, S. L., & Jex, S. M. (2006). Stress and quality of working life:

Current perspectives in occupational health. Charlotte, NC: Information Age Publishing.Thomas, L. T., & Ganster, D. C. (1995). Impact of family-supportive work variables on work�family

conflict and strain: A control perspective. Journal of Applied Psychology, 80, 6�15.US Census Bureau. (2000). Working at home: 2000 (PHC-T-35). Selected characteristics of workers

who worked at home by sex for the United States: 2000. Retrieved January 21, 2008, from http://www.census.gov/population/cen2000/phc-t35/tab01Table 2.xls

Wharton, A. S., & Blair-Loy, M. (2002). The ‘overtime culture’ in a global corporation. Work andOccupations, 29(1), 32�63.

Winslow, S. (2005). Work�family conflict, gender, and parenthood, 1977�1997. Journal of FamilyIssues, 26(6), 727�755.

Community, Work & Family 181